> I know the solution. Starting this year, students will be forced to contribute to a project they use, care about or, at the very least, truly want to use in the long term. Not one they found randomly on Github.
I expect that most of the open-source projects the students care about will be in GitHub.
So perhaps a better solution was to teach them to setup their own git repo in a cheap or free uni server.
Tell the students that they will receive a one week notice during the middle of the semester that they need to migrate their git repo to a new server, then teach them the 2 or 3 commands they will need to enter to do this.
They will then understand that it is extremely easy to move a git repo.
People like this CHOSE to use github despite it being a walled garden and now are complaining that they want to destroy the "open source" ecosystem.
You see, here's Stallman being right once again. Stallman many times on many topics said he wouldn't use a product that's a walled garden and where his presence would contribute to that products dominance (i.e. network effects - everybody has to use it because everybody else is using it). People like the author said "I'll use github because it's beneficial to me in the immediate horizon, despite the fact that I'll be indirectly contributing to their assault on free software". Well, hard for me to care about the author now.
Anyway Forgejo[1,2] is FREE and COPYLEFT software, and Codeberg[3] is a pretty big forge. Forgejo also has on their roadmap to add some federated-type features, so that different people/organizations can host their Forgejo instances, but interact seemlessly with projects on other instances.
If this stuff matters to you, donate to Forgejo[4].
Federation should be the way but time and again we have seen federation fail, first for chats (XMPP) then for Blurbs (twitter vs mastodon), and social networks where there's not even a big enough name to mention for federated social networks.
Depending on what specifically you're talking about, what you say might or might not make sense.
For chat, it depends on what you mean by fail. What actually happened/is happening, is that XMPP the protocol works so well, that most chat apps start by just being another XMPP app, and if they ever get traction they make their servers incompatible with XMPP. That's what WhatsApp did for example.
For social networks, Mastodon is quite large especially in the tech space. It's just not facebook-size. But given that Facebook's success is driven by getting people addicted to their phones, I see this as a sign that Mastodon is doing better. Mastodon doesn't try and get you addicted. And the federation aspect works AMAZINGly. I don't know/even care whether a user I'm interacting with is in the same instance or not. And I can subscribe not just to publications of Mastodon users but also blogs, photo apps etc which implement the ActivityPub protocol.
So yeah, what you said is not nonsense, but I 100% disagree with "doesn't have 1bn users therefore federation failed".
If federation really works that well, then it sounds like Mastodon/ActivityPub is ripe for carrying Git repository metadata and discussion. Obviously Git development works very well with mailing lists, other than the centralized server requirement, awkwardness (for most users), and technical limitations of mailing lists. So then you just need your federated discussions to carry patches or point to publicly accessible repos and put some decent UI (of choice) on top of that.
Even issues are mainly just discussions with some metadata attached and as long as they can be surfaced in a way to be attached to a project, then they could be created by anyone.
Hmm.. Surely there is already effort being focused in this direction?
You're right about federation having failed in those instances, or at least failed to take over the world as was expected.
But Forgejo is a git + associated services application first and foremost. It's clearly on the up, and this federation is no more than a planned feature that may or may not catch on. You make it sound like Forgejo will fail because of it, but it's just an add-on.
BTW Forgejo seems to be very similar to GitHub when it comes to bug tracking. There are so many project management systems and bug trackers out there, and I think GitHub (and as thus, Forgejo's) way of doing this is limiting.
I wonder if people would rather prefer Jira, Redmine, MantisBT, Bugzilla, or something completely different, or a choice to have X and Y and why, and so forth.
Or rather "centralisation" in the hands of BiGTech is eroding many of our rights, the right to repair (through open source) being one of them. Also, App Stores too seek to kill open source apps on the mobile platform as there is no money to be made from it and good "free" open source apps tend to poach the sales from paid apps. There many open source projects for iDevices that will never be popular because you cannot install it directly.
I have always been very interested in seeing repository metadata (issues, etc) kept in the code repository itself. Not technically easy, I know. I wonder if there are any current efforts that manage to make this work?
If you want something close to the GitHub UX, then Forgejo or Gitea.
If not, there are loads of options depending on your preferences and needs.
GitLab is still relevant. Personally I find it too resource-heavy (both server and web ui) to be my first pick but it's still widely used and actively developed.
Many other tools/platforms provide decent source control and issue tracking. Nothing else has the FOSS project market share of github, and this matters especially when you're looking for the canonical home of a project and trying to judge how popular/active/viable it is (stars/commits/issues/PRs).
If you want exposure and participation for your FOSS project, it's harder to not use github.
FWIW, Forgejo does the source control stuff well. I love it for self-hosted local mirrors.
GitLab is very relevant. In some spaces, it is the only one with enough features. But it is pretty expensive and the way i see it, only the ultimate pricing plan makes sense. Premium makes no sense unless you are just a small company developing proprietary software with separate workflow for customers..
GitLab, Codeberg, SourceHut, Notabug, Radicle, Tangled… & that is just Git but there are other VCSs hosted other ways like Darcs, Pijul, Fossil, & so on.
"Please don't pick the most provocative thing in an article or post to complain about in the thread. Find something interesting to respond to instead."
> I know the solution. Starting this year, students will be forced to contribute to a project they use, care about or, at the very least, truly want to use in the long term. Not one they found randomly on Github.
I expect that most of the open-source projects the students care about will be in GitHub.
So perhaps a better solution was to teach them to setup their own git repo in a cheap or free uni server.
> So perhaps a better solution was to teach them to setup their own git repo in a cheap or free uni server.
And then lose access after graduating. Great idea!
Tell the students that they will receive a one week notice during the middle of the semester that they need to migrate their git repo to a new server, then teach them the 2 or 3 commands they will need to enter to do this.
They will then understand that it is extremely easy to move a git repo.
Probably but I never got why that is. It seems a school could be there in all kinds of ways for the entire duration of the career.
It's just an educational exercise to teach students to setup and mantain their own git server.
Obviously students don't expect to use uni servers forever. They can setup their own with the lessons learned.
Any Git clone is also technically a Git server, so no, they don't lose access to their own filesystem after graduating.
People like this CHOSE to use github despite it being a walled garden and now are complaining that they want to destroy the "open source" ecosystem.
You see, here's Stallman being right once again. Stallman many times on many topics said he wouldn't use a product that's a walled garden and where his presence would contribute to that products dominance (i.e. network effects - everybody has to use it because everybody else is using it). People like the author said "I'll use github because it's beneficial to me in the immediate horizon, despite the fact that I'll be indirectly contributing to their assault on free software". Well, hard for me to care about the author now.
Anyway Forgejo[1,2] is FREE and COPYLEFT software, and Codeberg[3] is a pretty big forge. Forgejo also has on their roadmap to add some federated-type features, so that different people/organizations can host their Forgejo instances, but interact seemlessly with projects on other instances.
If this stuff matters to you, donate to Forgejo[4].
[1] https://codeberg.org/Forgejo/forgejo
[2] https://forgejo.org/
[3] https://codeberg.org/
[4] https://liberapay.com/forgejo
Federation should be the way but time and again we have seen federation fail, first for chats (XMPP) then for Blurbs (twitter vs mastodon), and social networks where there's not even a big enough name to mention for federated social networks.
Depending on what specifically you're talking about, what you say might or might not make sense.
For chat, it depends on what you mean by fail. What actually happened/is happening, is that XMPP the protocol works so well, that most chat apps start by just being another XMPP app, and if they ever get traction they make their servers incompatible with XMPP. That's what WhatsApp did for example.
For social networks, Mastodon is quite large especially in the tech space. It's just not facebook-size. But given that Facebook's success is driven by getting people addicted to their phones, I see this as a sign that Mastodon is doing better. Mastodon doesn't try and get you addicted. And the federation aspect works AMAZINGly. I don't know/even care whether a user I'm interacting with is in the same instance or not. And I can subscribe not just to publications of Mastodon users but also blogs, photo apps etc which implement the ActivityPub protocol.
So yeah, what you said is not nonsense, but I 100% disagree with "doesn't have 1bn users therefore federation failed".
If federation really works that well, then it sounds like Mastodon/ActivityPub is ripe for carrying Git repository metadata and discussion. Obviously Git development works very well with mailing lists, other than the centralized server requirement, awkwardness (for most users), and technical limitations of mailing lists. So then you just need your federated discussions to carry patches or point to publicly accessible repos and put some decent UI (of choice) on top of that.
Even issues are mainly just discussions with some metadata attached and as long as they can be surfaced in a way to be attached to a project, then they could be created by anyone.
Hmm.. Surely there is already effort being focused in this direction?
There is, it's called ForgeFed https://nlnet.nl/project/ForgeFed/
You're right about federation having failed in those instances, or at least failed to take over the world as was expected.
But Forgejo is a git + associated services application first and foremost. It's clearly on the up, and this federation is no more than a planned feature that may or may not catch on. You make it sound like Forgejo will fail because of it, but it's just an add-on.
There are more mastodon servers than GitHub servers
I agree.
BTW Forgejo seems to be very similar to GitHub when it comes to bug tracking. There are so many project management systems and bug trackers out there, and I think GitHub (and as thus, Forgejo's) way of doing this is limiting.
There was a recent submission about it on Hacker News: https://news.ycombinator.com/item?id=46460319.
I wonder if people would rather prefer Jira, Redmine, MantisBT, Bugzilla, or something completely different, or a choice to have X and Y and why, and so forth.
Tangled.org looks really interesting as well
Or rather "centralisation" in the hands of BiGTech is eroding many of our rights, the right to repair (through open source) being one of them. Also, App Stores too seek to kill open source apps on the mobile platform as there is no money to be made from it and good "free" open source apps tend to poach the sales from paid apps. There many open source projects for iDevices that will never be popular because you cannot install it directly.
I have always been very interested in seeing repository metadata (issues, etc) kept in the code repository itself. Not technically easy, I know. I wonder if there are any current efforts that manage to make this work?
What are free, good alternatives to github?
Is gitlab still relevant?
If you want something close to the GitHub UX, then Forgejo or Gitea.
If not, there are loads of options depending on your preferences and needs.
GitLab is still relevant. Personally I find it too resource-heavy (both server and web ui) to be my first pick but it's still widely used and actively developed.
For what part(s) of github?
Free as in freedom?
Many other tools/platforms provide decent source control and issue tracking. Nothing else has the FOSS project market share of github, and this matters especially when you're looking for the canonical home of a project and trying to judge how popular/active/viable it is (stars/commits/issues/PRs).
If you want exposure and participation for your FOSS project, it's harder to not use github.
FWIW, Forgejo does the source control stuff well. I love it for self-hosted local mirrors.
For small teams, fossil is unbeatable
https://fossil-scm.org
GitLab is very relevant. In some spaces, it is the only one with enough features. But it is pretty expensive and the way i see it, only the ultimate pricing plan makes sense. Premium makes no sense unless you are just a small company developing proprietary software with separate workflow for customers..
GitLab, Codeberg, SourceHut, Notabug, Radicle, Tangled… & that is just Git but there are other VCSs hosted other ways like Darcs, Pijul, Fossil, & so on.
I use codeberg
[flagged]
"Please don't pick the most provocative thing in an article or post to complain about in the thread. Find something interesting to respond to instead."
"Eschew flamebait. Avoid generic tangents."
https://news.ycombinator.com/newsguidelines.html
[flagged]