On Tue, 14 May 2002, Owen Taylor wrote: >> Be patient. >> >> We're all busy with other things, and there are plenty of patches still >> waiting in the queue. Please don't resend anything. > >Can I suggest, as a long term goal, having a publically viewable bug >tracker / patch queue? > >At least from my point of view, the current system isn't working very >well. > >If I find a bug in XFree86 (say, >http://bugzilla.gnome.org/show_bug.cgi?id=81783, which turned up 5 >minutes ago :-), it's frequently not clear how to proceed. > >Yes, I can send mail to [EMAIL PROTECTED] or [EMAIL PROTECTED]; b >bug report. But in either case it feel like a complete shot in the >dark. > > - I can't check on the status of my bug-report/patch. > > - I can't give someone else an URL to go to check on the the status. > > - I can't meaningfully give updates ... sending mail to "[EMAIL PROTECTED]" > saying "you know the patch I sent 2 months ago. Forget it, it turns > out to have been faulty hardware" at least seems like it won't > work very well. > > - There isn't any reliable way of telling if/when my bug has been applied. > > - If the person dealing with the bug report / patch wants to get > further information, they have communicate with me privately, > and > >I understand very well that something like bugzilla is considerable >amount of sysadmin work to set up and maintain that would take away >from someone's hacking. And there simply may not be the resources >currently. > >But for GTK+ and GNOME, we find it an extremely valuable tool to have this >stuff public ... not a panacea ... I still get plenty of people >annoyed at me for slow response to GTK+ patches, but at least they see >a milestone for the bug, and know it hasn't been lost.
I agree completely. We (Red Hat) receive a lot of bug reports against XFree86, many of which we just do not have the manpower to fix/look into, other distributions no doubt have the same problem. I think a lot of these problems end up falling between the cracks. What makes things a bit worse, is that a bug reported to a distro vendor that ends up getting fixed by that vendor, generally gets fixed in that one distribution at that point in time, and doesn't necessarily propagate to other vendors, or back to XFree86.org in a timely manner, or at all. Without pointing any fingers at anyone, some vendors are great with submitting patches back to XFree86 while others don't bother it seems IMHO. This puts more work on each distro vendor, to harvest patches from the other vendors packages also. When submitting a patch upstream, you sometimes do get immediate feedback on a particular patch, and sometimes do not. I never quite know if a patch I've submitted for example is applied, if the problem was fixed in a different way, or is still in someone's queue. I fully realize that everyone receiving the patches have full time jobs themselves, and don't have time to respond to the number of patches right away or apply them, since this is often the case for myself as well. The problem is though, that bugfixing/patching seems to not scale well at all. I've submitted 40-45 patches about a week ago or so, and I've had some feedback from a few core team members about a few of the patches, which was greatly appreciated. Many of the other patches I presume are in a main queue, or have been taken and put in personal patch queues of a given developer to look at in the future when they're working on that area of code and/or have time, etc. At least that is probably what I would do if I was receiving the patches this way. When a patch does eventually get applied though, one has to hope to catch it in a changelog message. I read changelogs and am on the CVS commits mailing list, however I'm sure many others are not, and would just appreciate knowing in a simple manner that their patch was applied or not, and if rejected, perhaps a reason. In the past I've experimented with submitting patches a bit, and I've found that submitting patches in an ongoing fashion as they are made, tends to not get them applied sooner unless it is a rather important issue with a straightforward fix. For our 7.3 development cycle I decided to submit the whole storm of patches all at once to save myself some work as I figured the bulk of the patches I was submitting, most likely would only be applied to the head of CVS, and probably only just before 4.3.0 was released. Submitting them in one shot was much less work than submitting them one at a time, possibly having to bugfix the patch and resubmit a few times, etc. when there was good chances they'd queue up until 4.3.0 was near. On the #xfree86 channel on irc.openprojects.net, I often point people to the [EMAIL PROTECTED] bug report list, or the current XFree86 web based bug submitter CGI. People who have sent a few reports there in the past have come back to me saying "Why should I bother reporting there? I never get a response back anyway." I generally tell them that XFree86.org doesn't have the resources to respond to every bug report right away, or at all in many cases, and that they should watch the Changelog for their bug to get fixed, as this allows developers to spend more time fixing bugs, and doing work on X than emailing people. Generally though, people tend to not care about that, they just expect to get an acknowledgment, and they seem discouraged when they don't get it. I get this sometimes even _with_ bugzilla. So I understand the difficulty involved with responding to every single bug report. The problem with our bugzilla though, is that we do not have expertise in all of the areas bugs are reported. Also, we do not support every tiny piece of XFree86 100%, so many issues reported are considered by us to be low priority and/or WONTFIX. In these cases, I recommend to people to report the bug upstream to [EMAIL PROTECTED] and/or [EMAIL PROTECTED] in hopes the maintainer will fix it, or at least that more people who could potentially fix the bug are looking at it. Since I am also on the receiving end of XFree86 bug reports (via our bugzilla, with emails sent to me upon new bug report submissions and updates), I know the volume of incoming reports can be a bit staggering, and that many are bogus and/or user misconfiguraton issues, or people just looking for free-ride tech support. However, for both bug reports, and for patches/fixes, I strongly feel that the administrative overhead that bugzilla would require would be small compared to the service that it provides, the patches it harvests into one common place, and to the two-way communication mechanism it would establish between X users, the XFree86 core team, and other XFree86 developers like myself. Red Hat used to deal with bug reports via email a long time ago as well. The problem with it was that it was not scalable, it did not allow a good two-way communication between bug reporter and bug fixer(s), and it did not allow for bug tracking, co-ordination, or allow for searching for already reported bugs, and their solutions/workarounds. Also, the details of a user's given bug report could easily be lost in email, or forgotten entirely. Developers can't keep track easily of what issues are being worked on by other developers. Bugzilla IMHO would allow more people who are not on the core team to be aware of more of the bugs that are being reported, and would result in more people fixing more bugs. People often come to me and express an interest in working on X, but don't know where to start or what to work on. I refer them to our bugzilla, as there isn't a way they can find out about bugs at XFree86.org currently. Some tell me "I don't use Red Hat", somehow thinking that all XFree86 bugs in our bugzilla are specific to Red Hat Linux, others not wanting to "contribute code that does Red Hat's job for them" like somehow fixing every bug reported is "our job". I would much rather be able to point people at a central bugzilla tracker on XFree86.org, and to be able to consolidate non-distribution specific bug reports in one place, which will remove some level of duplication among all distributions, and including the core team itself. This topic has come up before on private XFree86 mailing lists, and I've also discussed it a bit with Keith Packard (whom I've added to the CC) and Jim Gettys. Keith set up bugzilla a while back to investigate it, and I've poked around a bit locally with it as well. I dunno how far he's gotten with it, but I havn't had much time to dig into it, and I presume he hasn't either. Almost all major projects out there (Mozilla, GNOME, KDE, etc.) have a web based bug tracker, and most of the larger projects use bugzilla for the large number of features that it has that cater to large projects like XFree86. I think XFree86 really needs a bugzilla bug tracker in order to harvest the wealth of developers that are out there that currently see X as a black art, or a mysterious voodoo tarball. Lets make XFree86 development, and bug fixing more scalable, and pool the community of developers and users together. Lets work together to put up a bugzilla bug tracker on XFree86.org and use it as the primary tracker of bug reports. I think that would benefit everyone, and most importantly the end user. TTYL -- Mike A. Harris Shipping/mailing address: OS Systems Engineer 190 Pittsburgh Ave., Sault Ste. Marie, XFree86 maintainer Ontario, Canada, P6C 5B3 Red Hat Inc. http://www.redhat.com ftp://people.redhat.com/mharris _______________________________________________ Xpert mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xpert