#41

I have repos on all three of the main public repo services - GitHub, BitBucket and GitLab. For my workflow, they’re identical and I mostly use BitBucket and GitLab just because they have free private repos.

The main reasons to prefer GitHub over the other two for a public repo are third party integrations and discoverability. A surprising number of third-party services only support GitHub, and if you want to be found, you’ll do better on GitHub.

1 Like
#42

Neither of those seems particularly relevant to this particular repo.

#43

All I can say is I feel very conflicted about this. I personally am here because the principles behind Free Software motivate me. This essay by Benjamin Mako Hill is even more relevant 8 years later. But I just can’t imagine trying to take even upstream projects I created off Github right now; here’s my blog on the ostree move to Github - although that predates the GNOME Gitlab move.

I think the way Pagure stores issues in git is very cool; the comment history around pull requests and issues is almost as important as git log for long term maintenance.

1 Like
#44

+1 here.

Pagure is not perfect yet, but without proper issues reported it will not evolve.

1 Like
#45

Making a decision like this on a sample of just 61 votes (at the moment, 8h remaining), with a vote this close, seems a little bit excessive.

Fedora ought to mirror Flathub on Fedora infrastructure, contributing changes upstream as needed.

#46

The population of people who will be directly impacted by a change like this (outside of political/philosophical reasons) isn’t incredibly large to begin with, so I’d say that isn’t really that small of a sample.

3 Likes
#47

Considering this thread was used more for a general open source philosophy and principles discussion than Silverblue contributions and the practicalities of it, 61 votes with 57% for GitHub (or Gitlab) is a strong outcome.

After New Year’s we will start the GitHub vs Gitlab vote and act accordingly. Thanks everyone for the discussion! Silverblue is still a small community that needs more contributions, not only from those within Fedora who find time to contribute to Silverblue as well. Telling this community to just contribute to Pagure simply does not work.

We can all work towards a Fedora-hosted Gitlab solution but until then the Silverblue community will decide on whether to go for hosted GitHub or hosted Gitlab in the next vote. Until a Fedora-hosted Gitlab version is available, we will then definitely use that new home that is decided on via the next vote to improve Silverblue and grow our userbase from there.

1 Like
#48

If the eventual goal is Fedora-hosted GitLab, why would there be any reason for interim GitHub.com hosting rather than interim GitLab.com hosting?

#49

The simple reason is that there is no guarantee that there will ever be a Fedora-hosted Gitlab and if so, it likely will not happen in the next year or two. Someone would have to make it their priority to provide that and then gather the resources within Fedora to create this, which means going into existing structures, rallying those people that really want it, and then get to the task of implementing that solution. These things take time and as far as I am aware, no-one is currently on this task nor really wants to be. Or is it already in progress?

Let’s say someone did it and it’s now early 2021. Whether we’re on GitHub or Gitlab, in that time we have grown more awareness and have a mature process as well as possibly have gotten actual resources dedicated to Silverblue, depending on growth. On GitHub, we have the CoreOS community already there and the CoreOS developers which helps with visibility and possibly fixing things faster as rpm-ostree is also there. The growth factor is also the biggest there. On Gitlab, we half-satisfy the people who are against the fact that GitHub is proprietary and it is a good platform which gives us all the UX and features we need to simplify our tasks. Personally, I really like and prefer Gitlab but from a strategic point of view, GitHub seems better to me. There is also a chance of Microsoft open sourcing the platform, weirder things have happened.

So the goal is growth and resources. The community members can decide where they would like to contribute, there is plenty of information in this thread alone.

2 Likes
#50

Hello @walters,
I read Benjamin’s essay, and it is very relevant to this issue. I understand that the decision to leave Pagure seems to already be advanced. I also looked at your blog on the ostree move, and can realize that the end means of greater audience and ease of engagement are two major factors that weigh upon any open source or free software project. In this case, it seems to me as being someone who had difficulty with Pagure and asked for assistance consequently; that I became drawn into the greater discussion around the suitability of Pagure for the role it was being used in. I wasn’t too shy of voicing my displeasure and also noting my own errors regarding Pagure, while trying to contribute to Silverblue. I don’t lay a claim to any preference other than ease of contribution, and unfortunately this is landing squarely in the wheel house of a proprietary solution it seems. One alternative I think is to stick with Pagure and help fix the issues, if that ship hasn’t already sailed. Another is to look at an alternative to GitHub, which is I take to be GitLab. If the move is happening, and there seems to be direction of product offerings that would support a case for the longevity of Silverblue, why not just go with the more open and free solution of GitLab? I mean of course as a collective decision in the Silverblue community.

2 Likes