Jacob Helwig <[email protected]> writes: > On Thu, 28 Oct 2010 11:34:48 +1100, Daniel Pittman wrote: >> >> I am going to mention patchwork again - a bunch of Linux kernel projects use >> it to monitor patches that go past on their lists; it could easily capture, >> in >> fact, not just the deliberate dev list things, but also the user list set of >> patches.
[...] >> Otherwise I am gonna advocate Gerit, which every single project I know that >> has adopted has within a month had a rash of "wow, this is like the SVN to >> GIT >> change for us!" comments. >> >> Gerit is a much bigger change of process, though, than patchwork. > > I've brought up Patchwork internally a couple of times, and a couple of us > have worked on getting it up, and running, with very limited success. > Unfortunately, the documentation, and finicky version dependencies meant > that we never actually got it fully setup. I'd love to have a complete set > of directions on getting it setup (or something with the same features). > I think it's pretty much exactly what we need to help prevent community > submissions from falling through the cracks so easily. Well, as mentioned I didn't have too much trouble running it up, and it seemed to work. Perhaps it loves Debian derivatives more than other platforms or something. :) > As far as Gerrit...well...I rather dislike it, actually. I think that it > has a rather awkward work-flow. Especially if you're not all that familiar > with Git (and even if you are). I'd rather not raise the bar for submitting > patches quite that high. Mmmmm. I don't disagree, but it is one of those tools that enough smart people swear by that I would give it serious thought. It definitely does provide the review and management facilities required for the job... > Personally, I like an emailed patches based work-flow. I also realize that > it isn't quite working for us at the moment. I also realize that I have a > very different email setup from the reset of the development team, and that > it's mainly geared around this type of work-flow. *nod* I suspect that we both are in that position; the majority of patch reviews seem to cite the diff mangled beyond recognition, which wouldn't help in the least, while Emacs/Gnus handles it just fine.[1] > No matter what "internal" development does, I'd still love to see something > like patchwork to help keep track of community submissions. Anyone feel > like helping out with getting some step-by-step directions on getting > patchwork up on Ubuntu 10.04? ;-) 10.10 gave me: * install the mentioned dependencies as packages * configure the database for username and password access * write the local configuration as directed * sync the database * s...@python-support/python-dja...@pyshared@g on lib/apache2/patchwork.wsgi.conf * adjust paths to match my deployed location in the same file * include that apache conf into my Apache... So, it didn't seem like a huge load of work. Not exactly fun, but still much less than I would have expected from the description. Markus <[email protected]> writes: >> Gerit is a much bigger change of process, though, than patchwork. > > :) You say that like it's a bad thing. My biggest concern would be that it has the quirky 'git push' approach for submitting things for review. If github offered better support for review-before-pull I would probably lean to that, especially because github is such a big thing in the FOSS community these days. (...or I could just be ignorant of some of the things it does to support this and/or tools like github-gem that make it easier to manage...) Daniel Footnotes: [1] Also syntax highlighting for diffs, if asked, or they attached. -- ✣ Daniel Pittman ✉ [email protected] ☎ +61 401 155 707 ♽ made with 100 percent post-consumer electrons -- You received this message because you are subscribed to the Google Groups "Puppet Developers" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/puppet-dev?hl=en.
