Ask HN: What Are You Working On? (June 2025)

I'm working on a new app for creating technical diagrams - https://vexlio.com. It's an area with some heavyweight incumbents (e.g. Visio, Lucid) but I think there's good opportunity here to differentiate in simplicity and overall experience. I'm still in the fairly early phase, and I suspect I haven't quite found the best match of features to customers yet.

From a dev perspective this area has a ton of super interesting algorithmic / math / data structure applications, and computational geometry has always been special to me. It's a lot of fun to work on.

If anyone here is interested in this as a user, I'd love for any feedback or comments, here or you can email me directly: tyler@vexlio.com.

Some pages the HN crowd might be interested in:

* https://vexlio.com/blog/making-diagrams-with-syntax-highligh... * https://vexlio.com/solutions/state-diagram-maker/ * https://vexlio.com/blog/speed-up-your-overleaf-workflow-fast...

really nice work. I'm going to give it a roll!

Thank you! If you end up having any feedback, definitely feel free to drop it here, or email if you prefer.

seamless latex integration is a winner for me!! will definitely spread the words for this

Looks great, and smart differentiation!

Actually right up my alley. I have many frustrations and reservations against the current offerings. Super excited to see a new player enter the field

In the off chance you haven't seen Bret Victor, your app reminds me of him, https://www.youtube.com/watch?v=NGYGl_xxfXA

Visio and Lucid are trying to cover everything at the expense of practical convenience. Pick a lane and stick to it. Good luck!!!

It looks like a pretty interesting product so I really hate to be that guy but the FAQ page at https://vexlio.com/faq/ straight up doesn't work (whenever I click any of the questions, it does nothing). Also, wanted to know if there was anything in the pipeline to get a Desktop application which would work offline. In several places in the enterprise world especially, I do feel there would be scope for that. I would definitely pay for a desktop version which worked offline for example.

Whoops - FAQ issue should be fixed if you refresh (if it's still broken, give it some time for caches to be invalidated). Thanks for mentioning that!

Re: desktop version. The short answer is yes, probably, but I don't have a concrete timeline. I made tech and architecture choices from the beginning to make sure a cross-platform desktop version always remains possible. Frankly, the biggest obstacle for desktop is not the app itself, but distribution and figuring out a pricing model. The current solution for enterprise, business, and other interested people, is to self-host Vexlio, with separate licensing.

oh cool! I want to try this soon.

FAQ works fine for me now.

Cheers, thank you!

Definitely seems to be the case from my observation as well. Appreciate it!

Awesome, thank you! If you or your colleagues have other LaTeX-related goals or wishes, do let me know. There's a lot of untapped opportunity there as well (IMHO).

Would love to hear those frustrations and reservations - drop me a line if you're interested in sharing: tyler@vexlio.com.

This looks really cool. An application I would use this for is to generate code for FPGAs, as finite state machines are very common.

This is an example, https://terostechnology.github.io/terosHDLdoc/docs/guides/st...

But it only outputs an SVG, and there are no tools (AFAIK) that go from diagram to code, which should easy to setup.

So I'd consider extending this to both generate code and read in code and make these nice interactive diagrams.

Gave it a quick try and it's really nice, the aesthetic defaults are great. One thing I found unintuitive: I should be able to connect objects without having to select a new tool (the anchor points on hover should be clickable in any tool mode so I can connect objects on the fly).

Overall amazing though, will be using!

Unless you intend to be acquired by Overleaf I don't really see a future for your business to be honest.

Looks great! Your editor design is beautiful.

Looks really good an seems intuitive (from just browsing the landing page). Will look more deeply.

Diagram-as-code option?

ie. a language syntax from which a diagram can be generated?

I find a lot of the time taken up in doing diagrams is laying them out properly and then having to rearrange them when it grows beyond a certain size.

This may,however, be an old-man Visio user problem that's been better solved by more recent options...

Thank you for the feedback! This is a great idea and definitely fits into the vision.

Do you know if the FPGA and/or hardware communities use any type of formalism for design or documentation of state machines? One example of what I mean is is Harel statecharts - essentially a formalized type of nested state diagram.

Some type of programmatic diagram creation is definitely something I'm interested in supporting. It's not clear to me how large the audience would be, so it's been hard to prioritize.

Thanks for this feedback! This is one of those quality-of-life features that I think are really important for the overall experience - I will be adding this.

What's your long term revenue model?

Enterprise licensing? Donation based? Hosting fees with value-add mark up?

Super cool. Do you consider yourself to be a competitor with Mermaid?

Thanks! I would say no. Mermaid is strongly code-first diagramming, which is an excellent usecase and niche in its own right. I would be surprised if Mermaid ended up with a WYSIWYG editor on top of it, since that is pretty counter to its philosophy (as far as I understand anyway).

Looks pretty great! The free tier also looks reasonable. The pricing on the other tiers isn't outrageous either if you use it consistently. Unfortunately, I likely find myself in the big gap between the free tier and the Basic plan. I can't justify yet another subscription that I use only a couple of times a year. That said, I would happily pay the $6 on the months that I use the service. Given the churn issues, I'm surprised more SaaS offerings don't work that way.

Thank you for sharing this perspective! Your proposal is potentially a good middle ground, and I will certainly give this some thought.

Super impressive! Your software product looks phenomenal and your website is also extremely nice! Best of luck to you.

Years ago I was making a diagram editor with the intent of doing code generation from diagrams (like simulink, not stateflow). I started with splines for the connections and decided straight lines and junctions would be better for complex diagrams. I realized that a better way to internally define the connecting wires is via a set of lines and their connectivity (vs their endpoint coordinates). Imagine each line segment is defined by a direction (vertical or horizontal) and a position (perpendicular distance from the origin) Like ax+by=d where a and b are either 0 or 1. You also need to define which other lines it connects to. Given the list of connections you can then calculate the intersections at rendering time. By sorting the list of connections you can render the line without features at the start and end, and then draw junctions for any intermediate connections. The beauty of this would be to allow dragging blocks around and having the lines follow with the junctions passing through each other as needed. There is some housekeeping with this data structure (merging colinear segments that connect, breaking segments when needed) but the UI for dragging should be much better than anything out there.

Nice project, congratulations, it would be cool to see it integrated into LaTex, what do you think?

Interesting - do you have a writeup or a demo available somewhere? What types of junctions were you envisioning?

In the links you posted (2 comments up) there is a PID controller. Out of the difference block the signal splits 3 ways. In Simulink that junction would just be a round dot where the lines meet. In most diagram editors that junction would probably have 5 lines connected to it. In my scheme there would be one horizontal line right through it and one vertical line. There would also be 2 short horizontal lines connecting the vertical line to the P and D blocks. In my scheme you could drag the D block up to the top above the P (D,P,I) and the junctions would change (because they're locations are computed) even though the topology has not. I never did a full implementation. email me at gmail if you want a better explanation or a quick drawing.