Is there any chance of upgrading the web-apps/mantis to the latest version? I suspect it is not as easy as just renaming the ebuild! Reproducible: Always Steps to Reproduce: 1. 2. 3.
>I suspect it is not as easy as just renaming the ebuild! Maybe. But if not developer uses it himself, the priority for not widely used packages is low unless someone (in this case you) requests it.
Created attachment 44332 [details] Mantis 0.19.1 Ebuild I updated the ebuild myself. The tar file name had been changed to mantis-* (instead of mantisbt-*) so tis required a couple of changes from the renamed 0.18 ebuild
IMO calling Mantis "not widely used" is a bit misinformed. I suspect that mantis has rapidly become one of the bugtrackers of choice. I suspect there are many Mantis users like me who simply dont use the portage ebuild becasue it is terribly out of date. Lets get this version bumped.
> calling Mantis "not widely used" is a bit misinformed. Maybe. I'm not touching web application ebuilds. It just takes time (and more people, who are willing to get their fingers dirty and become developers). Not everything is as well maintained, as we'd like it to. Also we have currently more than 2700 open requests for packages to become added to portage - and the number is growing...
How do you get to be a "developer"? The gentoo docs indicate that you should submit patches via bugzilla and then just wait to be invited. I've submitted a few ebuilds and patches via bugzilla (such as the attachment to this bug), and they are almost always just left like this one and never added, with the excuse being there aren't enough developers. I think so far only one patch has ever made it into the main CVS tree. We use Mantis extensively in our organisation, using the current release version which is 0.19.2 (which I built with gentoo using the ebuild I attached here (renamed to 0.19.2)). If I had CVS access I could have added it months ago. 0.19.3's release is imminent (though there is some discussion in the mantis mailing lists to call it 1.0 to indicate its maturity and stability). I'd be happy to contribute/commit a new ebuild for it when it is released.
>The gentoo docs indicate that you should submit patches via bugzilla and then just wait to be invited. >I've submitted a few ebuilds and patches via bugzilla Well, it's not that simple and it's more needed, than to attach a few patches. If a developer notes, that you're a constant source of help, be it via bugs.g.o, irc or mailing lists, he may have a look at your track record. If he thinks it looks good, he'll ask you, if you're interested and want to go through the process to join. He (or another developer) will be your mentor then and guides you through the probationary period. You have to read a lot of gentoo.org documentation, to pass a quiz and another question catalog - both eyed by the developer realtions herd. If they are fine with your answer you get an account and your mentor will be held liable for your mistakes for some time.
How does that help with packages like this where there doesn't appear to be an active developer for it? So there is nobody to notice if someone is a constant source of help. There seems to be loads of bugs like this one where there is stuff ready to go into the rsync tree but no developer interested enough to do it, yet there are several comments about a shortage of developers.
By the way if anyone want to install 0.19.2, you can just rename the 0.19.1 ebuild that is attached to this bug and put it in your portage overlay.
It's not about missing interest, it's about time constraints. As a developer you're supposed to have every ebuild tested and checked for side effects before committing something to cvs. As a user, it's perfectly fine to have a look at bugs.g.o or to write it yourself and put it in your overlay as long as there is no decent maintainership possible.
*** This bug has been marked as a duplicate of 77253 ***
