Re: [PEAR-DEV] Hosting PEAR's git repositories on git.php.net
[email protected] (Christian Weiske)
| Newsgroups | php.pear.dev |
|---|---|
| Message-ID | <20120905102021.65c02ae5@cyberdyne> |
Hello Daniel, > Anyone else used crucible? > > I have never used it myself, and from a quick exploration I don't see > a "pull request/merge this patch" equivalent. If it had that, or a > similar flow. Gitorious has pull requests. > It's emphatically *not* open source itself, but it may be worth > considering as a way to mitigate vendor lock-in concerns re Github, > and the "on demand" side of it is just hosting. I don't want to run from one closed software to the next but like to have something self-hosted. This does not seem to be possible at the moment, except we run our own Gitorious instance. This doesn't give us the many-people-are-there effect of Github. Maybe we need to do something else: - Let people decide where they host their code, since it's the PEAR channel server that is the final karma control. - We provide code mirroring/backup preferably on git.php.net - We also mirror the code on github so people can easily fork the repositories and send pull requests. If tomorrow some other cool tool comes along, we can switch to it. - Somehow we need to get two-way mirroring working, which means that you can push either to git.php.net and get the code mirrored to github, or push to github and get mirrored to git.php.net. This should be possible as long as only one repository is used as main integration repo. - On the pear package pages, we offer links to the github repository as well as to the backup on git.php.net. - To not lose discussions in the github web interface, we need to pipe pull request discussion mails to the pear-dev mailing list. > My former workplace is literally across the road from Atlassian's > Sydney offices, I'm wondering if I could send a friend down there to > ask "what's qualified, and can we get PEAR setup?" I wouldn't like that. -- Regards/Mit freundlichen Grüßen Christian Weiske -= Geeking around in the name of science since 1982 =-