+ Headline includes the inflammatory term "addiction" without support.
+ Inflammatory term "monopoly" is immediately followed, apparently without irony, by a paragraph exploring the many alternatives.
+ Article gives no factual support why alternatives are better, except for a condescending throwaway statement that "ten years of suboptimal habits" have prevented readers from knowing any better.
+ Author rants about implementation of emoji typing.
+ Article claims that GitHub leads to bad coding by discouraging people from thinking. No factual support.
+ Article shifts suddenly to an apparent moral obligation not to use GitHub. The best justification I found is the claim that by using GitHub, you encourage others to use GitHub, because it's a popular site and you make it more popular. Unclear where the morality lies in that argument.
I think the author doesn't like GitHub, but it's possible to make the point better. Don't insult your audience, don't assume your reader will take your main premise as a given, don't lean on rhetorical terms like "addiction" and "monopoly," especially without support, and do provide a common ground that acknowledges that others can see the same facts as you but reach different conclusions.
Some real “old man yells at cloud” vibes from this post.
The Discover feed has been helpful for me to discover libraries I can use or projects to contribute to. I approve of it much more than the profile badges (which seem abandoned?).
To the author: believe it or not but open source is a community. People need help and others want to offer help, just because a new discover feature isn’t for you doesn’t make the company the ruin of all software programming.
> Developers are now raw meat encouraged to get stars, followers and commit counters, doing the most stupid things in the most appealing way to get visibility! Yeah! Engagement! Followers! Audience!
I'm not sure what's wrong here.
Pre-GitHub, most OSS happened via mailing lists (and for some core projects, it still does). The idea was to make a more polished, perhaps more social UI to browse code and repo (that wouldn't get overrun with spam and shady download like SourceForge did...).
And you don't have to use any of those features if you don't want to! A lot of devs only use the repo and PR UI and don't care or even notice the follows, start or issue tracker.
> Good code is written when people are focused, thinking hard about a problem while having the time to grasp the big picture. Modern Github seems to be purposely built as a tool to avoid people thinking about what they do and discourage them from writing anything but a new JavaScript framework.
That's such a bizarre take. How is GitHub incentivizing people to only write JS Frameworks? You can write pretty much anything you want on GitHub...
> I've migrated all of my own projects to Sourcehut (where I've a paid account) or to my university self-hosted gitlab.
It's great when you have a datacenter and full time sysadmins to maintain a GitLab install. That's precisely why people (who didn't have those resources) went with GitHub.
Plus, let's be honest, compensation from GitLab has never been competitive with Github, and they tried normalizing some weird CoL adjustments that made little sense. We all know where devs who got offers from both companies ended up working...
GitHub is like any other commercial social network. It appoints trophies and rewards to those who play the 'game' in some meaningful way. GitHub doesn't even do rate limiting on follows and actually encourages bot behavior for anyone wanting to artificially inflate their follower count (it's trivial to do this). This has lead me to believe most accounts are fluff accounts who chase metrics instead of meaningful contributions.
perhaps anecdotal but gh's newer features have actually noticeably increased my productivity. i don't particularly enjoy slack, so i have all but mentioned notifs off - i maintain several of our open source libraries and we've created a project on gh to track issues across all of these repos:
- the kanban board is alright, but i just like having everything in one place. ideally, i'll never have to log into atlassian anything, ever again.
- having the mobile app means i can also turn my desktop notifications off and use the 'focus' mode (with github being able to notify me).
- discussions have recently been revisited and they made a lot of improvements. the next team from vercel is extremely responsive on there and i think many open source projects could move from mailing lists to something like that. it's not for everyone, but worth considering.
- wiki's are relatively useless, in my experience.
aside from my own experience, i also support many open source wizards through github sponsors. it's pretty much plug and play, so there's that too.
politics aside, there is a lot wrong with gh - but i'm also strongly convinced that post-acquisition, things could have been much, much worse. i'll continue to support projects and maintainers based on the merit of their projects. whether they choose to use sourcehut, gitlab, or github, i think it's safe to say that creating software and distributing it has never been easier.
I just tried Codespaces last night, and I think after making two myself, this could lower the barrier for users to try out many projects, if they add a single devcontainer.json to their repo.
although it's annoying to have to need to do this, sounds like a good idea for a quick little side project. compile a list of your client-side grievances and make an extension?
Initially I thought weary was the right term (more of a sad, tired dejectedness that the : character was overloaded), but I guess both words work equally well, since now I'm also a bit cautious hitting that key.
I thought the article would be talking about why you don't really need Github... Instead it was saying that discovery (meaning finding developers/projects) was an anti-pattern. GitHub (GitLab and others, too) do have a strong value prop.
What does it say about us collectively that we've chosen a proprietary closed-source platform to develop the majority of the worlds Open Source Software?
It seems a bit strange, like going to use a gym inside a fast food restaurant, or perhaps seeking spiritual fulfillment through associating with a for-profit religion. Yes, these are both terrible analogies. I invite you to offer better.
Using the platform which is best at doing what developers need while also providing a network effect is best for open-source. Why does it matter that the platform is open/closed source?
I think a better analogy is that you could cook an equally healthy meal in the kitchens of fast-food, high-end restaurant, or at home. The tools are basically the same, it's what the individual puts in and gets out that we care about.
Just last night, with just a few lines of JSON, I made it a single click to try out my project on GitHub Codespaces. This is great UX & DX that enables open-source to flourish.
+ Headline includes the inflammatory term "addiction" without support.
+ Inflammatory term "monopoly" is immediately followed, apparently without irony, by a paragraph exploring the many alternatives.
+ Article gives no factual support why alternatives are better, except for a condescending throwaway statement that "ten years of suboptimal habits" have prevented readers from knowing any better.
+ Author rants about implementation of emoji typing.
+ Article claims that GitHub leads to bad coding by discouraging people from thinking. No factual support.
+ Article shifts suddenly to an apparent moral obligation not to use GitHub. The best justification I found is the claim that by using GitHub, you encourage others to use GitHub, because it's a popular site and you make it more popular. Unclear where the morality lies in that argument.
I think the author doesn't like GitHub, but it's possible to make the point better. Don't insult your audience, don't assume your reader will take your main premise as a given, don't lean on rhetorical terms like "addiction" and "monopoly," especially without support, and do provide a common ground that acknowledges that others can see the same facts as you but reach different conclusions.