Working on hal9.ai -- Long term, a Roblox for AI; short term, a Python customizable ChatGPT that is enterprise ready. Think of ChatGPT without the LLM and support for writing your own RAG.
https://www.inclusivecolors.com/
> Instead of only working with a handful of colors, you can create a whole palette of swatches at the same time so you can see if they look good together.
> Precise control of every shades/tints in each swatch rather than being limited by autogenerated colors.
> See which color pairs contrast as you edit so you can create a palette with built-in WCAG accessibility. This way you can plan in advance which foreground colors (for headings, body text, form fields and so on) should contrast on which background colors, so you can avoid running into surprise low contrast issues later when designing.
Currently working on adding webhook notifications for status pages.
The scale is RNG worlds like Minecraft. I've never seen that before with a Raycaster.
Here is my progress so far (I've had a month break)
https://github.com/con-dog/chunked-z-level-raycaster/blob/ma...
Not for profit, just for fun and exploration
Typed Relational Database access:
(and some philosophy on the subject :))
https://github.com/fedi-e2ee/public-key-directory-specificat...
The content needs some work, but I'm pretty happy with the framework / UX. I would love to get any feedback from folks who check it out!
(The first section is just multi-guess questions as part of the introductory content. Try any other section to get the full in-browser-code-execution experience, which uses client-side Pyodide under the hood.)
My library is specialized for parsing text. That had enabled some cool capabilities.
It comes with a `Span` primitive, which tracks where in a file a token came from, for implementing error messages. A `Span` can be the input or the output of a parser. At the front end a `Span` is an entire file, and as you slice and dice it, it tracks the metadata of where it came from.
Along with the standard `Sequence` (combining parsers in a set order) and `Choice` operations (branching between many parsers) that parser combinators are built around, I have come up two operations that are very handy. I suspect that others have made them before, they are both patterns I used in `nom`. (I've also only skimmed the original paper, they could be in there and I didn't see them.)
One of them is called `Compose`. As an alternative to a `Sequence`, instead of a group of parsers consuming the input in order, the first parser consumes the input, and the subsequent parsers consume the return of the previous parser. This is useful for instance when implementing escapable strings; the first parser grabs the entire string, the second one transforms escape sequences. (There is a mechanism for transforming the content of a `Span` while retaining it's metadata.)
The other is called `Fuse`. This is a small twist on `Sequence`, where after matching the parsers in order, the result is all concatenated together into a single token. This is useful for a "pattern matching" primitive, where you want to find a series of tokens in order, but you don't want to split them into different tokens, you want them all together.
It's been a wild ride, there's been a lot of thorny issues. I often think I should've just stuck with `nom 7` instead of shaving this yak. But I've learned a whole lot about writing especially abstract/DSL-yy Rust by combining tuples, traits, and declarative macros. There are also other programming language projects I'd like to pursue, and it will be nice to have a tailor fit tool for parsing text.
Special thanks to dtolnay's `paste,` the real MVP.
We’re also hiring for multiple roles if anyone’s interested in founding roles (ML Systems, DevRel): https://vlm-run.notion.site/vlm-run-hiring-25q1
- object capabilities
- implicit async/await
- immutable collections
It's small and (hopefully) fun, and quite usable already. If you try it out, please share your thoughts!
Daestro - https://daestro.com
...absolutely no one requested an RISC V port, but I did that too for laughs. Neat to see the whole thing run on a system the size of a postage stamp.
Not sure what to do with it next. Will probably just let it be what it is, and fix any bugs that people report. Maybe move on to a new little weekend project.
To fund it, I’m building agentic workflows and automations for insurance, finance, and real estate companies. It’s a way to keep things moving while I get the startup off the ground.
It is very much a work in progress with lots of commented out code which are just experiments.
in essence, the core of the project is a vector database that works for end users who want no fuss quick capture, semantic and fts search, the ability to create new relationships with marginalia.
im not trying to replace tools like obsidian or notion, i think ycb works better with them doing what they do best! i also plan to make the stack self hostable in the near future :)
The SDK is open-source and on GitHub:
https://github.com/dend/decksurf-sdk
It's a hobby project, but one that I love working on because it unlocks some _really_ great hardware to be open to do anything I want it to be rather than be constrained by out-of-the-box client software that asks me to sign in with an account to get an extension installed.
Plugins can be used to add new "result actions" and new sources of entries to filter and select. Eg. recent Jira tickets, email inbox, shell history, Notion pages, etc. The result actions are a way to easily perform common transformations on selected entries (eg. wrap in triple backticks, find and parse json, trim whitespace, ...) or kickoff some script with a selected entry as an argument.
Project started as a result of having to do a lot of work using Ubuntu and sorely missing Alfred and all the workflows I'd built with it. I wanted something for which I could build workflows once and have those workflows available on whatever system I'm on. Plus to be able to build some plugins that would be usable by coworkers regardless of what operating system they're using and with minimal runtime resource usage. There are some existing cross-platform solutions which could serve this purpose, like Cerebro, Ueli, Script Kit, some others.., but I wanted something lighter weight than is possible with an Electron app. Granted the current state of Epte is that it's built with Flutter + Go + Python so the final distributable and runtime memory usage are higher than is ideal.
Basic Windows support is almost there but there doesn't seem to be a great solution to switching to existing windows of an application instead of just re-launching it. The tool isn't intended to be as good or better than any given OS's built-in launcher so I'll probably just leave that as-is and upload the current state of the Windows build.
I have been working on this tool to create lead magnets. Magnetron researches the web for your topic and creates a well crafted ebook as your lead magnet.
You can try it here (https://magnetron.ai)
(This is still WIP)
I’m trying to really see and feel what’s actually missing in my life and trying to build it. Right now I just want to see what my friends are up to in a non-curated way.
Last month I added Dockerfiles and a docker-compose.yml file to the project to make building and locally running it a breeze. Earlier today, I finally had a chance to figure out and document how to debug the app, which should greatly improve quality of life for anyone trying to fix bugs or add features to the backend. https://github.com/OneBusAway/onebusaway-application-modules...
Also I have built a chrome plugin that can filter twitter by feeding your feed to gemini returning only tweets that match a criteria (E.g. no politics, only Ai or something more elaborate).
Edit: Three from prior published here: https://github.com/jaronilan/stories
We're now getting ready to launch a web portal for others to manage their own private hosting in a simpler way. The product also includes a directory of off-the-shelf applications which can be launched in a few clicks (eg. Deepseek chatbot).
If you're interested in being part of our closed-Beta in March, reach out! (e: james at below domain)
I recently added support for Cloud Run and am now building it out. Support for Cloud Function is also on the road map.
I’m also still maintaining the patch [2] I created which allows you run App Engine Python 3 Apps with dev_appserver.py on Windows. To test App Engine bundled API/services, you need dev_appserver.py
[2] https://github.com/NoCommandLine/dev_appserver-python3-windo...
Cool. I got interested in this subject recently. Have been checking out some text articles and videos about it. Unfortunately there is not much info available (and some of it is advanced stuff), or at least I couldn't find much, so far.
I am working on a library, which is not exactly a parser combinator one, but borrows some of those ideas, for use in other projects.
>One of them is called `Compose`.
About the escapable strings example: can you not just rescan the string for the escape sequences, after grabbing the full string?
Almost like a guitar hero thing but with just a metronome
It requires WebMidi to be enabled:
https://interjectedfuture.com/tag/lab-notes/
Subscribe to follow along if you're interested.
I would really love for you to reach out via the email in my bio so we can talk ontology!
The relevance to hackers is that unlike most fantasy where spells are cast with hand motions, magic words, or spell ingredients, there's actually an explanation for why that works and makes sense.
https://www.amazon.com/Dark-Sorcerers-Intern-Humorous-Fantas...
The first clients are here and I am working on "darker tech" now, a single codebase injecting their data in both MS Office and Adobe's software suites. That's quite a change from Elixir.
Also working on the reverse feature, reading/writing MS Office files inside Alzo. For that, I'm writing a Java app behaving as an Erlang node, connected to my main app via Erlang distribution, to leverage existing rich Java libraries for office tech.
This didn’t feel integrated enough and could fail if the phone was off, so I started looking into Swift and created my first app [1] with added features like contact import and notifications for other people‘s ages in days.
It‘s still very much a work in progress but the core functionality of the lock screen widget is something I use almost every day to quickly get the current number and use it for notes etc. I just like having an incrementing unique-to-me number to reference stuff.
[1]: https://apps.apple.com/us/app/days-of-life-milestones/id6738...
I'm also building a customer support app when I'm taking a break from Gorby. The idea is to make it easier to organize and quickly find/copy useful replies, discounts, screenshots etc. It's similar in concept to text expander apps, but those never worked great for me, I either forget the triggers or don't bother storing things I don't use daily. You could probably use Notion for this too, but to me it's too clunky and slow.
The hardware is based around a ESP32. The server that gathers and prepares the data is running on Symfony php. The app to configure the device is written in vue and is using capacitor by ionic. More technical details are here: https://sschueller.github.io/posts/turning-a-project-into-a-...
I pushed myself to do a couple of game jams cuz I thought it would make the burnout go away, but it’s basically only made it worse.
It’s the first time in my life where i haven’t had a billion ideas in my brain and im not sure what to do with myself. Been trying to listen to history podcasts and read manga to inspire myself again but it’s not working…
I wanted to have some data on how many people speed, the max speed recorded, that sort of thing. Things the city should be doing after many complaints of dangerous driving and people being almost killed on zebra crossings.
I have a doorbell camera, and by analysing the footage using OpenCV and some code, I can track how fast people drive if you see how fast they move between two known points.
Average speed: 46 km/h :(
[0] https://github.com/rumca-js/Internet-Places-Database
I use it for various projects, like simple search
[1] https://rumca-js.github.io/search?page=1&search=emulation
As well as finishing shipping the remaining boards to the kickstarter backers (many years late, but significantly better).
Been a long struggle overall...learned a lifetime's worth during the last couple years. Every single day has been spent doing something new it seems. Looking forward to what the next broken machine will teach me :)
[1] https://youtu.be/Vk53VsXkh9o?si=SU45-DkkjwZi6orp [2] https://github.com/Smoothieware/Smoothieboard2 https://github.com/Smoothieware/SmoothieV2
Donezo: https://donezo.pages.dev/
I am just going for the low hanging fruits - very basic stuff like location, person, event, activity, date (a few more things) and relations between them. One of the limitations is that I want to do deterministic extraction with suggestions from AI/ML, so there is much code to be written. I will email you, thanks a lot!
I would say the insights frontier models have given me, at a high level, match some of those offered by professionals in a theraputic context--which is one reason I'd be curious to make an affordable/accessible app. Although I tread lightly into depersonalizing such a human area with techno utopian naivete...
WOULD YOU BE INTERESTED?
The plan is for it to be federated, but that's a while off yet.
I recently spoke with a Lemmy developer who gave me some advice on making it easy for anyone to host. I was struggling with the mess of supporting both docker and VM hosting. He told me that Lemmy uses ansible provisioning to install docker compose on the target VM so that the effort can be focused on docker support, so that's what I've been homing in on for the last few weeks.
- The idea: https://carlnewton.github.io/posts/location-based-social-net...
- A build update and plan: https://carlnewton.github.io/posts/building-habitat/
- The repository: https://github.com/carlnewton/habitat
- The project board: https://github.com/users/carlnewton/projects/2
https://github.com/ironcalc/IronCalc
I have been doing this as a side project for over a year now. It's progress is slower than I would like but there we go!
Parser combinators are a bit hard to get into, the most helpful resource for me was `nom`'s "Choosing a Combinator" document [1], which is dense but gives you an overview of all the Lego bricks which you can then start imagining how to fit together.
I've not really read it, but there's also the original paper on the subject [2] (as linked to by the `parsec` documentation [3]) which describes the nuts and bolts theory behind it.
> [Can] you not just rescan the string for the escape sequences, after grabbing the full string?
Absolutely, this is just a convenience around that pattern that allows you to express that like:
let string = quoted_string.then(escaped(json_string_escapes)).parse(&input)?;
And also more generally for any parsers `foo`, `bar`, and `baz` as:
let quux = (foo, bar, baz).map().parse(&input)?;
[2] https://web.archive.org/web/20140528151730/http://legacy.cs....
It’s a curated directory of personal blogs and a blog search engine. I started to build a simple RSS-reader for myself, just wanted a HN-like list of links. Slowly it grew, and now it has full-text search across tens of thousands of blog posts from 700+ blogs (adding new ones every day), related blogs and posts recommendations, lists, link blogs. Now I’m working on adding email newsletters, curated collections, and text-to-speech generation.
Historically, UPSes had various proprietary communication protocols over serial ports. Nowadays they usually have a USB port, but I have a bunch of very old APC UPSes with just the serial port and/or the expansion card slot (which is usually just another serial link plus power on a edge connector).
A normal person would just use NUT or apcupsd over serial and call it a day. A bored person would write a USB HID power device stack and serial protocol acquisitors to give these UPSes USB ports. An insane person would add projectors from the USB HID power device stack to serial protocols so that they could use whatever communication card they want on any UPS they have (for example, a CyberPower RM205 card plugged into an APC Smart-UPS).
Why? Because apparently I'm insane and I needed a break from working on delinking executables back into object files, another heretical project I've worked on for the past couple years.
I've just started and I don't know if I'll finish that, but it's something I need to work on to exorcise that particular nagging thought out of my head.
Once I get this done, I get back to the actual project of a 2.11-BSD based handheld computing appliance.
Also there is this thing called "day job".
Our reader's fav reads of 2024: https://shepherd.com/bboy/2024?only-published
Feel free to message me if you've got any tricky or tedious problems to do with creating color palettes that extra tooling like this might help with! I have more feature ideas but I want to understand more what others need.
I'm planning to write some articles for giving a more intuitive sense about WCAG color contrast rules and picking accessible colors too. From working with designers, I find many give up here because it takes a while to get your head around and it's often not obvious how to fix designs with failing contrast.
Sounds very cool. Good luck with it!
It’s going to be fully decentralised from day 1—we borrowed the PDS model from Bluesky to allow users to run their own “knots” to self-host their git repositories.
https://www.tldraw.com/s/v2_c_E1XeFuW0tGbDxqnhDiLRT?d=v-286....
Been trying the "do the thing, and desire comes after" for many things (baking, piano, skating, ..), but that hasn't really worked. What has seemed to work is connecting with people (crucial that they know how to connect back).
Made a little web app that helped me communicate: https://azriel.im/tears/
(I could just point to the number when I couldn't talk/listen)
- Last month I demonstrated the ability to build Nintendo 64 ROMs with Zig¹, making some headway on Zig-native APIs for interfacing with the N64's memory-mapped hardware. Taking a break for a moment; will probably resume when Zig 0.14 drops (within a couple months IIRC). Next planned milestone will be to implement interrupt handlers.
- Gradually migrating my code repos from Git to Fossil (with plans to continue to mirror to Git). Experimenting with bidirectional syncs in order to preserve the ability to handle merge/pull requests from the various Git repo hosts on which I syndicate my repos. The above Zig64 project will probably be the first real guinea pig.
- Migrating my personal website away from Jekyll has been an ongoing project (going on almost a year now) with multiple parallel efforts: using Fossil's wiki features², using Scroll, and (most recently) using Typst's newly-announced HTML export feature. All three approaches have their pros and cons.
- I've been tinkering with my PowerBook G4; recently swapped in an SSD (using an mSATA→PATA enclosure) and installed the latest OpenBSD (with all partitions except for '/' encrypted; working on documenting that process and the associated kinks - and possibly turning said documentation into installer and initscript patches so that hardware platforms like macppc that lack support for encrypting '/' can still enjoy not-quite-full-disk encryption). Next on the list is rebuilding the battery.
- That PowerBook is also the only working machine I have that has an optical drive, so as soon as it was consistently booting right, I took the opportunity to back up the stack of burned CDs/DVDs I'd accumulated throughout my lifetime.
- I have a bunch of my dad's old photos and schoolwork and such that I've been meaning to digitize and organize.
----
I'm also building an application and training materials to help people with annual strategy. I've spent 20 years in marketing and ops putting up with people doing it badly so this is an attempt to help people running businesses actually come up with a strategy likely to result in something valuable.
i simply want mine to be able to fill in forms in preview with a passport image as context. also to be able to do recurring tasks as if i was the desktop user. e.g., i’m going to bed keep working on this spreadsheet.
it’s working and built but very slow and buggy atm. uses multimodal LLMS and OCR but lots more optimizations needed. need to make it a lot faster. can demo it and need help if anyone is interested.
Be warned it's in early stages, difficult to use and code is big ball of mud. But the basic functionality works, so maybe it will be already useful for someone. And I have been using it and working on it consistently, so hopefully it'll only get better.
I’m building a tool to make monitoring setup a no-brainer. I’m talking about basic website monitoring setup in 5-seconds — literally.
The problem is not a lack of tools. The existing tools are not even that complicated, but they still require too much thinking to set up.
In practice, no one cares. Cops will even ask you nicely for footage if something went down in your street.
Anyway, best of luck with your project.
https://github.com/Krisseck/ROCm-Docker-Scripts
Needs more documentation and more projects, but all contributions are welcome!
Two new projects of note this month, one specific to Finnish language learners, and one that is probably useful for language learners in general:
* https://github.com/hiAndrewQuinn/tsk - A Finnish pocket dictionary with a TUI interface. This is the first nontrivial thing I've built in Go, by which I meant I had to implement and tweak a randomly pruning trie by hand to get the performance characteristics I wanted (it wasn't actually that bad). I chose Go mostly because of the fantastic cross-compilation story.
* https://github.com/hiAndrewQuinn/audio2anki - This Python program wraps around `yt-dlp` and `whisper` to create Anki decks for listening practice. This should work for any (monolingual) video in any language. There are many such projects on GitHub, I'm aware, but it was surprisingly hard to find any that actually wrapped around Whisper instead of needing an SRT, VTT, etc file to come from somewhere else. In that sense mine is a "one command" solution - just provide the YouTube link and go. It does not provide a translation for those subtitles yet; in keeping with the all-in-one approach, I'm thinking I might wrap around LLaMa 3 to let the user specify that we should also --translate-to {en,es,eo, etc} if desired. For now my reading skills are advanced enough that I don't need that.
There is a key difference between recording vs publishing. There are more restrictions on publishing and an objective assessment needs to be made between the interests of the person in the footage and the general public or publisher.
I would argue that recording the road to collect speed data, not keeping the recording longer than needed and not for example recording license plates, would pass in the Netherlands. Since you're making an assessment between different interests and the is limited privacy impact. Of course assuming this is happening on a public road and not someone's property.
Publishing the recordings instead of just the average speed data would be a very different story, especially if the cars or drivers can be identified.
The website is a bit old, but lots of exciting changes are happening under the hood and I finally have the time to make big architectural and performance improvements.
I have been working with LLMs and VLMs to automate browser based workflows among other things for the last couple of years. Given how good the vision models have gotten lately, the perception problem is solved to level where it opens up a lot of possibilities. Manipulation is not generally solved yet but there is a lot of activity in the field and there are promising approaches to solve (OpenVLA, π0). Given these, I'm trying to build an affordable robot that can help around with household chores using language and vision models. Idea is to ship capable enough hardware that can do a few things really well with the currently available models and keep upgrading the AI stack as manipulation models get better over time.
Have also been working on making a customizable self-tracking app for my thesis.
And learning Korean, working on an app to learn words faster: https://game.tolearnkorean.com/
It's based on some learnings I've had in the past building where building on managed platforms like Heroku and Render, and watched our costs explode, with an annoying amount of vendor lockin.
It uses Kubernetes under the hood (which you can now get fully managed for $12 / month on linode), which lets you take advantage of a ton of things that Kubernetes does really well, like automatic healthchecks, zero downtime deployments, auto scaling, etc, while also making it easy to use for solo developers or small teams.
The additional benefit of Kubernetes is that it's also possible to host a bunch of other stuff in your cluster via Helm charts, that you’d normally have to pay for like: Sentry, Wordpress, Postgres, etc.
As I get older, the desires that "work" are very frequently the same ones as I had when I was a child.
just started it this week (won a bunch of things at a hackathon with it)
this is going to be free (there's a different product for enterprise this will feed into - but this is going to always be free). join the discord for announcements.
https://www.youtube.com/watch?v=_8K7yWjdyJQ
We didn't win :(
Also, thinking about building a catio for my cat.
Right now, I'm building a bank statement PDF converter to track my past spending. I’m about halfway there, with a semi-automated way to categorize transactions too. So far, it’s working great!
Done the first demo: https://www.osequi.com/studies/list/list.html, now focusing on "diagrams as code": https://tonsky.me/blog/diagrams/
https://store.steampowered.com/app/3539310/Nitronauts_Demo/
It’s built on a custom C++ engine (using SDL2) and uses WebRTC for networking, so a browser version is coming very soon. It’s a 2-6 player couch/online party game with Bomberman-like mechanics, plus wacky items and power-ups across nine stages.
You see, Stripe is very powerful, but also very complex. Coding a straightforward subscriptions implementation will take you a couple weeks at best.
That is without handling all those edge cases like: prevent starting a paid subscription without a billing card on file (yes, you read that right)!
The gem is ready, I'm currently working on getting the website up. If you're working with Rails and need a solution for subscriptions get in touch at hacker.news@railsbilling.com - I'd love to chat!
I had the idea 2 years ago, but starting building in earnest 2 months ago. Spending all my time on it now, minus 3 or 4 days per week of earning money. Currently looking for a GTM/sales-oriented cofounder in NL.
Recently added Similar Topics feature which uses Scikit-learn's TF-IDF vectors
Just create/upload your resume, update it, choose the template and colour and download.
1. I have a cloud platform for the movie industry (although in reality a lot of different industries use it for different things) that allows you to share files and get feedback from your team that I’ve been rewriting in Rust. Didn’t necessarily intend that but I started replacing Apache with Rust and liked it enough that I kept on replacing stuff.
https://www.kollaborate.tv (current version on cloud is not the Rust version, but on-prem is)
2. I work with another company that uses a really rudimentary way of time-tracking employees. So I’m working on a system to use their device MAC addresses to count their hours when they’re connected to work Wi-Fi. I was surprised that such a thing appears to not exist. I’m still working on it so it’s not anywhere public right now.
Just recently actually I published a Chrome extension for it too, so you can access and copy all the Tailwind CSS colors directly from an extension: https://chromewebstore.google.com/detail/tailcolors-tailwind...
Yes, I'm aware I relying on WhatsApp and that it is a risk.
Although n=1, I do have the feeling it works ( but that could also be the IKEA coginive dissonance)
You can’t seriously believe this right?
I’ve gotten to some degree of a protocol, surprisingly.
——-
I had made this LLM prompt (about a year ago?) wildly useful in helping me think.
It’s helped with everything from realizing you have burnout, relationship issues, arguments with family members, Mondays
https://chatgpt.com/g/g-Cdq3drl87-two-guides-3
I made another version, which showed its “thinking”.
I moved to London some years back, and was pleasantly surprised by the vibrant cinema scene, that seems to be in a steep decline in so many places. On any day of the week, one can find independent films, old and new classics, Q&A's with filmmakers etc. playing in one of the many theaters across the city. Staying on top of it all is a chore though, and I found myself missing out on screenings regularly, because I didn't check that one cinema's website on time.
This is also my first time building and releasing an independent app. The journey from research, backend development and learning SwiftUI has been a trip. Released on TestFlight a couple of weeks ago.
After spending a lot of time in an acquired startup and becoming more specialized in my role, I realized I needed to switch back to “build mode.” It’s been a rewarding exercise to try generating some organic traffic (no-ads by design) and a much needed escape from excel sheets.
Wondering if being still on a third level domain is messing up my SEO efforts.
I'll think about this for sure, especially since I've been thinking of making it possible to save and read local files. If you'd like to try Gorby, send me an email and I'll be happy to give you a free license code :)
Another project that is currently only happening in my head - I am thinking about security operations teams that I think often do the same things in different companies. Namely there is a lot of tinkering with detections and alerting, often for the same services. I think this could be cost optimized by being offered as a SaaS.
A lot of professional sports clubs, S&C coaches, etc.. use timing gates for measuring sprints, but those are a pain to set up, only capture split times, and are expensive. I think radar (+ optional video overlay) provides a far superior solution.
Quillbot makes it sound better for essays and presentation. I want to do the opposite and make it sound elliptical, turn a rant into a jab, that kind of thing. You don't say, "You're an idiot," you say, "Thanks, I thought I was the dumbest person in the room."
https://free-visit.net/fv_users/garance/vis/VisiteBNF001-004...
GF products are expensive and hit or miss, I really wanted something where I could keep track of my favorite items. I also want to let people rank them, so maybe I can discover the best gluten free hamburger bun (Rudi's Brioche), or beer (Glutenberg Blonde).
I'm also making a user submitted recipe section, so say you want to recreate a Taco Bell Doritos Locos Taco, it's easy to link to the products you need.
I'm not sure where this project is headed but I couldn't find any jobs working in this space so decided to make something to help myself.
https://www.producthunt.com/posts/reflect-ad2b97ed-13af-443d...
* Track most active users in a period (day/week/month)
* Track which users post/react the most
* Know (and now it is possible, but not implemented) who is 'lurking' just reading messages
* Summarize content and frequently asked questions to help users in the future
* Check which channels (groups inside a WhatsApp community) are most active and how many messages were posted in each channel in a specific period (day/week/month)
* Track (external) links posted to communities, how common they were, which users posted these links
--
The tools:
- A WhatsApp bot capturing all events in the group (reactions, messages, etc)
- A (same) WhatsApp bot to onboard new users in the community, check your membership, talk to
- A web application where the community member can check their membership + admin panel for me to check the stats mentioned above, think of it like a circle.so but built for my specific case.
--
The tech stack:
1/ a WhatsApp instance running on web, I'm doing this because WhatsApp oficial APIs do not support the "communities" -> I use Z-api [1] but you can use something like Baileys and self host [2]
2/ Elixir running on a simple machine
3/ OpenAI/Claude for summarization and topic extraction
[1] Z-API https://www.z-api.io/ [2] Baileys https://github.com/WhiskeySockets/Baileys
(is the domain for sale?)
MMPs collect user data across apps to help apps run ads. This is needed because mobile apps are downloaded without the additional data you could get passed along an HTTP url like you'd see in a regular email marketing campaign or YouTube affiliate link.
My goal is to create a way for mobile apps to self host their advertising attribution, keeping their user data in house and not sharing it to a 3rd party like AppsFlyer/Adjust/Branch. There are only a few companies that do this in the world and NO open source non 3rd party option.
Please reach out if interested, I'd love to chat!
Learn Dutch through immersion, look up any word and get lots of examples in very simple Dutch.
Just having a lot of fun playing around with LLMs and language learning basically.
Basic functionality has been implemented, and I am working now on polishing the UI and workflow. Big features like art strokes, path offsetting, colorizing, etc. are also in the making and will be added later. I hope there is still a commercial market for products like this.
There is a theoretically stable algorithm for the classical problem called the Remez exchange algorithm, and an extension to complex domains due to P.T.P. Tang in his 1987 PhD thesis at Berkeley. Theoretically Remez and its complex extension are very stable, but unfortunately implementations my advisor and I are aware of seem to struggle with large degree polynomials, where large is bigger than say n=45 -- errors begin to explode.
In any case, independently of this I've been learning more of the nitty gritty details of deep learning for a project at work (I'm a SWE in my day job, the math is more moonlighting), so to ground my efforts there I've been exploring deep learning approaches to this problem of complex uniform approximation, implementing results from various papers and tweaking things for my use case, and coming up with questions. That's much of what I'm thinking about this week!
Also, I'll be having a half-day long ADHD evaluation session on Friday -- so a bit apprehensive about that.
Back in 2005 I remember working with some physical sec company that were setting up cameras in a factory, and they wanted the cameras to 'not record traffic, be activated on if THIS part of the screen has motion')(sidewalk vs sidewalk-right-on-our-doorbell vs road). Also, sudden changes in lighting would trigger it :)
Then you need to have retention period (good luck). Most people use those door-cams are violating GDPR. UNLESS when people complain and take you to court (very very very rare), you can prove that "I auto-delete records after 24h when there is no incident", "I have proximity scanner so it is only 0m-2m from my front door", etc.) (violating GDPR because "hey you pervert why do you record my kids EVERY DAY going in and out")
Privacy and Data protection is very very very difficult with GDPR (and thank you Facebook for messing up back in 2015ish!!!)
You can set up your cam but have the "AI" automatically pixelating all license-plates, and the video recording (if any) should be post-pixelating, and not the original feed. How about you put something with a speed-measuring-sensor (that is NOT a camera), so you only get 'anonymised' data, i.e. "20 moving items", and their speeds. But you will not be able to tell if the 300km/h was done by a bicycle or a Hayabusa ;)
LLM driven 3d packing written in F#
- There's no speciality grade coffee in Zambia, and all the coffee beans in Zambia are from Zambian coffee farms. I've bought a small roaster and will start sourcing speciality grade coffee beans from Malawi, Kenya and Ethiopia and roasting for a few of the small stores and cafes around us.
- Converting a beat up Suzuki Samurai into a capable 4x4 rock crawler/off road vehicle to enter my own team into the Elephant Charge 2025 https://elephantcharge.org/ec-charge-2025/.
- Growing and propogating cuttings of coffee plants in my backyard to start an outgrowers scheme in on the border of Zambia & Congo.
- An emotion recognition app that has animated fruits and animals that dance and respond to emotion. I'd like to create something a bit more responsive for my child than the YT videos that exist (Unity).
- Helping my wifes company prototype and spec out some lease management software.
- Sourcing the equipment and ingredients to process my own coffee cherries into green coffee. I'll likely buy coffee from nearby growers and start processing as my own plants still have a couple years before they produce any fruit.
- Migrating a Flow project to Typescript.
- Learning Haskell by building a back-office API for another project in it.
It’s a bit hard to spread efforts over all of them, but at least most of these projects have lasted several years now, so not constantly doing new things that never finish.
generally using customized whisper based models with better performance
The road next to my house has a speed limit of 20mph but most cars go 45mph because it's a straight road 4 lanes but space for 6. No bumps, no curves, wide. Effectively it feels like you should be driving fast. If I go the speed limit in the center lane because I'm going to turn left people will get angry and speed around at 60mph pissed off
I didn't have any Python experience but it was surprisingly easy to pick up (MVP in an hour). Wrote it in notepad, which, imo, was a distraction-free experience. Prolly would be scrolling autocomplete than reading docs if I was in nvim. Took me back when I was used to completing coding exercises on paper.
If there is an implementation to read presences without using Discord client, let me know. Would be helpful to skip Discord altogether.
That said, my main goal for now is just to make it work for personal B2C use first. I do think there’s some potential here because major cities are pretty much cashless now, and there aren’t any good existing solutions for B2C.
There are some other decent options, but they mainly focus on B2B (that’s where the money is), so they’re quite expensive and overkill for what I need.
After seeing an example image on wikipedia someone took of a tram [2] I want to try do it to the trains that run near my house.
[1]: https://writing.leafs.quest/programming-fun/line-scanner
[2]: https://en.m.wikipedia.org/wiki/Line-scan_camera#/media/File...
``` @Synth def tone(freq=440, amp=0.2): return SinOsc.ar(freq, 0) * amp ```
It's not available publicly yet, but works well for my purposes and I'm working on productionizing it. Sign-up page for updates: https://letmeknow.jkoff.ca/infinite-ci
ATM I'm making some videos to show how it works and how it saves time for us. It's free, 100% private, local-first, and has E2E browser sync for subscribers.
And you can definitely add some ref links for a bit of revenue.
For example, you can search "cheese", and it'll show you results for cheese on Google. If you search "!b cheese", it'll search "cheese" but on Bing instead. "@yt !b cheese" will search "site:youtube.com cheese" on Bing.
I built it mainly so I wouldn't blow through my 100 Kagi trial searches quite so quickly.
There are plenty of apps that do outfit tracking, with some basic stats. But they all have a few or more of these shortcomings (from my perspective); unpleasant UI, no cross device syncing, lack of detailed usage statistics (e.g. cost spread over time by garment category), some categories just not supported, pushing a specific lifestyle such as Capsule Closets, or just plain focused on recommending what to wear using some mediocre algorithm that doesn't understand cuts and how different pieces fit together; basically only suitable for capsule collections.
These apps all have a lot of downsides too in common, which I haven't been able to solve either yet; ultimately you must start with an inventory of your clothes, and then work from there. It takes ages to catalog and import your clothes, and I haven't found many existing product that lets you export if you've even done it before. And on top of that, you have to be quite rigorous at tracking what you wear; the more data you have the more insight you can get from your choices.
I finally published on iOS a couple of months ago. No traction, and I don't expect there to be. I won't argue that my offering is better than any of the competition, but I've tried most of them (and wasted colossal amounts of time onboarding onto them) and found none fit my need properly. It's still very much work in progress, but I find myself reaching for it multiple times per week to inform my purchasing habits.
As part of this effort, I have developed a preliminary prototype (named deLLMiter) that I am refining.
This method is founded upon the hypothesis that there exists a correlation between first-order expressions and more complex, hierarchical forms of expression (https://glthr.com/llm-delimiters-and-higher-order-expression...).
With this project, my goal is to improve the security of LLM models.
- Reading advice like https://eamag.me/2025/Good-Research
- Figuring out problems like https://openreview-copilot.eamag.me/
- Doing analysis like https://eamag.me/2024/Automated-Paper-Classification
- Figuring out where scientists are located like https://eamag.me/2024/ICML-2024-on-a-map
I think people used to use science to set speed limits.
Unlike queues (SQS) needing state hacks or pricey orchestration tools, Duty uses your existing Postgres to ensure tasks survive (retry) failures, retain state between runs, and eventually finish.
Pre-release, star if interested
Once I feel comfortable, I'll probably open an Instagram account and hand out free personalised rings. For some time I've thought about how I should price them, and I've reached the conclusion that giving them out for free feels right.
How much we're willing to pay is a whole other question. I feel like this is the kind of thing that Cursor already does by itself but it's just not releasing a user-readable output of it.
It won't likely be a subscription thing, but one off payments per repo makes sense, and there should be some kind of satisfaction guarantee or say, charge to have the output in a human readable format.
Monorepos are also a pain. On the front end, they sometimes share design. On BE they may share databases. It would be cool to break it down into DDD-style domains if applicable or propose things like anti corruption layers. More often it's like a "pacific ocean meets atlantic ocean" kind of thing, where you can tell there's a difference in the way things are done, but it's not entirely clear where the border is. This would probably be worth a lot more.
To a much lesser extent, an architectural copilot would also make sense. On the front end, we have a lot of redundant components. Say a button might be PrimaryButton, but the same thing is GreenButton or FilledNoOutlineButton by other devs. We tried documenting this which just ended up being a waste of 1 week because nobody read the doc. It's worse with complex components like TwoButtonModal vs TwoButtonModalWithClose. And what happens is code is always built in parallel; people don't realize that the designer's new style applies to both teams so you get two people building the same components at the same time. Not a major problem, but I think this is worth a few cents every PR.
Ultimately it's hard to gauge. Like Copilot underdelivered, Cursor overdelivers, and yet both essentially do the same thing. I guess the amount we're willing to pay is just vibe-based.
Lot of cool cpp to write
It's been months I've been stuck on the description of note groups because of the insanely complex 2D semantics.
- Use natural language for telling offloading requirements.
- Just year of the LLM launch of HF url can help if it’s an outdated LLM or a cutting edge LLM.
- VLMs/Embedding models are missing?
This year I’m making it production safe and open sourcing it.
How do you deal with different kinds of groupings and connections? For example, some things could be connected because they are “integrations”, or because they deal with notifications, or because they’re available only in the enterprise plan. Not all related things are related in the same way.
Time investment has been _massive_ so far, but I just hit the first $100+ profit month, and despite the distance from my normal dev salary, the positive reviews/feedback have been an incredible reward that drives the motivation to continue. I will also say that it is quite the humbling experience to ship physical products and the experience has given me a whole new appreciation for the things we have in this world.
Early days, but check it out :)
[0] - https://freedomfrenchie.com [1] - https://www.etsy.com/shop/FreedomFrenchie
I am already several years into my career and I have a spouse to support, so I'm ambivalent about formally attending graduate school -- at least anytime soon -- since that would introduce lots of time pressure and administrivia for little apparent benefit. My relationship with my supervisor is mostly informal
I was just about working on the Twitter api when Musk bought the company and restricted access. Real bummer. I got employment weeks later.
It's in open testing phase for Android users if anyone wants to check it out: https://play.google.com/store/apps/details?id=com.manugo.bit...
This app doesn't directly answer those questions, but it gives the data needed to stop and think about the answer (quickly). I don't consider myself a vain person, but I consider myself a person who makes poor decisions with their disposable income (fashion).
The main selling point would be this: You could avoid buying 2 shirts that will be unused then thrown away every year, for the cost of 1 shirt. Save money, and textile waste.
But also, why must everything be profitable? I most I could ever hope for is that the hosting costs are paid for.
It would be more real-time and give me heartaches everytime I go out of budget :D:D:D
In my off-hours, I'm working on an old school pixel art RPG, but in 3D.
Oh, and finally I'm also working on finding a new job :-(
https://docs.google.com/spreadsheets/d/11jLH3UISmRId_gTFXiTQ...
The Quran itself is referred to as the reminder. Because it's supposed to remind us of why we're here, our purpose, who made us, and all that.
Currently I'm restructuring a big chunk of the code for readability, trying to apply good refactoring practices (work in small bits).
After a lot of frustrating experiences with Python linear optimization DSLs, I'm thinking of writing my own at some point as a side-project :)
I don't have a blog post to share. My experience here has been largely positive, modulo the obvious financial caveat: My take home salary really is about ~30% of what it would be had I stayed back in the United States and followed a similar career trajectory. In the long run I think this is an eminently fixable problem, however :)
I'm also doing some electronics - I'd like to make a tool that gives blind people without light perception light perception by putting a lightweight device on their forehead that delivers haptic feedback based on light intensity. I'm doing that with a friend, and we're planning on open sourcing the specs.
I'm honestly really surprised about how much I get stuck on business logic decisions. I went into this thinking making appointments, basic managing of employees and all that would be simple and relatively similar across salons.
Additionally I'm considering where I should move to. I wish to live in a place where owning substantial land for homesteading in a relatively climate safe area (relatively doing a lot of work there but imagine not already arid or with high storm risk) is not completely out of grasp. My region of Belgium is too densely populated for this. Even if I'm not moving to a different country even next year I figured it's the kind of thing that takes a stupid amount of preplanning.
Started building this with a friend, as I was personally frustrated at the lack of good options when it comes to "true IMAP" email clients.
So for my first contribution to FreeCAD I'm working on fixing this.
The underlying CAD operations are done by "OpenCascade", and at first I thought OpenCascade had no support for aborting operations part of the way through. So my first implementation was to move the operation into a child process and give the user a dialog box that would allow terminating the child process.
But it actually turns out OpenCascade does support aborting the operations! So now I'm working on doing it the OpenCascade way.
My PR is here: https://github.com/FreeCAD/FreeCAD/pull/19796
Existing apps such as MyFitnessPal and HealthifyMe fall into two ends of the spectrum where you either need to add ingredients one by one, or your food is logged with a standard macro count where you cannot change the ingredients used.
Weit ideally provides a seamless experience in taking a picture to retrieving ingredients to retrieving macros per ingredient. Once that's sorted, food tracking should be granular enough to build intelligence around it to improve one's diet based on their requirement.
Honestly, I used to constantly struggle with the realisation that none of my ideas are unique and whenever I see someone having built something similar, I feel like I'm wasting my time. I'm getting better at dealing with it now though.
1. "Web-native" as in the protocol is designed with HTTP and modern web browsers in mind. Consequently, it can be implemented using Service Workers so that no additional software (nor even browser extensions) are needed to access files.
2. Files are addressed by their cryptographic hash of their content (a) to ensure the authenticity of the data received from mirrors and (b) to avoid hard-coding specific locations/servers (i.e. content addressing).
3. Files can be mirrored by anyone and users can retrieve files from any mirror; no party requires any permission from any authority. This is in contrast to traditional mirroring schemes, where mirrors have to "register" with the owner of the content (e.g. to mirror a Linux distro).
Demo: https://webmirror-demo.netlify.app/
Code: https://gitlab.com/webmirror/webmirror/
Work in progress!
Website is just a placeholder, but I'm documenting my progress on Bluesky at https://nickfisherau.bsky.social/.
I'm glad to say it's gaining traction - here's an unsolicited post by Adam Wathan the creator of Tailwind: https://x.com/adamwathan/status/1889134672866582617
And here's the site: https://arcadium3d.com/
Note that I'm not always a CTO in the strictest sense of the word, I like doing complex technical challenges with software companies and sometimes just lead a complex project like implementing ISO 27001 or re-packaging a software suite for on-prem deployment.
Currently, it's set up as a daily (short) newsletter with a different link each day, but I'm trying to learn marketing to figure out how to get others interested. I've enjoyed creating it, but would like to see if others like it before moving on to a new project. Link to project: https://www.thedailydetour.co.uk/
You feed in your docs and you get a dashboard that shows your categorized “flow” of money (think sankey, stacked bar), as well as some simple grouping tools (Show me all grocery spends on my credit card by month.
I initially wrote it in Haskell, but I don’t really know Haskell and I didn’t feel very productive with the stack, so I’m reworking it in something more familiar now.
this week, thinking of adding a new feature where users can create forms just with prompts..
Made in Love2D, mostly because it's limited in its simplicity (good for creativity) while still allowing me to make something usable. That, plus I love Lua, which is how the project event got started - just me wanting to mess around with the language. From then on it quickly spiraled out of control - 2 weeks to make the core logic of the game, 2 months to create a basic UI library from scratch, just because.
Probably the best way to develop your taste and understand the spectrum of different coffees available is to do comparative tasting aka try a small number of coffees in parallel to compare and contrast. I was having trouble finding tasting sets so I started freezing a little portion of every coffee bag I bought to create a collection for doing these tastings at home.
I needed something to keep track of them all (as well as my tasting notes in general) so instead of using a spreadsheet I built a full app for it. The app supports NFC-tagged containers which I've found to make my workflow a whole lot easier.
I also set up an online store to sell the NFC-equipped single-dosing tubes: https://store.coffeelibrary.app Planning on adding more containers that work well with the app in the future.
How are you building the app? The interface makes me think it's native SwiftUI?
I've been having a lot of fun with AI agents lately. Have tried a lot of them - Cline, Roo Code, Windsurf, and finally settled on Cursor now with Claude 3.5 sonnet. It's been a big boost for my productivity.
AI helped me write a synchronous API proxy in Go that I'm almost ready launch. One of the main challenges with Ruby on Rails is that it's terrible at handling long-lived HTTP requests. Especially a lot of them at the same time. So our PDF generation API was forced to be asynchronous and our customers need to poll for status updates (or set up webhooks.)
This new synchronous subdomain will handle all the polling logic for you, so you can just make an HTTP request, wait a few seconds (or longer), and receive a link to a PDF that's ready to download. Even with AI, it was still very difficult and took many weeks to get it right. Challenges included security, load testing, data races, concurrency, and setting up reliable, secure infrastructure with an internal load balancer. I learned a huge amount about both Go and Kubernetes. But it's almost done and I should be launching in the next day or two.
After that, I'm finally launching support for template versioning. This will allow you to pin your API requests to a published version, so you can keep making changes to a draft version without affecting production. It's long overdue so I'm excited to get this launched as well.
Also working on a side project from time to time: VisualCI [2]. We have a lot of PDF integration tests that use image diffs, and some browser tests where I compare screenshots. So this is a tool I've wanted for a long time, and the paid services I've found can be a bit pricy. I'm going to try to build a very simple MVP that just does what we need, and maybe others will find it useful too.
I'm also writing again. A story that's becoming more cyberpunk than I originally intended. It'll probably never be read by anyone but me, but getting it out of my head feels nice.
Also started going to the gym and working on my health.
It generates reports to show you your numbers in a bunch of customizable ways, it generates these reports in less than a second and uses a single CSV file as your data source.
I've gotten things to the point where I can do my books every quarter in about 5 minutes with complete accuracy since it supports importing arbitrary CSV files such as bank exports with a way to automate categorizing things in any way you see fit. I currently use it to track my income, business expenses and personal expenses.
Basically I ran into issues using different finance tracking tools over the last decade which always made me feel unhappy to use those tools so I built Plutus with intent to resolve all of those issues I had and make me happy while using it.
I've been into 'small web' and unbloated websites for ease of use and privacy reasons, and all of the available gopher browser apps on iOS are not great, so I have decided to make my own. Maybe someone will find it useful.
The idea is to have srs flashcards, like Anki, but without the pain of creating example sentences, translations, images and audio.
As a bonus, I've added the option to add vocabulary from ebooks, YouTube videos and websites.
Edit: if you meant which dialect, it will be whatever grapheme->phoneme ruleset I get a hold of.
I want to build a local company in my city of Utrecht, primarily on-site. That gives me the most energy and fun and is something that I want to optimize for.
It is more an excuse to create a simlpe Electron app, though.
Recently open sourced everything.
More broadly, I’m working on replacing my day job with something more exciting and impactful.
I think the most excitement and impact can be found in a startup, so that’s what im trying to get into now!
I’m a big privacy advocate but when you are handling a killing machine on a public road, there is no privacy IMHO.
There's a wait-list sign-up though: https://aiconstrux.com
I'm in the process of refining the filters determining which movies will be included. And at the same time I am trying to acquire news users, which is going quite well. Slow and steady increase, I am currently sitting at around 100 visitors per day.
I also already know from an insider that a decently sized Belgian chain has people really frustrated with Salonised so I hope to get my foot in the door there. However before I try that it needs to be a bit more than an MVP.
[1]: https://github.com/psviderski/uncloud
[2]: https://uncloud.run
The displays are really expensive so I’m looking at taking 12 kindles apart and mounting them in a 3x4 grid. They cannot seamlessly touch at the edges so I’m looking to include that as part of the larger aesthetic then ignore it.
I’ve figured out a few possible approaches and the software/service side - next step is to order 10 more kindles and get to work.
I'm also developing an online store for media files. At this point it would have been cheaper to pay for a ready-to-use service, but I felt like refreshing my knowledge of web development. I'm still unsure if going with react-router was the right choice.
And yes, this type of driving will produce annoyed drivers that "drive crazy", and I don't accept that this is just their fault.
Mostly, these same drivers doing 20km/h will not even stop for pedestrians on a crosswalk — slowness does not equal attention and safe driving!
Traffic, in essence, is a collaborative effort that requires all participants to be empathetical to other participants — as such, we need to be most mindful of the "weakest" participants like pedestrians (especially kids, who can also be very inattentive), cyclists, motorcycles but also of other car drivers — if we care about each others' experience, we'll reduce the risk for everyone involved, while getting everyone where they want to go in a timely and efficient manner — and that is the goal!
Registration will be free (compare Form N-400 to become a U.S. citizen which costs $640 plus an $85 biometric services fee, totaling $725), you just get free benefits.
There isn't any signup form yet but you can email the Founder Robert Viragh at rviragh@gmail.com with the message "request for citizenship in Utopia" and I can give you citizenship, by our laws anyone gets citizenship upon their request. (I will reply with confirmation within 24 hours, you can reply here if you emailed me and I didn't reply to you.)
I can hear you thinking there's no way a sovereign nation will be run and owned by AI and give out free money, goods, and services. Well here's our complete game of chess: https://taonexus.com/chess.html made by AI purely for your amusement, it's a 1500 year old game people obviously get utility from (spending $10 to $1000 on chess boards for example, with tens of millions of boards sold per year). So clearly this type of game is of some use/utility to people. I have fun playing it for example. AI just made it for you for free.
The goal is to bring inspiring emotions like these to the classrooms - https://youtube.com/shorts/waLXYiV-2cE
- [GitMentor](http://gm.srecraft.io) → 300+ users so far - [NurtiLens](http://nl.srecraft.io) → 50+ users so far
Blog post on GitMentor: https://blog.srecraft.io/posts/gitmentor
- Terminal in Notion: https://www.notion.so/Terminal-in-Notion-195668ab1a058044b0e...
Dropping Resume Optimizer: https://resume-maker.up.railway.app/ Made a video building it from scratch in 30 mins—completely freestyle. Accidentally exposed some API keys, so had to rotate them since I don’t know how to edit videos :) Video: https://youtu.be/OCcAjZ4Q-iM
Conclusion: You steer the LLM, don’t let the LLM steer you.
- [LLM Bootcamp](http://llm-bootcamp.srecraft.io) → making content
Another day, another drop. Not taking this one to production because PineconeDB costs too much for cosine similarity search. Built it anyway, here's the video: https://youtu.be/f5SIELet8JU Codebase: https://github.com/avirajkhare00/youtube-answer-finder
YT Answer Finder goes to prod today: https://ytf.srecraft.io/ Only two videos processed so far: - https://www.youtube.com/watch?v=P_fHJIYENdI - https://www.youtube.com/watch?v=R7WPEYGr1Vs
More projects: - Real-time flight tracker: https://lofi-atc.up.railway.app/ - Music recommendation app: https://music-again.vercel.app
It's RPA for browsers which is not fundamentally new, what I'm trying to do that is new is use AI to make it as easy as possible to create automations. Most of the existing tools require you to locate CSS selectors, XPaths, etc. whereas this is just point, click, type, describe data you want to extract in English, etc.
Still early days and it works much better for some tasks/websites than others but it's improving rapidly and I'm quite excited about it.
Also hoping that the likes of OpenAI Operator, etc. are rolled out in a way that I can use them to build a better product rather than being runover by them.
I played around with turning an old kindle into a wall mounted display a while back and got frustrated at how buggy the process was.
I managed to get my own image displayed on the device but sometimes it wouldn't update, sometimes it would drop power way too fast after a full charge.
I think I would use an eink display and a raspberry pi if I ever revive this project.
It's built with Elixir and Phoenix LiveView, backed by SQLite. Records are imported via MusicBrainz, and data enriched via Last.fm.
I'm looking now to add notes for each artist and record, along with arbitrary associations. Think supergroups, side projects, etc. and some trivia/quotes/stories that I can easily add for my own reference.
LeadSparkLabs is an agency that assists small and medium-sized businesses in generating more leads by quickly creating customized Lead Magnet Mini Apps. These mini apps are designed to engage your target audience and convince them to give you their email.
https://play.google.com/store/apps/details?id=app.puzzle.pad...
On your site it says "The calendar data can come from one or more calendar providers". Is the connection direct (e.g. data flows from Google to epaper calendar), or via the app installed on the user's phone, or does it need your servers in-between?
I've been obsessed with making it easier to handle tab overload in the browser without requiring any sort of active "tab management".
I have a working extension that replaces the "new tab" page with a clean view of all open tabs, along with simple ways to search and select which tab to switch to, including search over bookmarks and history. There are also some simple tools to allow for creating and reorganizing tab groups.
For a small group of people, it revolutionizes the browser experience. I'm still trying to decide if there is a widely-useful product there, or if it's just a niche use case.
Any and all feedback welcome!
WebDSL, fast C-based pipeline-driven DSL for building web apps with SQL, Lua and jq: https://github.com/williamcotton/webdsl
Search Input Query, a search input query parser and React component: https://github.com/williamcotton/search-input-query
Guish, a bi-directional CLI/GUI for constructing and executing Unix pipelines: https://github.com/williamcotton/guish
Using Gemini LLM and Python.
Highly focused on my needs, with special prompts and few shot matching.
Code is a little bit convoluted, maybe I will have time to open source it.
Helping people find their ideal place to live (in the US)
I have been working on it for about a year now. It is not yet public because I am yet to apply for Spotify API quota extension, but I'd be happy to allow access manually if anyone wanted to take a look.
To enable WASM properly you should be on the latest Chrome version since other browsers still have some issues supporting it, otherwise, it will fallback to canvaskit renderer which is slower.
I've written about the implementation itself here if anyone is interested: https://blog.dmilicic.com/posts/writing-a-personal-website-i...
Even tho I'm a rather smooth talker and potential salesperson it took me a fair bit of courage building initially to actually approach people about my product. I'm sure that'll get better with experience.
Down the line even if I hope to catch most customers trough unpersonal means such as online advertising, mailing, etc I'll still have to simply try and approach many larger business/chains directly.
https://github.com/WillAdams/gcodepreview
Writing it in a TeX editor using a Literate Programming system developed in the course of working on it, the PDF should give both a good overview, and provide all the code.
It has greatly expanded/restored my math/geometry and has me looking forward to how to implement Bézier Curves and surfaces using similar math (since G-code and CGAL are fundamentally limited to lines and arcs and what can be easily made from such constructs).
https://okuread.com/ is a desktop App that works completely offline and helps you read foreign language texts and learn vocabulary that way. No AI-garbage included.
Right now I'm working on an open source platform for enabling human pronunciations in Oku. Anki/Flashcard integration and a UI redesign are also all scheduled sometime in Q2.
* An open-source desk scheduling solution: https://workplacify.com/ (GitHub: https://github.com/igeligel/workplacify)
* A bit passively: https://sheetsinterview.com/ - Hackerrank for Excel/Google sheets (I need to add some more task templates, hopefully with AI)
* A Chakra UI v3 component library for SaaS
So far I’ve put all the side projects I manage (5 of them) and it’s working great. I can follow and query the logs, the JSON payloads by path, see live metrics like number of users currently online, etc. Even if I don’t get any customers I’ll continue developing it for my own needs! Will soon be adding alerts through webhooks, slack, discord, etc.
If you want to try it out: https://app.getboringmetrics.com (no landing page yet).
Edit: just want to say that if you want to try the software but not keep it, you can create an account in under 10 seconds, send a curl request to see logs arrive in realtime, and delete your account in 5 seconds. I do not track anything and do not keep a single piece of data.
- poesia.pics - to generate poetry (spanish) from pictures
- podtafolio.co - a podcast directory (colombia). it generates a summary of the episodes and provide a search experience
https://coolify.io, An open-source & self-hostable Heroku / Netlify / Vercel alternative.
(conversely, I drive a motorcycle sometimes which puts my head at or just over roof height of most cars, it makes 80 km/h not feel as fast. Mind you, the added road overview also helps)
Anyway. Narrow / winding roads and speed bumps will definitely make you want to drive slower. We have 'cars are guests' roads inside cities too, which are roads designed and coloured like bike paths (= red asphalt).
But the opposite is also true; I got a speeding ticket once, the road was a 4 lane, separated directions asphalt ring road... but the speed limit was 50 km/h.
"I'd rather put up with the speeding" implies you don't have kids and you only consider the noise nuisance, not the safety risk. Very self-centered.
For this I am jailbreaking the Kindles and relying on the screensaver functionality to pull down an image every 3 hours (will adjust based on power usage - only need 1 update per day tbh). This should require less power than the experimental browser using a web page with Javascript to refresh.
Collectively this will be a bit more abstract (no seamless NYT front-page experience) but I think I can get creative.
I need to see if 12 Kindles with USB cables across 1 or 2 hubs, each hub powered by a removable battery-pack is safe to hang on a wall :)
2. A AI assisted brief generator -> clients often have a hard time articulating their requirements for new projects.
3. Prototyping the UX of "my" version of the perfect "process aware" editor. More organized than a Wiki, more flexible than tools like Jira, Aha and all. Not ready to share a public link yet. My goal is share my mockup in a week or two.
Who the heck is so brain broken as to talk about their grandchild as a startup project and their family as a frigging team.
I have to almost need to see this as AI in order to maintain sanity because there’s no way an actual human being talked about a child—their human grandchild like he was a product.
I have a month sabbatical coming up and that was when I was planning to do all the physical work. Right now just validating the idea on paper and working with 2 kindles to ensure the approach is somewhat viable before I try and get 7 or 10 more Kindles of the same model
If I actually move forward in a meaningful way I'll do a write up and follow-up here.
There are apps that help kids on the autistic spectrum to communicate, and flashcard systems, and we're experimenting with these, but they're more geared towards encouraging the child to communicate. In our case, he communicates fine with gestures, nudges, pointing at things he wants, bringing flash cards of foods he wants, eye contact etc. And he seems to have good cognitive skills in terms of puzzles, basic arithmetics and counting, memory, etc. It is learning language as an auditory system that seems to be really difficult.
Adam can 'read' in the sense of knowing and recognizing all the letters (he takes delight in that) and pronounce the few syllables he's able to when he sees them written out (mostly consonants m,n,h with vowels a,o,e). His phonematic understanding for other syllables exists but is poor (e.g. he has trouble choosing between a BAH card and a PAH card when I say one of them out loud, whereas the letters B/P in isolation are easy). My idea is to build an app/site which teaches him and reinforces three-way connections [picture]<-->[written form]<-->[sounds] by letting him "type", initially by pecking at large squares with letters on screen, rather than an entire keyboard. So for example, there's a picture of him at the top, a row of 4 big blank squares underneath the leftmost of which is blinking, and 7-8 letters strewn around at the bottom, from which he can type in sequence A-D-A-M and get a sound effect of victory. For words he doesn't know or remember, there's a mode where he just needs to repeat e.g. C-A-T which is already written in identical squares in a separate row just above, then after a few successes the hint row goes away. For an MVP in which I can quickly backfill 100-200 simple words like that, and track progress, this would already, I think, be valuable; then maybe I can add a mode where the words sounds (with or without the picture) and he needs to type it.
If all this works for simple words, and he takes pleasure in typing, the stretch goal is to turn from words into short sentences, and both teach him phrases like I WANT [X], or WHERE IS MOM?, and let him request stuff with such phrases. None of this directly addresses the apraxia problem of actually learning to move his lips/tongue/throat/etc. appropriately, but I hope it can create more scaffolding around our efforts in that area (which we try very hard to work on daily) and together help him build an understanding of language/syntax. I'm very worried that, despite ongoing (very slow) progress in both speaking and understanding, phrases, sentences, syntax seem to elude Adam's grasp, and time is running so very fast.
I've been a backend/systems developer almost all my life, with not a lot of frontend experience (although I do know basic HTML/CSS/JS), and no app development. So I'm thinking for now to prototype this as a web page/pages, maybe using a lightweight framework rather than vanilla HTML (not sure), and let him interact with it on the iPad. I'll try to get the basic visual elements (picture/rows of squares for typed letters/bag of letters to choose from below) right with CSS/JS, and see if I can iterate from that. That's the idea, currently.
The photos too. That kind of stuff with the text overlays is way easier to build in SwiftUI; especially in a tableview.
Turned in to a pretty boring post, since I gave no detail!
I’m also working on the next version of HacKit, a native macOS reader for Hacker News. You can already download it on the App Store: https://apps.apple.com/app/id1549557075, and you can read more about it here: https://github.com/anosidium/HacKit-Feedback-And-Support.
the process of creating APIs for testing and automation should be as easy possible. the tools that exist nowadays aren't good enough, they require you to use their programming language of choice or complex procedures for a task that should be simple. I built mock to try to solve that and still continue to maintain it.
Currently adding support:
- loading/saving models
- model library
- simple math in parameter definition (for example, defining beta=2*eta, where eta is defined previously)
- viral intra-host models
- demographics
- arbitrary seasonality functions
The goal is to have it all ready for when my Cambridge epidemic modeling review gets published in a couple of months.
It's my first serious package, so I would love any feedback
reciperium is a small platform to write recipes and easily fork other recipes to adapt them to your liking
What absolutely grinds my gears is when a loud motorcycle or sports car drives through my residential neighborhood right after patiently rocking my baby to sleep.
Interestingly, as a motorcycle and sports car owner myself, I never thought about that aspect for even a second until I became a dad—I drive much more gently nowadays (especially in residential neighborhoods)!
Recently it got a surge of users (1k+ reviews on Google Play and 500+ reviews on Apple, really sucks that Apple don't show all reviews, but you can check the Google reviews here on desktop https://play.google.com/store/apps/details?id=me.audiodiary....) and people are writing in every day to say how much they love it and that it's changing their lives.
Also working on a new app in a similar vein that's way more technically complex and uses AI in a hands-on way, and looking for help on it.
My goal is to build a comprehensive database of font usage across the web. By collecting and analyzing this data, I believe we can uncover valuable trends, such as:
* Common font pairings * Popular heading fonts over time * Market share of commercial fonts * Top font foundries based on actual usage
I originally built a version of this four years ago and saw a surprising amount of organic interest. I've now rebuilt the tool from the ground up, switching from a Puppeteer-based crawler to an invisible iframe approach. (More details in another post)
Check out the current version at https://fontofweb.com. I would appreciate any feedback
I've never thought they worked really, but this is a new one. It has very prominent red and blue flashing lights that trigger if you are 5+ over. I've seen countless people slow down immediately, it's that jarring/terrifying
- A little free library, but for e-books. Having a bit of trouble with this one because I think that the move to e-books inherently removes much of the magic of a little free library of physical books. Plus there's the whole "letting users upload things is hard" thing.
- E-ink picture frame. It's been done before and it's mainly just a use for an old rpi laying around.
- Looking to start a tech meetup in my small locale. It's hard to meet tech people in my area, let alone people who are willing to present.
- TUIs to aid me in my day job. Claude makes whipping up proofs of concept super easy and quick, so this one is the most fun to me right now.
This made me chuckle.
Initially was aiming to use MLIR or at least LLVM but will probably try to handroll to a) reduce dependencies and b) as a learning experience.
The bootstrapped is written in CL with no dependencies and hopefully soon it will be self-hosted.
Right, because the car's cabin and motorcycle helmet protect the driver/rider from it. The noise they create are specifically designed to be heard by everyone but the operator. It's like me sitting in my back garden wearing earplugs, while blasting music out the front into the street.
This is first attempt, working on another now: https://www.instagram.com/p/DGdORIKNQ5w/
It's similar to Hacker Newsletter where I pick the most important news, features, reviews, etc of the week and send that every Friday. The gaming world has a lot of stuff going on and I always found that there was a missing newsletter to curate the important stuff from all the noise.
The challenge at the moment is growing it, I've been doing it for 5 years now and still haven't found a way to increase the number of subscribers, they come mostly from newsletter directories and referrals I would guess.
When I'm starting a new hardware design, I find myself pulling up familiar boards (like Adafruit or Sparkfun's dev boards) as often as the chip's application note. I sometimes prefer a full reference project so I can get useful context like which voltage regulator they used or how the USB port is connected.
But, it's kind of an awkward process because I'll have to download the design files from Github and open it in the native CAD software (Eagle, for example).
I've been toying with how to solve this. I made a script to crawl Github for open hardware designs, then generate a schematic and interactive BOM for each design. Now, hopefully, you can search for "ESP32"[1] or "WiFi"[2] or "Bluetooth"[3] and get a number of designs to view in browser.
[1] https://www.openappnote.dev/tags/esp32/1 [2] https://www.openappnote.dev/tags/wifi/1 [3] https://www.openappnote.dev/tags/bluetooth/1
Interested in this? Email me. My email is on my profile.
I had this written in Kotlin several years ago but now I want to do it all in Java, use as little 3rd party deps as possible and add more extensive unit testing.
As far as I'm concerned FreeCAD has been the world's leading CAD program for at least the past 8 years (which is how long I've been using it).
All of the competition is at least one of (worse; proprietary; doesn't run on Linux), which means FreeCAD beats all of them in my value system.
Apparently finding and setting up monitoring is too problematic for all of the people who are running their freshly-created startups and e-commerce shops without any monitoring.
For someone who does not know HTTP from HTTPs or all the different kinds of probes, reading about them and making decisions about what to set up is a time investment. Complicated might be the wrong word.
All you really need to do is to provide a domain name and an email address for alerts to get basic monitoring set up. My goal is to set up monitoring with zero decisions under 30 seconds.
Getting pretty close with nixmonitor.com
Recently I landed C exceptions support (I didn’t know that was a thing but it is, look for attribute cleanup if you want to know more) and ifunc support.
More info about the project here: https://github.com/pizlonator/llvm-project-deluge
And a Linux/X86_64 binary release if you want to play with it: https://github.com/pizlonator/llvm-project-deluge/releases/t...
Writing the specifications of a file format which I'll be using for the second iteration of an high-performance material simulation code I have written in my Ph.D.
Detecting CPU and GPU specs browser-side is almost impossible to do reliably (even if relying on advanced fingerprinting and certain heuristics).
For GPU’s, it may be possible to use (1) WebGL’s `WEBGL_debug_renderer_info` extension [0][0] or (2) WebGPU’s `GPUAdapter#info` [1][1], but I wouldn’t trust either of those API’s for general usage.
[0]: https://developer.mozilla.org/en-US/docs/Web/API/WEBGL_debug...
[1]: https://developer.mozilla.org/en-US/docs/Web/API/GPUAdapter/...
I started this for a bunch of reasons but mainly to allow non technical team members get any data insight they might need, without wasting dev resources on creating dashboards, queries etc.
I personally dislike this "everything in LLM/AI has to be a chat room" approach.
working on making it generally available but for now its early access only https://askquery.ai/
If you have any ideas, thoughts or concerns please let me know.
I aim to create a stable and affordable tool that allows me to eliminate most of the support code I write for web tests (page objects, locators, etc.) and replace it with human-readable actions and assertions. These actions and assertions are then translated by an LLM into browser instructions. The tool, however, should still leverage all existing infrastructure (test runner, CI/CD, Selenium infrastructure).
So far, it's working well on simple websites (e.g., a calculator, TodoMVC), and I'm currently working on scaling it to large web applications.
https://www.waitwhensyourbirthday.com/
something I am working on to help people keep track of birthdays. many people I know use facebook only to keep track of birthdays, so this hopefully will be a replacement
Most existing solutions only validate at the destination (dbt tests, Great Expectations), rely on aggregate comparisons (row counts, checksums), or generate too much noise (alert fatigue from observability tools). My tool:
* Validates every row and column directly between source and destination * Handles live source changes without false positives * Eliminates noise by distinguishing in-flight changes from real discrepancies * Detects even the smallest data mismatches without relying on thresholds * Performs efficiently with an IO-bound, bandwidth-efficient algorithm
If you're dealing with data integrity issues in ELT workflows, I'd love to hear about your challenges!
That’s the spirit! A buddy of mine created a tool/API to solve his own problems, opened it to the public and launched it, but he got very few sign-ups, so he just continued to use it for himself. Recently, about two years later and out of nowhere (without doing anything other than the original bit of SEO he did), he started to see a bunch of sign-ups (including paid) and then started receiving feedback and support request emails from customers. Most folks would’ve just called it quits on the product at least a year ago, but he just kept using the product for his own projects and left it open to the public just in case it was helpful for anyone else. Obviously, YMMV, but good luck with this!
I’m on mobile right now, but I think I’ll give it a try when I’m back at my computer.
Any way to follow?
After the chrome MV3 migration which was an absolute rigmarole, I lost my taste for beauty. I want to get back to making beautiful software.
1. It's unreasonably expensive. A single test "2+2=4" for a web calculator costs around $0.15. I run roughly 1k tests per month on CI and I don't want to spend $150 on those. The approach I took with Alumnium costs me $4 per month for the same amount of tests.
2. It tries too hard to make the test pass even when it's not possible. When I intentionally introduced bugs in applications, Computer Use sometimes pretended the everything was fine and marked the test passed. Alumnium on the other hand attempts to fail as early as possible.
Edit: It would be great if you could set the context and AI would generate it. It would make as an amazing addition to a standard Readme.
Further, at one level it could show endpoints and function signatures with parameters and how the argument usually looks as a value.
Which brings up another point, why doesn't Cursor or others allow me to say, "I'm in debug mode, show me if a value is dissimilar the values you normally get."
Turns out the whole app needs only ~ 10 SQL requests, and it's way funier to write modern SQL than fighting the ORM.
The new code looks like this :
db_message = Model.save_new_message!(@conf.db, DbButler::Hn, item, DbState::Processing)
def self.save_new_message!(db, butler, external_id, state) : Message
sql = {{ read_file "./db/save_new_message.sql" }}
db.query_one(sql, butler.to_s.downcase, external_id, state.to_s.downcase, as: Message)
end
The land page is not ready, but the bot has been working for me for months https://newsbutler.xyz/
Plus I having fun plugging it all into ChatGPT and reading the stories it comes up with.
Generally speaking models seem to be bucketed by param count (3b, 7b, 8b, 14b, 34b, 70b) so for a given VRAM bucket you will end up being able to run 1000's of models - so is it valuable to show 1000s of models?
My bet is "No" - and what really is valuable is like the top 50 trending models on HuggingFace that would fit in your VRAM bucket. So I will try build that.
Would love your thoughts on that though - does that sound like a good idea?
Now i´m trying to port to Cython to make as fast as possible. Here o3 is almost useless, but I´m progressing.
When it comes to "how to do the math" this repo was my starting point: https://github.com/Raskoll2/LLMcalc
Over at macos9lives.com a group of hackers figured out a way to get Mac OS 9 running on these late model G4s that previously never supported it. That combined with an SSD upgrade makes them close to the fastest machines that can run Mac OS 9.
I’ve taken advantage of this hack, now having sold about 80 -90 machines. But I’ve hit a wall with finding ways to advertise it. eBay has been okay. I tried Reddit Ads on the vintage Apple subreddit and they were so so—probably lost money doing it but got the word out. Google Search ads have surprisingly been ineffective. I’ve posted on various vintage Mac forums but they don’t allow formal advertising (otherwise I would buy it). I probably will try Facebook ads next. Open to other advertising ideas!
"Visit QR code to:" – You don't "visit" a QR code. Perhaps "Scan the QR code to:"? (Optional grammar fix included.)
"Comment on the QR code" – "Leave a comment"? (Unless you really want a comment about the QR code! ;-) )
I just get a busy spinner when I scan the QR code in the screenshot. A working live example would be nice.
Released an updated version of free app [0] for watching the news show Democracy Now [1]. Let's you browse 29 years of back episodes. Learning SwiftUI - mostly great, but when there's problems, it can be pretty frustrating.
[0] https://kenschutte.com/democracy-now/ [1] https://www.democracynow.org/
I did look at auto-detecting before, but it seems like you can only really tell the features of a GPU, not so much the good info (VRAM amount and bus speed) - is that the case?
I looked at the GPUAdapter docs, and all it told me was:
- device maker (amd)
- architecture (rdna-3)
and that was it. Is there a way to poke for bus speed and vram amount?
It's not 100% finished yet, but I've been using it myself on my freelance projects.
It's meant to work entirely offline and the service worker acts as the backend for the application.
«Text-editors are dead as a concept. What’s needed is a text-input system. Mobile phones got it right more than 10 years ago. Both Android and iOS can catch the text-input context: «ah here we can input text, let’s show the virtual keyboard!».
This project is inspired by very same idea: catch the text-input context globally (across all system, not just one process) and do what’s needed: change the UI, keybindings, etc. Emacs got some part of text-input right with modes. But modes should be global, on Window Manager level (or even deeper).
In GUI it’s possible to “catch input context” using Wayland::InputMethod
It should also be possible on pure-tty with readline or something.
The system should be very hackable. That’s why it’s written in Common Lisp»
2. SaaS Sales platform on Clojure(+script)
Where can I go to learn more about your invisible `<iframe>` approach/implementation?
I appreciate that it's a heavy site, but just being honest with you - it doesn't seem worth the time optimising this by moving to another lighter framework at this stage of the project.
Sorry!
For the first project using this technology I decided to go for a simple media converter app. I have a 64MB sampler so I have to convert a lot of samples from lossless formats to mp3, there are two options I know: any random online tool from the google page 1 or ffmpeg cli. The former almost physically hurts, because I have to upload my files to some server which does the same ffmpeg instruction and then get them bytes back, that's a heck of an overhead!
ffmpeg-cli and some knee-made bash scripts is what I have been using for a long time. I love my console, and I spend 90% of the time inside, but then it comes to ffmpeg, it instantly feels tedious to use it. Finally I decided to make yet another GUI wrapper:
https://github.com/ilya-lopukhin/conversimp
The idea is to drag the files in, or select them via dialog, then run an ffmpeg conversion template with each file as input in a separate thread (need to limit theese btw). I decided to go fully open-source, and maybe promote it's usage over the online converter ad-farms which are really abundant. When I decided to publish this, I instantly understood it's going to be a tons of extra work, but in the end I want it to look nice and do it's job flawlessy, and at the moment it's a weekend or two from release.
Let me know what you think ;)
- Computer: Used Dell Optiplex, bought on the streets of Medellin Colombia
- Operating system: Debian Linux / XFCE
- 2D design: Inkscape
- 3D design: Kiri:Moto (https://grid.space/kiri/), soon FreeCAD
- CNC controller: Universal G-Code Sender (UGS) --> GRBL
I am passionate about "maker-space entrepreneurship" (it's the dream job), so I'm meeting potential clients in my city to understand their ambitions and challenges, so we can work together to make useful prototypes to solve them.
Useful forums:
- https://forum.makerforums.info/ - This is my "new Hackernews"
(2) Affirmator - I'm building a system where you provide your goals as affirmations (example: "I feel healthy, fit and strong", "Today's the day to make it happen"). The system then uses Text-to-speech to generate voice audio files. Next Affirmator uses `mpv` (media player - similar to VLC) and `cron` to automatically shuffle-play these affirmations every day in the morning & evening. I recently used Python and FFMPEG to add "vocalization pauses" at the end of the affirmation audio, so that you have time to say the affirmation out loud.
Some of the driver & inspiration for this program is Earl Nightingale's "The Strangest Secret", Napoleon Hill's "Think and Grow Rich", James Clear's "Atomic Habits" and other personal development programs. These books always have these "every day, you should X" (meditate, write your goals, etc) - and I became frustated - Just how many "things" are you "supposed" to do per day, and how do they fit together, and how can you create helpful reminders & to turn these into habits? So far, I've been testing the system on myself, and it's been a HUGE help towards giving purpose to my life and day, and reducing feelings of depression & insecurity.
The solution is completely open-source (useable as web app, local docker container, and installable on a stand-alone computer - which I recommend), and I'll also soon offer it as a for-pay service.
If anyone is interested in these projects, I welcome contributors!
// JRO
I’m currently creating a new fan site for Marvel Rivals (https://marvelrivals.app), and I’m trying to introduce new types of features using predictive analysis, and further, use some DL maybe to understand specific player behavior and do stuff like find cheaters. I am failing so far.
I thought it’d be easier to throw data at the magic AI monster, but it’s still garbage in -> garbage out. It makes me respect AI engineers a lot more.
I wish there was an easier way to apply AI to this kind of stuff, on the how to do better data analysis. Ideally, I’d hook up some tool to my Postgres db, which has a couple tables but everything is named appropriately and has references. Then the tool would output correlations, patterns, stuff people would find useful and interesting. Instead, right now, I think I have to make those guesses and then build models that will either support my hypothesis or reject it, but I don’t know ahead of time and it relies on my gut feelings.
- Use natural language for telling offloading requirements.
- Just year of the LLM launch of HF url can help if it’s an outdated LLM or a cutting edge LLM.
- VLMs/Embedding models are missing?we originally built this for our previous agent startup as an internal solution to ensure agents could find the relevant data on apps they're using. We then pivoted to this after some early positive reactions and decided to open-source it.
here's a short demo: https://tinyurl.com/demo-airweave
we're two engineers/friends based in Amsterdam, NL. We just launched the project, so it's rough around the edges ofc, but we're very eager to get some feedback!
feel free to reach out to me personally if you like this! - https://www.linkedin.com/in/lennertjansen/ - https://x.com/lennertjansen
All the work is in collecting and entereing data and hopefully recruiting folks around the country to go to their local church/synagogue/mosque, government building, or glass shop/studio and taking photos and collecting information on glass pieces.
Site is still a work in progress, but if anyone out there is interested you can find it here: https://www.stainedglassatlas.com
... also continuing to not add features to my (not-much-of-a) system for getting more done each week. https://carpeweekem.com
I've also been cautiously adding AI-powered features to my Heroku autoscaling tool (https://flightformation.com/) and a simple free-text time/date input has been the most popular (demo: https://x.com/ejschmitt/status/1893268742760448497)
Eventually someone died, and they added a lot of traffic-calming changes to the road. It's much nicer now, but a shame that someone had to die to change it.
Augmented Chess: Normal chess with conventional ELO ranking system but you get additional penalties & rewards based on common bad practices, and recovering from bad positions.
> The first person that breaks from the book line loses 10% of the start time (unless book was not an option / the line was exhausted)
> Missing a forced checkmate forces you to wait 5% of the start time before your next move
> Achieving any principled position good grants some time (Passed pawn, connected rooks, rook / queen / bishop battery, etc)
> Doing any principled bad position loses some time (Knight on the rim, blocked bishop, king past the first rank in early game or mid game
----------------------
Continued Position: Chess but you continue a position from a high-level chess championship. There are a couple value-adds here
> Provide lower level players with a way to start a middle or end game positions after a highly skilled player followed all the correct principles. My theory is this will reinforce why those principles exists, how they can benefit you, etc.
> Provide high level players a way to be forced into positions out of their comfort zone / their preferred styles
> Provide differently-skilled players to continue play from unique positions with the desired amount of odds. So a GM might play a 1000 ELO player but starting from a position with -9 evaluation, etc.
If you have any ideas, comments, or feedback LMK.
Most of it involves taking advantage of data structure properties (and limits) by using zig comptime to derive functions that either compute offsets relative to existing pointers or use pre-computed offset tables, when relative isn't possible, to reduce function size further without inhibiting the ability to take full advantage of SIMD.
One of the next task for this is statically computing update graphs for archetypes such that a multi-thread runtime can mix strategies (last thread (detected by an atomic counter on nodes that require all dependencies to be complete) to reach a node broadcasts new work it unblocks, starved threads steal work from others, etc) to speed up the world update loop when running on larger targets while also remaining lock-free.
It's fun to explore how far one can go with statically declaring all limits upfront and managing even larger targets (steamdeck, servers) as if they were embedded applications.
https://www.langturbo.com (No signup required)
I have Python experience as a data engineer and I want to revisit Django. And my kid is getting into the Roblox studio GUI tools and I want to work on a Lua project with him to get him started with SDLC concepts and an IDE.
TBD exactly what that will look like or what direction it will go.
It's somewhat similar to how browser proxies like: https://proxyium.com/ and https://www.proxysite.com/ fetch the html on your behalf.
https://github.com/mech-lang/mech
Mostly a research project until I find some more people interested in pushing it further.
A recent blog for anyone who wants to check it out: https://mech-lang.org/post/2025-01-09-programming-chatgpt/
And a 10 minute video: https://www.hytradboi.com/2022/i-tried-rubbing-a-database-on...
You might want to pass the landing page copy though a spell checker or LLM:
User-Firendly InterfaceIf you ever have feedback/advices, don't hesitate to reach me out on contact[at]halftheopposite.dev and I'll happily answer.
A few months back, I got excited about pulse generators that had rise times on the order of 15 to 30 picoseconds. There aren't a lot of those available, and I was curious about what would go into their design. so I decided to build my own. https://voltative.com/pulser
It allows us to control the algorithm. It’s all LLM translating to YouTube search queries under the hood.
Visually it looks the same.
The suggested videos come from predefined buckets on topics they love.
E.g. 33% fun math, 33% DIY engineering, 33% creative activities.
Video recommendations that have a banned word in the title/desc don't get displayed e.g. MrBeast, anything with Minecraft in it, never gets surfaced.
For anyone interested in using it, send me an email.
I'll put you on my list. And you can contribute ideas to our community Google Doc.
jim.jones1@gmail.com
The best option I've found for my son (8yo) is this, but it's a little dry for a child: https://programming-25.mooc.fi/
Right now it works as a mock interviewer for algorithmic (leetcode-style) problems, you can sign up for the waitlist here, I'll send you an invite right after:
It actually works pretty well, but we're having trouble getting users (some sign up but don't end up doing even a single interview?!).
We're thinking whether we could sell a version of this to companies to do their technical screens in, perhaps with problems that are more similar to the actual software engineering work (e.g. debug existing piece of code, write tests, and extend it).
We're generous with interview credits if you give us good feedback =)
We built an agent that can make sense of your website, understand how it renders on search engines, its weak points and strengths. You get actionable advice that can make a huge difference in search visibility, often taking less than an hour to implement changes.
But I imagine, if you realllly needed that info. You could go the hard route and render the font on a canvas, vectorise and perform some sort of nearest neighbour search.
This makes sense. Though annoying to drive over, speed bumps work. Also, my parents small town converted their main street from a four lane to a slalom-style two-lane and it so much nicer to walk around because people have to drive slowly. Not quite car-free, which I'd prefer, but it's a decent compromise.
1. API for Clean Transcripts – Extract structured transcripts with references, code snippets, and images.
2. AI Video Assistant – Interact with videos using AI-powered tools for summarization, Q&A, and more.
Would love to hear your thoughts! You can reach out to me at sg@o14.ai
So I built buckaroo, it combines a high performance scrollable table (built on top of ag-grid), with summary stats, and histograms. All of this is customizable and extensible. I recently built a dataframe compare tool [2] on top of buckaroo that uses coloring to show differences between dataframes intuitively.
Get in touch if you want to talk tables, data science tooling, or exploratory data analysis.
With English not being my native tongue, it gets difficult to speak some words at times and I have to reshoot each takes several times. But it's worth the effort and it's different to what I do in my main channel!
Danger World (https://linktr.ee/playdangerworld) is well underway and should be coming out later this year! Danger World is a 2D narrative adventure written using Flutter's open source game engine, Flame. I am the solo developer. I've been working with a very talented art team to do the art and animations in Spine. The composer has just finished the soundtrack!
Danger World is running on iOS, Android, MacOS and Windows. Last week, we got Danger World compiling on my Steamdeck for SteamOS instead of having to use the Windows .exe + Proton. I'll be getting the Linux depot setup in Steam soon.
In the spirit of doing things that won't scale to get some interest bootstrapped, I've been doing a sticker giveaway. I have a bunch of stickers featuring the cast of the game and the Danger World logo and I have been mailing them out to anyone who wants them. I've sent out 40 envelopes or so with about 10 stickers in them each.
So far, in experimenting with the various social media platforms, Bsky seems to be where my potential users are. I posted a boxart concept that I was using for IndieDB and got 100 followers from that single post of who all appear to be real actual users. Certain engagement metrics like impressions are non-existent in Bsky, but in exchange for real actual engagement from people who seem interested in my game's art, who cares?
X feels extremely P2W, except you buy their checkmark and they still don't show your content to anyone. I'm still experimenting heavily in social media land. Meta and Tiktok seem to show more promise, but since I've been scoping out and planning a visual novel mode I have also been considering taking Reddit up on their free ad credit.
I'd love to talk with anyone more experienced in marketing games like mine.
It's my personal decision and if I ever get fined for it, I will gladly pay the fine... with the money that my doorbell camera has already saved me. It helped me catch hit&runners that bumped into my parked car twice already, and the camera is now almost two years old.
It's not connected to the cloud, saves data locally, and only stores a couple of days of video. It's not very ethical to unknowingly film public spaces, I know. My lame excuse is that I personally think that catching people that damage my property with a camera is a lesser crime than damaging someone's property and running away. The sad truth is that living in a place where parents drop off and pick up their kids twice a day do not give two shits about others. Hit and runs happen every day here.
Right now we're trying to figure out our funnels. We've had some pretty good success doing things that don't scale particularly well, like going to immigration events. Now I'm working on getting better visibility across the country (SEO, Google Ads, etc...).
I'm learning a lot, but it's slow going.
These are audit data like external system request/responses for possible investigations. This saves a lot of space. Initially written in Python, now practicing with Rust. Container images is 2.2MB small :)
Our neighboring town has a beautiful system in place. It tracks your average speed over +- 100m. If you speeded, you get a fine and even better: the light always turns red and you have to wait for a minute. If you run that red light, you get another (heftier) fine. If you drive normally, the light is always green.
It's beautiful. People actually drive slowly and safely there. And good, because that system is installed perfectly around the school and doesn't hinder anyone else. If they did this anywhere else you'd argue it's a moneygrab, but this is imo a great way to actually get people to drive slow in a school zone.
In contrast the only thing our town has done is install a bunch of big concrete flower boxes (?) that hinder your vision and causes conflicts on the road. The idea being that you'd have to slow down and that you're no longer on a straight wide road. It's absolutely only made things worse. You need to give way to others but you literally can't see them, or you need to pull weird manoeuvers to let them pass. And in the meantime cyclists need to swerve around them or are hidden behind those boxes. It's terrible.
https://github.com/btahir/open-deep-research
An open source alternative to Gemini/OpenAI deep research. I'm experimenting with this cool flow chart workflow that I think can be a cool twist on generating reports for deep research.
Check it out and lemme know what you think. :)
https://github.com/turbolytics/sql-flow
The goal is to create a stream processing framework that supports SQL jobs. Apache Flink supports this but is very heavy-weight overall. I work with cost constrained companies that just can't run Flink but still want access to high performance streaming primitives.
We are taking a slightly different approach from other competitors by building cloud-native tools for engineers. SQLFlow is built on DuckDB, Native Kafka Library, and Arrow. This allows SQLFlow to handle ~70k+ events / second with low memory overhead (~250MiB).
Would love your questions, thoughts and feedback or feature Ideas! Thank you
Don’t have a background in ML, so mostly just for learning purposes
Been playing a lot with the MNIST dataset. Trying things like training only on the examples the model gets wrong, or training only on random samples of the image (ie. using only a small subset of the pixels of the image as the input), or creating one model per label to overfit the data and then merging the models for testing, or just testing performance of different architectures and frameworks on the same problem
Honestly it's already pretty narrow and crowded here. I'm a regular driver here (duh) and it honestly feels dumb to drive faster than 30 because a cyclist can pop out of every corner. It's crazy to me that people want to, and do drive faster than that. But here we are. People suck.
Otherwise (changing lanes, left turn across traffic), I think it’s more “not ahead of ME!” than about the signal itself.
[1] - https://github.com/KaliedaRik/Scrawl-canvas/blob/write-initi...
Applying the same approach to chain-of-thought reasoning gave me the feeling that I might be looking at a form of realistic UX for some sort of science-fiction neural AI augmentation - you can let the CoT run on and do its thing, but also interject at any point and insert a "thought" of your own, or go back and revise a thought you did not like, and then let it continue. Imagine such a stream hooked up with a two-way pipe into your phonological loop (https://www.sciencedirect.com/topics/psychology/phonological... - perhaps more attainable with existing tech).
Marketing/distribution/adoption is maybe the hardest thing about creating a new product or service
I think you have the right idea about focusing on companies, because they can be recurrent users, whereas people preparing for interviews will only briefly use the product
some suggestions:
* have ways for multiple views and/or start-points and/or both up/down directions. e.g. hierachy vs reason/effect vs dependence vs whatever-else. Then think about animating those in time
* heat-map over the views, as e.g. churn(changes)-in-time, or usage(number of dependents), etc
* requirements engineering kind-of-view ~ may overlap with dependency (both directions!) but with explicit requirements/assumptions tied to respective stakeholders. Though this may need links to/from JIRA and similar issue-trackers etc
* check Wardley maps - yet another view, starting from customer/stakeholder/vendor-points. Also may move in time. It may need user decision on which things are big/separate-enough to surface on that view - sometimes a single script is on-par with whole subsystem
* future maybe - growing above into zoomable per-project thing (more proj.mngmnt than just code, incl. related e-mails etc) - described here: https://news.ycombinator.com/item?id=43060108
* probably more.. will add if something else dawns on me
have fun!
p.s. fractional CTO? i am looking that way too..
The visualisations generated by the AI from the agreement.
Click on legend text to drill down and on circles in the spiral. UI is 3D cube animation for now.
Former Deputy AG in charge of fighting corruption and fraud affecting US consumers: “I like how it balances the intent of the law for the consumer”
Both academic and practicing lawyers have looked at it and given us good feedback.
Not looking for a problem to solve. Found our niche. We applied for YC but we are all on the older side and the first question in the YC application was our age. LOL. Not gonna work. Doing another round of family and friends fund raising then VC post market traction.
Samples of ToS and financing agreements
Https://labs.sunami.ai/feed
Still needs some work for mobile/responsive version.
DM marc fawzi on LinkedIn.
Video: showing drills downs etc
By this logic, are you saying cameras on private property, that happen to capture public space in their frame are also illegal?
We're starting with cat, dog, horse, and pollen allergies.
We're close to peanut allergies.
The science is known as allergy immunotherapy. https://www.wyndly.com/pages/immunotherapy
What do you think? https://brainblitz.app
I professionally have some projects where we are doing custom CDC from kafka -> dbms. In that context, we're just using spring boot apps. How is the troubleshooting/recovery story for sqlflow?
Which feels an arbitrary line to draw. What about other road rules? Driving on the right side? Using unregistered vehicles?
Countries like India, China and South East Asia, have much less enforcement of these rules… Would you call their driving experience optimal?
Not sure what you mean by values software privacy. I don’t really have a view on it? So no, I don’t think I am.
At work, I'm finalizing a platform to quickly set up and deploy Apache Flink clusters and pipelines. It's going surprisingly well, with several teams interested in using it! Unfortunately we couldn't use the extremely cool Flink Kubernetes Operator because it's incompatible with the existing infrastructure stack at my company... but rolling out something similar enough hasn't been too hard.
At home, I'm about to receive my new robot vacuum and the gear I need to install Valetudo in it. I'm excited because I've had nothing but bad experiences with cloud-connected vacuums. To give you all an example: my current Shark vacuum sometimes starts by itself in the middle of the night. When I get up to stop it and check the app, I see that the floor plan it tried to clean is not my house... so it's receiving cleaning commands from other users.
I'm also learning JavaScript through Daniel Shiffman's book Nature of Code. As someone who has a background in modeling it's a very engaging read, and the exercises are fun!
At my organization, we have a corporate-approved password manager, but shared folders are disabled. If someone needs a FID password, or something of the like, there are people that send the password in plain text over Slack, Teams, etc. and not everyone deletes their message containing the password afterwards.
You enter your secret on the website and it gets encrypted on the client-side before being sent to the database. A one-time-use URL is generated and then you send that URL to whomever needs the password. The backend never sees the secret in plain-text.
The tool is meant to be self-hosted - but I'm in the process of deploying it publicly for people to try out. I got the idea from my brother -- he was doing an internship at Tesla and said they had this being used internally, so I figured I would create my own as a little side project that I can then implement at work.
There is a lot of potential here to help, and I don't have big ambitions to scale this to a lot of children. There are also schools that need help. Other people that volunteer their time to run a daycare center for those that can't afford a regular one. One school is looking for someone to teach programming, but doesn't have the budget to hire a professional. I could volunteer to do it, if I could find a sponsor. It doesn't even require a full-time salary.
* There is possibility of adding relations between schemas (1-to-n, n-to-n, 1-to-1)
* API allows to filter and transform responses to your own arbitrary schema via custom query language
* support Oauth with custom Oidc of your choice and API tokens for easily configurable but strong security
* from the start I designing it with extensibility in mind so there is built-in system of flows and custom extensions
* flows support running via custom http endpoints, and system events (i.e. data changes); cron schedules and external events are planned
* flows are defined via Json files
* flows are also customisable and allows to define your own blocks and flows that can be injected into other flows for bigger pieces of logic.
* all json files are source for code generation so performance should be the same as using your own manually written code or at least very close.
* there is built-in template subsystem, used for flows (i.e. for saving files with content from template) that I also plan to use to generate static files for FE.
* custom pieces of logic are supported by flows so you can create flow to update some read-only values based on other values (i.e. FullName from first and surname).
To achieve that I had to write my own fluent API for C# code generation. I had idea mainly during writing my own API for my house automation: 'Why I have to write another DAL layer, and auhorization, and filtering, and database schema? I did this dozens of times already?!' There is no need to figure this one out every time.
It's also a hodge-podge of other annoying things I want removed on the web, like when articles have pull quotes that repeat what they just wrote one paragraph ago, or when sites block the ability to paste passwords.
The extension is available for Firefox and Chrome/Vivaldi, but it has a lot more development to get it where I want it.
Would appreciate any feedback and tips on both the extension itself as well as advice on monetization.
* wav -> mp3: lame
* anything -> wav: mplayer (or the specific format decoder if one bothers)
* resampling: sox, it does much better quality. But it does not do mp3, so wav > wav
have fun!
The core components are on GitHub:
* https://github.com/m3047/shodohflo
* https://github.com/m3047/rear_view_rpz
* https://github.com/m3047/rkvdns
* https://github.com/m3047/rkvdns/blob/main/SHARDS_Command.md
* https://www.zipcon.net/~m3047/observability-dream.mp4
https://github.com/xyproto/algernon
A snappy and configuration-free little editor/IDE for the terminal:
Next I'd like to add a Pi into the espresso machine for some remote fun. A couple of the control panel pcb-pop-style? buttons are failing, and I thought it might be fun to try and control via a touch screen. I suspect I'll want some kind of timer on the machine side to ensure a safe/timely cut-off, but I haven't done anything like that since using a 555 a very long time ago. Maybe I'll just drive a relay from a digital Pi i/o pin, but suspect it is switching mains level V.
Feature request - Have a leaderboard of LLM for x/y/z tasks or pull it from existing repo. Suggest the best model for given GPU for x/y/z task.
If there is better model which my GPU can run, why should I go for the lowest?
Code-wise I didn't take inspiration from anywhere, I followed "C Programming - A Modern Approach" and based my C knowledge off there. There was also a YouTube video I watched regarding raycasting, and watching people optimise Minecraft etc.
Other than that, its just been alot of trial and error and trying many, many different ideas out.
What has helped is I don't worry about cross-platform ability, I just implement what I want and what feels fun and "innovative".
But I have never seen anyone do a Raycaster with as many vertical/z-levels as I'm planning (1000's), maybe the most I've seen is ~10.
Also most of the approach I'm taking is not best-practice, I'm trying to come up with how to approach everything from my own mind which has been extremely rewarding
Computing the nutrients from the ingredients should be reasonably accurate post this because that has a direct mapping in the USDA FoodCentral dataset.
The idea is to reduce the amount of redundant work (clicking a button to add a single ingredient over and over again), and rather push for minor modifications instead.
Omini parser lets you split section of the UI to hash and watch for changes that are relevant.
For 2, can you give some examples?
I'm surprised at the relatively sparse state of streaming. I feel like there are:
- Flink - Spark - ~Benthos - Arroyo - Then mostly custom / bespoke frameworks
Maybe there just isn't that much money in it? But I think there's still lots of opportunity to improve the Dev/UX over the JVM and the enterprise solutions.
-----
I would say recovery is achieved through kafka consumer groups right now, which results in at least once processing semantics.
We also support websocket input for the bluesky firehose but that is completely ephemeral. I have a story in the backlog to write ahead to disk which should allow for tolerating process crash/failure.
The tumbling windows stores state using duckdb. The end user can configure a disk-based duckdb databases which achieves durability as well.
The troubleshooting was actually pretty funky, i didn't realize how tricky it could be with stateful stream processing. I certainly don't have a good story for it, but what I did was added a sql debug HTTP handler. When this is enabled it exposes the duckdb execution context over HTTP. This was how I debugged the tumbling window logic during development. I would start SQLFlow, query to make sure the windows were empty, send some known messages, query the windows to make sure they were aggregated correctly, wait for the windows to close and flush, query the tumbling window state, etc.
For troubleshooting it also supports prometheus metrics oriented towards stream processing, # of messages, processing duration, success/failures, etc.
SQLFlow also ships with a dev framework, this allows you to execute your sqlflow pipeline against a json input file to make sure the sql processing logic is correct. I wanted to give the ability to decouple testing the logic, from actually having to stream data.
I would love to know about more what you're hacking on!
Is it that much work in your mind to be social with the people you care about or are you just unable to drop the corpo think at home?
I've wanted to work with code thus, but it becomes a problem in that things are only really readable/relatable in single-screen chunks, and when one tries to show more than is accurately related on a single screen it becomes an unreadable, confusing, blurry mess.
Screengrab?
Unfortunately, I’m not aware of any way to reliably get this type of information browser-side.
If you really want to see how far you can get, your best bet would be via fingerprinting, which would require some upfront work using a combination of the manual input you already have now and running some stuff in the background to collect data (especially timing-related data). With enough users manually inputting their specs and enough independently collected data, you’d probably be surprised at how accurate you can get via fingerprinting.
That said, please do NOT go the fingerprinting route, because (1) a lot of users (including myself) hate being fingerprinted (especially if done covertly), (2) you need quite a lot of data before you can do anything useful, and (3) it’s obviously not worth the effort for what you’re building.
So I'm experimenting with ways of making it easier to write useful logs.
The ironic thing is that not only are the tradies expensive, but prices for renting specialty machines are also crazy high where I currenly live. Like, I looked into renting a stand-up concrete grinder/sander -- $100 for 4 hours! Plus, you need to buy the sanding attachments which are another $200 if you go for diamond disks. I'm now actually considering using just a hand grinder and spend a whole week on what I could probably do in a day or two otherwise.
Version control could reliably track movement between files, we could identify copy and paste (and prompt when refactoring), and code coverage changes would be accurate.
Alongside that, I've also been coaching kids (ages 4 to 11) at a local Benfica academy since September, and I've had a lot of fun helping them improve in many areas of the game. It's a truly enriching experience.
Most days the road noise itself was no big deal, if the wind blew it was gone, if you listened for it you could hear a whooshing/droning in the distance.
But all it took was one loud motorcycle or old car/truck to really ruin your peace for a solid 3 to 5 minutes, being able to hear it on approach, and it went by, and exit. And it wasn't that rare to get them back to back, seeing as it's a 4 lane highway with a decent flow of traffic.
It wasn't really until then I realized how a single vehicle could cause so much annoyance for so many people.
Saving Voyager 1! - Bruce Waggoner at !!Con 2024 - https://www.youtube.com/watch?v=dF_9YcehCZo
Edit: Also, please do share when it's playable :)
I have never seen a system like that. What town? Or what system?
content wise, i found the first lesson to dry. i'd rather start with something interactive, and explain necessary concepts along the way. the print lesson introduces the separator feature which is rather rarely used and should only be introduced when there is a practical need for it.
I spent almost 8hours a day over the course of a month working on and off on the project.
Did give me tunnel vision but, no regrets
I've built a large solar array on a repurposed mobile home chassis and have been digging a trench through rocky ground to lay the conduit for the cabling to backhaul the DC power to the shed where the inverter and battery bank will live. It's the rainy season in Ireland, and that field was marginal at best, so it's been swampy work and half of the digging is trenching additional drainage channels so the conduit doesn't flood and path water back into the shed. There's a housing order due to a bird flu scare, so I'm not running poultry on pasture until it's lifted (or until I can get the government to agree that my pasturing system qualifies as housing). Annoying, but that's farm life.
Oh, I'm doing some shit with AI also, but that's a secret :)
Not planning to open source, working on a commercial offering but haven’t launched anything publicly yet.
Would love to hear any more thoughts on the concepts here or my email is in bio
ATM we basically have lawless highways in California. Every time I go out I see outrageous traffic violations. People driving 15-20 mph over the speed limit on the freeway, weaving in and out of the carpool lane. People pulling into the left turn lane at a signal, then just crossing on a red light as soon as they think it's safe. People passing in the gutter (no lane). People cutting across 3-4 lanes of the freeway at ~70mph, across the painted barrier to exit a freeway.
I'd say the average of these incidences I see is 3 per drive and that doesn't include just you average weaver or tailgater. In other words, I take my car out for a 10-15 minute drive, (so 20-30 mins round trip) and see someone do something outrageous endangering other people's lives 3 times per drive on average.
I'm happy to reach this first milestone, but I'm already envisioning future iterations. My goal is to transform this into a proper celebration of Unix – a system whose elegant simplicity and powerful design principles continue to inspire me. There's something deeply satisfying about connecting with this foundational piece of computing history.
Contributions are always welcome.
Special shout-out to @abetusk for all the support and enthusiasm with the project.
https://github.com/StarsRail/Cocai
Better than a human, it can draw an illustration of the scene in meme seconds.
It can also roll dices in 3D: Demo: https://youtu.be/8wagQoMPOKY?si=oCa2erHvheEyEKM_&t=55
It’s for iPhone only (yes, sorry!), and you can play for free once. Apologies in advance for the large download. Get it from here: https://www.inotherwords.app/
> I'm struggling to comprehend why Apple is ditching Objective-C
I am too studying a lot and trying to build one from scratch. It has been very humbling and enlightening experience, something that AI will never replace ;)
Kudos sir, and best of success. Check Maestronet (Luthier section, ie, Peg Box) for insightful comments and suggestions.
Either way, I didn't appreciate the 100+ entries in my history.
Currently working on setting up a blog with some dev details about the newest game. Also working with a UI/UX designer to make it look and flow a bit better.
* reversed engineered 90% of the data files
* finally got copying banks between projects working (a 10-ish year old feature request no-one has implemented yet)
* now tidying it up before I publish it on basically every public forum octatrack people hang out on and hoping I don't completely bork people's stuff in the process.
We’re also building our own CDN to optimise asset delivery and improve performance across deployments. But beyond infra, we’re betting on something bigger: developers want everything they need to ship an app in one place—not just UI kits, but fully hooked-up components that handle state, auth flows, and analytics automatically.
That means forms, a help desk, user management, and more—all tightly integrated so you can focus on building rather than stitching services together.
Would love to hear from anyone who’s been frustrated by piecing together third-party tools just to get a product off the ground!
- A webassembly bindings library for go, with "namespaced" / subpackaged adapters for Web APIs, so the final wasm binary won't be bloated
- A git management tool for my repositories, implemented in pure go (and using said wasm library) to find out what I need to build a small little app with it
We're also building OpenEWS (https://github.com/somleng/open-ews) which will be the world's first Open Source Early Warning System Dissemination Platform.
We're a small 2 person team trying to make a big difference.
https://www.youtube.com/watch?v=ZmCNMQ7TOgY&t=9s&ab_channel=...
Right now I am adding a couple of features as well as improvements to the UI and I want to make a proper pricing section for the landing page. Also need to figure out marketing. I enjoyed building it so much!
- get tldr of a long article
- draft personalized outbound linked messages
- search if a specific company or type of role is posted on a long HN WHO IS HIRING thread
- ask questions when reading research papers on arxiv without context switching into chat gpt, etc
think virtual table top, live sharing for players, map manipulation
a platform (not an engine) for building “native”, cross-platform games. exposes WebIDL inspired APIs, including audio, GPU, input, storage, etc. you can think of it like unity without the engine. eventually will support user libraries (read: engines) written in whatever language, usable in whatever language.
very rough, don’t expect everything (anything?) to work. might change the name due to the collision with the Rust library by the same name.
I live in EU city center. Stupid noisy speeders are bane of my existence. Both noise and safety (for biker and pedestrian) ... In some circles even raising this topic makes you whining leftist crybaby.
In summary, our tool lets you generate dynamic links, track attributions, and analyze referrals. You can easily see which links and marketing campaigns drive the most installs, reactivations, reinstalls, and views.
Beets takes almost 5 minutes per incremental update of ~1000 folders of tagged flacs with my current configuration, when all I really want it to do is:
- fetch album art if not present
- create folder structure readable by Subsonic server
- symlink relevant files
Very raw and unfinished, currently only implemented adding new albums. However, 5 minutes to <1 sec is a promising improvement.
i figure writing online would be a fun way to try and start conversations with people who are interested in the same topics. also, hopefully any technical posts can go to show employers that im not just another junior engineer whos reliant on LLMs for coding, and have some deep knowledge.
Why: - because stable voice AI is still extremely hard. I built multiple such platforms (inbound, outbound, support, phone based) last year. And learned, it is very hard to build something complex with current tool - the interview paradigm is common enough to be useful. E.g: off boarding, business insights, pre/post lab experiment interview, market research
A tool that builds job-specific tailored resumes. It's not a novel idea, but the ones I've seen use a subscription based model, and as a user I feel more inclined to purchase some tokens for usage based approach. Also my wife has been looking for a job for a time and I thought it would be handy for her.
Pretty close to getting ready for launch.
It's quite jarring to come home and drive an expressway in South Australia where the "speeders" are doing 105 km/h on a 100 km/h route. Someone doing significantly more is a rarity where I live.
Pycharm AI agent (just started): https://github.com/zerubeus/aladin
Alexa skill to set Muslim prayer reminder, ask for prayer times, and so on: https://github.com/zerubeus/alexa_adhan
Why? Because I thought it would be a fun challenge. I have a win 98 machine. I have Vim 7.3 installed. It has not been compiled with lua/python. I want a minimap.
It's been fun to find clever compromises to make it work. I've almost completed the project. (at least it works on my modern computer, I'm sure when I transfer it to my Win98 I'll find I'm using some modern vimscript functions and then I have to invent fun workarounds again).
If you want a good dev experience on a Win98, I highly recommend Vim or if you don't vim, then notepad++. Vim is better though.
I am buying all the tools. AliExpress price and quality level if I don’t need them much. Good quality ones for the bigger tasks. I tried renting at the beginning, but it’s very expensive at the end. Sometimes it’s more about convenience and not money. Having the equipment at home allows me much more relaxed planing.
Ever had an amazing app idea but felt overwhelmed by the thought of actually building it? Say hello to Tiram.ai! (https://www.tiram.ai)
Tiram.ai is here to turn your vision into reality—whether you’re a startup founder, a product manager, or just someone with a killer idea. All it takes is a simple voice or text input, and Tiram.ai does the heavy lifting to create fully functional web, mobile, and logic apps tailored to your needs.
Why Tiram.ai?
Voice & Text Input: Just speak or type your idea, and watch it come to life.
Tailor-Made Solutions: It intelligently extracts user personas and crafts solutions that fit real-world needs.
Seamless Development: From concept to deployment, Tiram.ai makes the process smooth and efficient.
Whether you're building the next big thing or streamlining your business, Tiram.ai is your go-to platform for effortless app development.
Ready to bring your ideas to life? Check it out here: Tiram.ai (https://www.tiram.ai)
I'd be surprised if it came close to the accurate amount of calories even 50% of the time. If by there you are talking about accuracy.
How would it know how many calories my curry has from a photo? It cannot. It could be full of oil or have none.
How can it know how big the bowl of food is? It cannot.
It cannot know the actual fiber content of any baked good just by looking at it.
The promotional photo gives 7/10 nutritional score of pancakes with syrup? Why? because it has a couple of blueberries sprinkled on top? I'm sorry, but this does not seem like a useful app.
The promotional text on their site says: "It is consistently 90% accurate when predicting nutritional value of food" They might as well say 100%. They get to cherry pick the food they photo. It would be impossible to be 90% accurate if they actually included similar looking foods with different fat contents. Mash potatoes made with butter, without butter. Mash potatoes in a big bowl vs small bowl.
The site just reeks scam.
I realize it is a topic done over and over but it is 100% for my learning. I have generated some graphics locally using Flux model, will play around with sounds and music generation and spice it up in Ableton, but now I'm stumbling on the basics and I learn a lot through solving problems.
Far-reaching goal would be to release it somehow, somewhere - I don't expect to earn even a dollar on it, just for fun and even more exp.
3+ years ago i landed as architect onto a working system having 300Kloc js codebase across 20+ repos, plus other ~100 unused ones with unknown Mlocs in them, running over 25+ containers in 2 datacenters.. without neither good architectural diagram nor deployment diagram nor runtime flow diagram... took me quite some months to build some mental image of which is what and where and why. It was event-sourcing engine so event-flows mapping from parsing the source was doable (few weeks, python+esprima+graphviz), but all else... nope.
Anyway, do ping me if you feel like it, any time.
----
[0] https://news.ycombinator.com/item?id=43161332 - subimage.io
[1] https://news.ycombinator.com/item?id=42521769 - gitdiagram.com
[2] https://news.ycombinator.com/item?id=42976467 (manned service?)
[3] https://news.ycombinator.com/item?id=41393458 - codeviz.ai
We tested our competitors apps and make our own blend of tools that actually ease the process. Effortless script breakdowns, moodboards, scene-specific venue suggestions, storyboard from script generation (here we consider us the best), nice pitch deck generation and projects sharing.
The market is still a bit conservative about using AI, but we are looking for testers and enjoyers: https://blooper.ai/.
There's a free trial period for people who don't like to f a lot on pre-production (especially during conversation with stakeholders).
Price is also mild for industry, 49$ for up to 5 projects and 199$ for big guys who need process optimization on scale. So...that's it.
It's a way to create a visualization of your life in weeks, based on Gina Trapani's version that I saw here last week. Would love feedback if anyone has any!
https://www.amazon.co.uk/Art-Violin-Making-Chris-Johnson/dp/...
The main motivation for creating Circuitscript is to describe schematics in terms of code rather than graphical UIs. This makes it easier to track changes and version control with a text based schematic. I have used different CAD packages extensively (Allegro, Altium, KiCAD) in the past and wanted to spend more time thinking about the schematic design itself rather than fiddling around with GUIs.
The main language goals are to be easy to write and reason, generated graphical schematics should be displayed according to how the designer wishes so and to encourage code reuse.
Interestingly, there already were a number of times where a tool came in handy that I had bought as some point in the past not when I had a need for it at that point but because it was on sale. And if not, I could always resell it, although I probably won't because it's not worth it.
As for the concrete grinder, I actually looked into buying one coming from the same mind set. But the kind of machine that I could rent for 100 bucks costs North of $3000 if I bought it new. And I will likely never need it again after using it once. So this time, I'm surely not going to buy one.
With the windows, I know there's quite a difference in terms of installation depending on where you live. For instance, in the US, you basically buy a window with a flange on it, put it in the whole and screw it into the studs. Then you foam spray the gaps and trim the window. In Germany, installing a window (to code, mind you) is a whole production. Even just measuring what kind of window to order is non-trivial. And I'm sure there's lots of variations in other countries as well.
It has taken a couple of months to go from idea to a product that's polished enough for other people to use, and I've been full time on it. It has a couple of dozen companies using it now, almost all from the last couple of weeks. That's been a big boost!
My plan over the next couple of months is to build the option for users to enter the kind of things they want to scan for, have AI convert that to keywords, use the keywords for the (fast) scanning, and then apply additional filtering using AI to the small number of posts that match.
Not built yet, but I think there's a bunch of promise to using AI to find relevant conversations online.
Re: APIs, yep, all APIs. I'm not doing any web scraping at the moment
Currently working on a few rust tools for drone stuff.
https://github.com/BWStearns/ulog-rs/ https://github.com/BWStearns/flight-engineer
"Violin Making - a practical guide" by Juliet Barker is also very good
I purchased the DVD series "When Trees Sing" by Peter Paul Prier (he was the founder of the Violin Making School of America in Salt Lake City). The DVD series wasn't cheap but it was worth it.
There is also an online course by Lucas Fabro, https://www.lucasfabro.com/en/online-course/, that I know is also very highly regarded. Fabro also has a number of free YouTube videos that are great to just get a feel for what's involved.
The Strad has a whole series of posters (https://www.thestradshop.com/store/product-category/lutherie...) with scans and measurements of famous violins to serve as templates.
International Violin, https://www.internationalviolin.com/, sells a lot of lutherie tools and is where I bought my tonewood.
https://www.reddit.com/r/selfhosted/comments/1iuqoyi/devs_pl...
It doesn't look like much, but there's a lot going on behind the scenes. I'm proud to finally make it live.
how cool is that?
If you're up for it, give it a shot at https://www.tiram.ai
It's super easy and fun!
"I want to sell coffee premix online"
And its give me this as output
https://app-01jmyp44-coffee-enthusiast-tiram-ai.vercel.app
isn't it cool
i’m learning golang and made this library that parses ham radio ADIF logs. my goal was to match the speed of the golang json parser. i managed to surpass it by about 2x!
i’m currently employed writing c#, but looking for a job elsewhere and golang seemed like a good way to level up :)
(even if poorly managed and overbloated enterprise codebases may lead you to believe otherwise, they are quite detached from what modern C# is supposed to look like)
I could tell you some horror stories about how my current employer does C#. It is very strange.... can't use var, nuget packages for _every_ class library... and it gets stranger from there on out... unfun!
How would you determine that something changed in UI by just looking at a screenshot? Would you additionally compare HTML/DOM or approximate the two screenshots?
> Omini parser lets you split section of the UI to hash and watch for changes that are relevant.
I wasn't aware, thanks for sharing!
> For 2, can you give some examples?
Specifically, if you take the Shortest tool (https://shortest.com), a test runner powered by Computer Use API, write a test "Validate the task can be pinned" for https://todomvc.com/examples/vue/dist/, and run it — it passes. It should have failed because there is no way to "pin" tasks in the app, yet it pretends that completing the task is the same as pinning.
The tools... I think, this is not my last construction site. I must buy some smaller property for my retirement, a house for old people is too much.
I’m not saying it’s not worth it, even 1 life saved would make it worth it, but it’s not beautiful.
Additionally, existing tools that I used struggled interacting with sites like reddit. So I set out to skip DOM and focus on a generalized approach.
I tried to go cheaper by using ui-tars, open source model by bytedance to run test locally without needing anthropic but it wasn't reliable enough.
That short test link is interesting. I didn't know they existed. Wow, the field is moving fast.
We found a nice design someone had released online under creative commons BY, modified it to fit our needs (though we’re thinking mark2 might need a complete redesign), then have been printing the towers out and growing veggies under grow lights!
We have a long way to go but I’m building the online store presently to sell towers (that should be up and running in a few weeks - I’m kind of exhausted after my day job - and we’re hoping to have enough towers to sell veggies at the farmers market by the summer.
Long term I’ll likely buy some land and build a less ad hoc setup, for now, we are still experimenting. The idea is to slowly build this up as self-reliantly and simply as we can so that anyone can grow their own food and so we can use dirt simple tools to run the business. We plan on doing as much as is practical with open source tooling and Linux. No subscriptions, no spyware (once we get to tower design mark4 or 5 and we have some raspberry pis running the tooling), and we want this to be a model so other people can do the same thing we’re doing!
Basically, as we approach this AI inflection point, I really strongly feel like the best results for the future will be to leverage this technology for good. Also, I think there’s a niche in using tech for extremely practical and non-SaaS purposes. Everyone and their brother wants to create this crazy billion dollar startup idea where everyone pays them a subscription to their AI bot or whatever, we’re just hoping we can make people a little more resilient in our community and maybe feed my neighbors. We think we can leverage these new tools for good.
I’m not a very experienced web dev (my day job is in automation), but I’m learning and actually it looks like I might have some side business where I throw together some small websites for local companies too. The goal is modest, practical, and ideally something that makes my community better.
Come check out my terrible flask website here:
www.pragman.io
In a couple weeks I should have my garage cleaned out and the online store ready, but for now enjoy!
Now you paste in the job description, we use an llm to generate a new resume grounded in the details you added earlier. You get a pdf that you can submit to the job application.
Been on this the last few months. Have worked on a few other ideas before that but those didn't fly.
I found that boilerplate code needed for handling defaults, environment variables, and CLI variables could become unreasonably large and error prone. I just wanted to have a struct hold the settings needed for the project, with sane defaults, helpful messages, and handling of environment and CLI variables at the same time.
So I created Settingo.
Settingo is a unified solution to handle defaults, environment variables, and CLI arguments. Settings are a boring aspect of a project, and Settingo will allow a dev to focus on the project. https://github.com/Attumm/settingo
But there definitely are good craftsmen out there, and generally the education they receive today is much higher than one or two generations ago. But, unfortunately, it's true that "you get what you pay for", which makes high quality work often pretty much unaffordable for many. That's the big reason I'm getting myself dirty at our house every day.
There are some jobs that I can do at least as good as hired labor, if not better, but there are other jobs where a qualified worker will give me a better result in the end. You gotta pick your battles.
Includes some nice-to-haves like payload encryption, carrier image creation
I am thinking of starting a venture that will increase the amount of nonesense in the world. The intention is to get people playing, writing and exploring more. I would keep track of "wins" by recording the work that participants/readers create in the world. Very grindset. Much hustle. If you need a push - or a cheerleader - on a creative project that you're working on, then reach out to me.
Users complained the app was too annoying, so I've revamped it after asking for feedback properly [1]. Learning more about user research, UX, marketing (even making TikToks!) and so on has been pretty neat. Learnings are:
1. People want to avoid losing track of time on social media
2. Instagram is great to connect with friends, but Reels are dangerous!
3. Complete focus is only necessary in cases like work or study.
This has been one of the most creative projects I've done. It's the closest thing I've done to art as a programmer :)
Also, Kotlin Multiplatform has been a joy! Only faced minor issues because it's too new, but it's been easy to create animations and reuse components. XML on Android was a huuuge pain. I'm happy with myself for prototyping without wasting too much time overall.
Slight typo: search for "NodeSytem" on your home page.
What do the police do with that register? That sounds like it’s just helpful for them. If it’s only a registration and not an approval process?
I don’t think you are responsible for data protection laws… like if it’s a ring, that’s amazons problem.
I think this is controversial, because how’s it different from me taking a happy snap on my phone and accidently capturing you in the background.
The single scary thing is the roof. I fall, I die (best case). So I will hire professionals for it.
Edit: some numbers. A guy plastered whole house inside and outside. Took 18000€ in cash and wrote a bill for another 18000€. That’s 5-6 months engineer’s salary. Plastering took 3 weeks. I had many offers for week’s task asking my whole monthly salary in cash. This kind of money attracts not only artisans, but also all possible scammers.
Right now I'm working on getting the data more complete, adding more sources and improving the learn (https://computeprices.com/learn) section.
Any feedback appreciated! Here's hoping the tools and info help people learn more about training models, deep learning research and generative AI.
There are a few others in this space, often using Python as their base.
Maybe add a map to show nearby users (that accept to chat under certain criteria)?
I'm also a student at WSU and will be pitching at our business plan competition. https://carmapay.com
I asked Gemini 2.0 to describe it and it is what was produced.
Discover a simple yet powerful way to organize and share your cherished memories with m.emori.es. Designed with ease of use in mind, even for those less tech-savvy, this private photo and video storage platform lets you create logical folders and effortlessly invite family and friends to view them. Perfect for sharing precious moments with elderly parents or loved ones, m.emori.es offers a secure and intuitive experience. With plans to integrate contact management, private chat, and genealogy tree support via GEDCOM files, m.emori.es is evolving into a comprehensive hub for family connection and memory preservation. Visit https://m.emori.es today and experience the joy of sharing made simple.
Ignore the pricing page and if you want a test account in exchange for honest feedback, please contact email me at alain AT aoware DOT co DOT uk
The idea came from an interesting "tail wagging the dog" situation. I had an old domain (faxbeep.com) lying around unused which I'd been renewing for 20 years. When I finally built something there, I discovered users were more interested in a fax testing feature that I'd added as an afterthought - the abitlity to send a fax to our test numbers and see it appear on the website for 30 days.
That insight led me to dedicate faxbeep.com entirely to fax testing, and to buy payperfax.com for my original idea: straightforward, pay-as-you-go faxing. People still need to send faxes surprisingly often - for tax authorities, immigration paperwork, medical prescriptions, etc. The service charges a fixed fee for a 3-page fax (cover + 2 pages) with a bit more for additional pages, and you only pay if the fax sends successfully.
My current challenge is visibility - Google has essentially black-holed the site. Even exact match searches like [pay per fax] don't bring up the site. If anyone has SEO advice on how to climb out of this hole - I'd appreciate it!
Here's how it works:
- Define your GPT
- Choose a base LLM (OpenAI, Claude, DeepSeek, etc.)
- Upload knowledge base
- Set pricing and start making money
I’d love to hear what the HN community thinks!
I was inspired by skidl (https://devbisme.github.io/skidl/)'s python usage and did explore that direction. However, it felt strange and clunky to try and fit the circuit drawing concepts I had in mind into python via a domain specific language (DSL). The user would require a decent understanding of python and the DSL to create circuits. This would complicate my original aim of creating a language/tool that allows electronics engineers to focus on the circuits that they would like to design.
By creating a new language, I was not limited by python's syntax/grammar. I could craft the language closer to what was needed for schematic design and maintain simplicity. Overall, it was a great learning experience on language design and also using ANTLR for the implementation.
I think you should check out the Light Table kickstarter[0] which originally had a similar premise to yours, and raised several hundred thousand dollars. I personally put in $50 to be in the beta, and that was almost 15 years ago.
I think you would get a lot of takers if you could make a convincing demo.
[0] https://www.kickstarter.com/projects/ibdknox/light-table
haystack editor is a neat canvas based IDE
codesee recently got acquired by GitKraken (wish they'd sell individual licenses for function maps rather than only for enterprise)
Sometimes I wish I could code within an obsidan.MD canvas
I think a lot of developers generally want to code on a useful node graph IDE ---- the Gource time travel idea is interesting!
Would love to see your idea pulled off! I'm rooting for you
From what I’ve noticed from my friends we all tend to have different addictions. For some it’s YouTube for others Instagram, etc…
https://imgur.com/a/miniature-weather-station-kkw7RLI
The data path will be the same as my DIY full-size weather station; ESP32(MQTT)-->WiFi-->Rpi-->Node Red-->InfluxDB-->Grafana
I’m In the US male hair market as a user these days and i think you could add value by increasing LTV of a customer for the stylists. I always wonder by they don’t offer a discount if you book your next appointment during your visit. Or offer a discount for annual prepayments, etc. The goal would be to have their customer 1) repeat and 2) repeat often. You could move the needle with some good incentives and marketing strategies for the stylist. If their walkins converted to repeats more often and their regulars had better “adherence” coming every 18 days on average instead of 20 days on average then you have a strong value prop for any stylist should use your tool and pay you for it!!
I’d definitely use an app that gave the kid recommendations and some discover-ablilty that the YT algorithm provides (probably with some more pre-filtering) except just as tiles/gifs previews and they had to “request access” to any video to play. Or have some redefined categories as you mentioned. Then once a video was allowed, it would appear as available in the app via yt-dlp or similar.
This does 2 main things I prefer as a parent. I get control over the content. It removes the instant feedback /random access loop of on-demand streaming. For some things it might even be good to control how many times they can watch it , or if they have to ask permission again, etc. Not sure if you could bolt on / use as a base but this kind of customizations for stacher would be perfect.
I generally don’t like AI, but I could see having a chat prompt “making changes to a car file” that I’ve described in my natural language could be a killer feature. Especially if you let me describe components like layers if a wall. Then use said wall in an area. Basically BIM level of detail.
It works by connecting to apis of services like github, gitlab, jira, trello and more with user's account and provides widgets for those services with various configurations.
Our most recent feature has been a widget that allows configuring a table that can connect to whatever api the user provides, all handled in the browser so only user can see the data. You can connect to any api endpoint you like and customize the table to your needs.
It was a side project that we needed for our own projects but came to realize there was a huge demand for it when we showcased it on ProductHunt.
Check it out and let me know what you think.
It was frustrating to apply for jobs, with no rejection emails and no updates, not knowing whether to keep waiting or move on. The plan is to answer questions like:
1. Has anyone heard back from this role?
2. How long does it usually take to hear back?
3. What are the online assessments/interview rounds like for company X?
I tried to understand the code and how it's used among the nodes. The thoughts in my head right now are similar to "wow this is so cool but I have no idea what it does".
I've been experimenting with DNS a lot lately, and e.g. used DNS exfil techniques to build a mesh router (mostly for privacy related reasons). [1]
But this kind of protocol layer inside PTR/TXT/URI(?) records could be really useful, as it could bypass firewalls which is super important while remote debugging a potentially compromised node. With DNSSEC or OPT record support, you could probably also reflect an authentication workflow, so that the following sessions could be handled in an encrypted-per-session manner.
I'll discuss some more with local salons who use or have used competitors instead i think. Perhaps it's dumb but I'm a bit skittish about being obvious when i'm looking at features of a competitor worth hundreds of millions. (I think a stake in treatwell went for 180 million a good while ago.)
> I always wonder by they don’t offer a discount if you book your next appointment during your visit.
This as a pre-configured option in the scheduling UI on the business end of things is something i haven't added yet. I'm guessing this is not an option glossgenius?
Ps: How can I hit you up once i have an easily shared demo?
Lots of fun to think about the future of crypto if you're able to handle unstructured data (and I'm not really a crypto person.)
Feel free to check us out/hit us up if you're curious: hello@ambient.xyz
Stella – Inspired by Jack Quaid being interviewed at the Emmy's (https://www.tiktok.com/t/ZT2a1mcfT/). I created Stella (https://usestella.app/) an app that acts as an assistant that helps actors run lines! It is at the very early stages now, but the goal is to have AI do the voices, just in case you aren't as talented as our friend Jack over there! Also the famous little women scene is on their to try out, cause I think running lines is fun just do to when bored.
Doctours – Similarly inspired by TikTok, I started something to help people getting hair transplants overseas. It was crazy for me to think people are still paying in cash for these procedures. Like idk about you but I am not bringing $5k across the border to any country. So I mainly created this as a way to solve the cross border payments problem (https://www.doctours.health/).
edit: added the link to the tiktok.
A realtime webOS interacting with it is also nearly ready to be released.
It's probably simpler than you think it is. ;-)
1: there's something producing diagnostic information which needs to be monitored
2: identify and construct evaluatives
3: put evaluatives in redis (e.g. rkvdns_examples/totalizers, but you can also write directly, see shodohflo/pcap_agent)
4: put RKVDNS in front of that (this is where the control plane ends)
5: put e.g. BIND in front of RKVDNS (BIND sits at the segment edge)
6: lie to the world that the BIND instance is authoritative (optionally hack BIND to always return AA, but never mind, nothing cares)
7: use the full set of features provided by BIND for access control, integrity, privacy
DNSSEC is problematic for something like RKVDNS (as are zone transfers), doing negative proofs in particular. DNSSEC does not provide encryption; it provides proof of authenticity, and nonrepudiation / integrity. DNS can be used to publish public keys (DANE), and DNSSEC provides proof that the key belongs to someone / something.
I need to say an important thing here: DNS is a wire format, an algorithm, and an application layer protocol. That last one in particular seems to be something people want to wish away. Every (recursing) DNS server is like an HTTP proxy: what happens to proxy headers when you chain HTTP proxies?
Encryption is per-hop, each DNS server has access to the entire DNS question, and assembles its own response. DNSSEC provides identity across the entirety of the DNS; there is no encryption / privacy across the entirety of the DNS.
On a per-hop basis, encryption is provided by DoH and DoT. DoT is as simple as putting nginx in front of the nameserver (config templates used to be included with the BIND distro, I think Ondrej was responsible for that). You're more likely to have success implementing (key based) access management with DoT; if you don't need encryption, you can get access management and non-repudiation with TSIG (from BIND, RKVDNS doesn't support TSIG), on a per-hop basis.
The most common santa claus and easter bunny question I get is "why didn't my super special EDNS data make it to my authoritative server?": people truly wanna believe. That's because EDNS is per-hop; I don't care what Cisco or Microsoft say, ain't gonna happen. (That's another reason why you need true DNS at the segment edge, not forwarding.)
("My favorite programming language is the one with the least maintenance burden. If you find it, please let me know.": In my somewhat grizzled experience lack of dependency management is the biggest burden. I try to minimize dependencies and live off the land as much as possible.)
I'd love to see a demo. We should demo for each other. I will stalk you and attempt to send you an email. (Oh I see. I'll open an issue on GitHub then.)