Most active commenters
  • matheusmoreira(4)
  • Yokohiii(3)

←back to thread

311 points lukeio | 22 comments | | HN request time: 0.789s | source | bottom
1. pedrozieg ◴[] No.46233265[source]
There’s something refreshing about explicitly saying “this editor exists to delight me, and that’s enough”. The default script now is that every side project should either be open-sourced or turned into a SaaS, even if that pressure is exactly what kills the weirdness that made it interesting in the first place.

Some of the best tools I’ve used felt like they started as someone’s private playground that only later got hardened into “serious” software. Letting yourself park Boo, go build a language, and come back when it’s fun again is probably how we get more Rio/Boo-style experiments instead of yet another VS Code skin with a growth deck attached.

replies(3): >>46234633 #>>46236932 #>>46237537 #
2. mghackerlady ◴[] No.46234633[source]
I'm very much for people open-sourcing their projects in terms of releasing the source code. Just don't accept patches or whatever, keep the repos closed
replies(1): >>46234840 #
3. mirashii ◴[] No.46234840[source]
Unfortunately, and I think to great overall harm, GitHub does not let you disable many of the collaboration features. I was just having a discussion today with someone who would be fine open sourcing their code, but is uninterested in any contributions, questions, or community interaction. Since GitHub won’t allow that, their options are to host it somewhere themselves where nobody will see it, or just don’t publish it, which is ultimately what happened.
replies(7): >>46235043 #>>46235315 #>>46235989 #>>46236376 #>>46236456 #>>46236688 #>>46241833 #
4. matheusmoreira ◴[] No.46235043{3}[source]
> GitHub does not let you disable many of the collaboration

I wish they'd allow making issues and pull requests sponsor only. Could enable a business model.

replies(1): >>46236035 #
5. munificent ◴[] No.46235315{3}[source]
I have a hobby game up on GitHub. The README explains that it's open source for people to fork it and file issues, but that I don't accept contributions. So far, it seems like that's been very effective.

We don't always have to solve problems with technology. Sometimes you can just tell people things.

6. hnlmorg ◴[] No.46235989{3}[source]
Odds are, you’re not going to get any contributions even if you do want them. So they could just upload regardless.

And if the README explicitly says the project isn’t open to contributors nor feature requests, then you’re even less likely to see that (and have a very valid reason to politely close any issues on the unlikely scenario that someone might create one).

The vast majority of stuff on GitHub goes unnoticed by the vast majority of people. And only a very small minority of people ever interact with the few projects they do pull from GH.

replies(1): >>46238311 #
7. Yokohiii ◴[] No.46236035{4}[source]
It's weird that this thread argues to keep the fun in hobby projects and you ask for the exact opposite.
replies(1): >>46236557 #
8. mghackerlady ◴[] No.46236376{3}[source]
The obvious solution is to just not use github but that's probably not super easy for people without the resources to just throw a tarball on a server somewhere and link people to it
9. Lammy ◴[] No.46236456{3}[source]
I use an Action to auto-close any Issue or PR in my hobby repo for same reason: https://github.com/marketplace/actions/repo-lockdown
replies(1): >>46236716 #
10. matheusmoreira ◴[] No.46236557{5}[source]
It's precisely because of the hobby nature of my projects that I want this feature. Support and collaboration are a lot of work. I have trouble conjuring up enough motivation to work on my projects as it is.
replies(2): >>46236639 #>>46237459 #
11. Yokohiii ◴[] No.46236639{6}[source]
Sponsors can have quite a bit more entitlement then the average github dude. But well, maybe if you lock it down for sponsors the stress level is overall lower.
replies(1): >>46236744 #
12. zzo38computer ◴[] No.46236688{3}[source]
I hardly get any contributions, questions, etc even though I have published them on GitHub (although some people do watch and/or star them, but I don't really care much how many stars it has).

I think you can disable issues but not pull requests, as far as I know.

It might be helpful to allow to disable pull requests too, and possibly to hide how many stars/watchers there are and hide the list of forks (people could still star, watch, and/or fork the repository, but they would not be listed on that repository if the display of those features are disabled).

Whether or not GitHub accepts these ideas, it can be an idea that other services (e.g. Codeberg) can consider adding such options if they want to do (as well as other things).

13. zzo38computer ◴[] No.46236716{4}[source]
I use GitHub Actions to affect issues and pull requests also, but to assign them to myself (so that they are visible in searches), not to close them. However, for some reason it does not seem to work properly for pull requests, even though it works for issues.
14. matheusmoreira ◴[] No.46236744{7}[source]
> Sponsors can have quite a bit more entitlement then the average github dude.

Is this some sort of unwritten agreement? When I was setting up my sponsor page, I explored the sponsor pages of other users for ideas. I don't think there were many sponsorship tiers with special features. Some people offered advertising space on the README, others offered access to an exclusive Discord channel, most just thanked the sponsor.

I'm still new at this so I wouldn't know. I only ever had one sponsor. Happened organically after my work was independently posted here on HN once.

replies(1): >>46238122 #
15. BeetleB ◴[] No.46236932[source]
> There’s something refreshing about explicitly saying “this editor exists to delight me, and that’s enough”.

(Emacs)

16. tacone ◴[] No.46237459{6}[source]
I found working with AI as the code buddy to be motivating (ironically). You get to chat about the project, ask opinions and in general have somebody do the work you don't find inspiring.

AI often doesn't do things your way, but if your doing something for yourself you usually care more about the goal than the technicalities. Also AI working on a hobby code base is less prone to overcomplication since it basically copies what you've wrote yourself.

replies(1): >>46237820 #
17. hgs3 ◴[] No.46237537[source]
> The default script now is that every side project should either be open-sourced or turned into a SaaS

I think its worse then that. It seems the narrative is everything needs to be enterprise-scale by default. Those who value small languages and tools, experimentation, self-hosting, and the do-it-yourself mindset are the counterculture.

18. matheusmoreira ◴[] No.46237820{7}[source]
I had a similar experience. Just chatting about stuff, shooting ideas and concepts back and forth with the AI is quite stimulating. I get to be an obnoxious help vampire without draining other humans of their patience and motivation. It's like having a developer friend with infinite patience to chat with.

In terms of productivity it's having something of a mixed effect. It gives me very clear ideas and direction but at the same time everything just feels done afterwards. All that's left is actually executing the tasks which is... Boring.

I'm not sure I trust ChatGPT to do it for me like an agent. The examples it gives me are never quite right. It's probably a lot better at generating frontend javascript code than programming language interpreter code.

19. Yokohiii ◴[] No.46238122{8}[source]
Oh my mistake I was thinking of individual donations, which may be implied as some premium service. I think a company/org sponsor should be more professional. In theory you can just cancel a sponsor if it doesn't fit. You can turn your back on on donations, but you somehow owe the donators forever.

Edit: https://pocketbase.io/faq/

Look at the bottom, just an example how sponsors/donors may affect you.

20. mirashii ◴[] No.46238311{4}[source]
> Odds are, you’re not going to get any contributions even if you do want them. So they could just upload regardless.

This is not my personal experience nor the experience of a number of folks that I know personally. I think it's pretty hard to generalize about this.

> The vast majority of stuff on GitHub goes unnoticed by the vast majority of people. And only a very small minority of people ever interact with the few projects they do pull from GH.

So what? It's probably not going to impact you, so it's okay and we just have to deal with it? I reject that logic entirely.

replies(1): >>46241799 #
21. hnlmorg ◴[] No.46241799{5}[source]
> This is not my personal experience nor the experience of a number of folks that I know personally. I think it's pretty hard to generalize about this.

I think it’s pretty easy to generalise because public repositories are public, so the data is available.

The vast majority of repositories on GH has between 0 and 10 stars and no issues raised by other people.

Even people (like myself) who have repos with thousands of stars and other GH members “following” them, will have other repos with in GH with zero interaction.

> So what? It's probably not going to impact you, so it's okay and we just have to deal with it? I reject that logic entirely.

That’s a really uncharitable interpretation of my comment.

A more charitable way of reading it would be:

“Worrying about a minor problem that is easily remediated and likely wouldn’t happen anyway isn’t a strong reason to miss out.”

If we were talking about something high stakes, where one’s career, family or life would be affected; then I’d understand. But the worst outcome here is an assumption gets proven true and they delete the repo.

Please don’t take this as a persuasive argument that someone should do something they don’t want to do. If people don’t want to share their code then that’s their choice.

Instead this is responding to the comment that your friend DID want to share but was scared of a theoretical but low risk and unlikely scenario. That nervousness isn’t irrational, but it’s also not a good reason by itself to miss out on doing something you said they did want to do.

If however, that was really just an excuse and they actually had no real desire to share their code, then they should just be honest and say that. There’s no obligation that people need to open source their pet projects so they don’t need to justify it with arguments about GHs lack controls. They can just said “I don’t want my code public” and that’s a good enough reason itself.

22. planb ◴[] No.46241833{3}[source]
Or to simply ignore issues and PRs. You don't have to answer people for something you do as a hobby.