On 10/20/18 9:09 AM, Stephen J. Turnbull wrote: > Kevin Fenzi writes: > > On 10/19/18 6:43 AM, Neal Gompa wrote: > > > > You know why the usage numbers bear that out? Because the upgrade to > > > HyperKitty was mishandled and delayed over and over. We were screwed > > > over by the fact that our infrastructure doesn't run on Fedora, so > > > that made it harder to get it working. The initial deployment was very > > > slow and unoptimized. Bugs in the UI remained unfixed in Fedora's > > > installation even though upstream fixed them. I would not be surprised > > > if upstream ignores us because we don't seem to be upgrading. > > > > Huh, you do realize that things take as long as they take, and there's > > no magic wand for 'it's magically done'. mailman3 was a massive > > undertaking with a very small group of developers, many of whom were > > wanting things to be really done before releasing them. > > Yeah, as a core Mailman developer I was really disappointed when the > whole crew of Fedora/RH-supported developers just disappeared without > leaving behind successors. I understand why that happens, but I wish > I'd known they were able to participate only as long as they were > assigned to it. Huh. The only person I know of from Fedora at least that was working on it was abompard. While he's working on other things now, as far as I know he's still working on mailman3/hyperkitty as time permits. > Unfortunately, it's clear from the current > installation supporting this list that no, they didn't get things > "really done", or at least they were restricted to their direct > relationship with HyperKitty -- in Fedora Postorius, even the > explanatory blurbs in the user config screens are frequently > incomplete (eg, lack information about current, default, and inherited > values) and at least one is outright confusing (the semantics of the > associated variable are the reverse of the option name). Again, I > understand why such things happen, but *somebody* is going to have to > commit to better care and feeding of the channel, whatever software is > supporting it. Sure, but it could be anyone who wants to fix those things. By "really done" I didn't mean that the software was bug free and implemented 100% of it's intended uses. I just meant that they want to ship a usable product, of course there are still bugs or things that in hindshight would have been good to fix before release. > We at Mailman are very happy to help. We're also a small crew of > part-timers, so that's going to be limited, but at least we're aware > that Fedora's is one of the most heavily-used Mailman 3 installations, > so we have a strong interest in it working well! Máirín's mail to our > dev list got immediate and enthusiastic reaction. But we can't help > if you have no support for upgrading to upstream current release; > that's not our job (unless paid, and I'm not even sure that would > break our developers loose from other responsibilities and core work). Sure, and I don't know of anyone who said it was your responsibility. We have a ton of things going on, so we just had this as low priority currently. If it needs to be moved up higher the reasons for that would be great to know/discuss. > I'm not sure you can count on such support from Discourse, but I have > said more about that elsewhere in the thread, so I won't belabor it here. > > > You can always ask why we aren't upgrading. In this case it's because we > > are moving stuff to python36 from 34. If these fixes are urgent let us > > know and we can re-evaluate and try and get things faster. I was under > > the impression that the fixes were pretty minor. > > HyperKitty is a fairly complex piece of software. I never did make > head or tail of it (most of my time is devoted to core Mailman, > especially email security such as DMARC and crypto), and there's > nobody associated with the Mailman project to teach me about it > anymore. To me it's not surprising that the only things people are > willing to touch are minor. And even the original developers are > unlikely to be familiar with the current state of upstream, as > upstream has changes, some significant I think. While I know he's got many other things on his plate, I thought Aurelien was still doing upstream work and helping folks. CCing him on this, I could be mistaken. kevin
Attachment:
signature.asc
Description: OpenPGP digital signature
_______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx
