Is the bug database a circular file or is it used?
I would like some confirmation on this from an official source.
The OpenGL bug database appears, from all visual evidence, to be nothing more than /dev/null. Every bug filed here is left in the "new" state. Bugs are almost never fixed or responded to.
The only exception to this seems to be bugs for the man-pages, which Graham Sellers seems to be fairly responsive to. Though there are plenty of bugs that haven't been addressed there too, at least Graham seems to be at least someone present in the DB. That is, if you file a legitimate bug, there is a reasonable chance that it will be seen and corrected, and the bug closed as fixed.
In order for a bug database to be useful, for both the owners of it and its users, the database needs to be maintained. The owners need to respond to new bugs, even if that response is just to close them outright without comment. At least it shows that the bug was seen and responded to. If bugs are fixed, the the corresponding bug reports need to be closed. If there is an intent to fix a particular bug, then the bug report needs to be updated with that information. This allows users to trust the bug DB as a useful mechanism for getting things done.
So I just want to know: is time spent filing bugs there wasted? Or are these going to actually be seen by anyone with the power, time, and inclination to do something about them? If it's the former, can you suggest some alternate means of contacting the ARB that is more likely to result in action? And if it is the latter, then I would suggest investing 30 minutes a week into cleaning up the bug DB, so that it at least looks like it's an active place that the ARB uses.
The bug database does get looked at and cleaned up periodically, although I'll admit there does seem to be a long string of unresolved issues. Most of this is related to time - most of us are volunteers and, while we are paid by our respective companies to work on OpenGL, nobody at AMD (in my case) asked me to write man pages. I address bugs as I am able. I don't want to close out legitimate bugs without fixing them, but some require more time than I have available to get things sorted out. I will get to them eventually.
As for the bits outside the man pages... I'll see if we can get some of it cleaned up and the legitimate issues escalated.
hmm... last time you visited amd developer central is may, 2nd. i thought that was the place you can report amd-specific bugs, but it's abandoned now. i have no idea about current state of the bug i've reported several months ago. so what is the primary place to report bugs for amd drivers?
While we're at it: what about the Catalyst Linux Bugzilla? Activity there seems to lag behind a little as well.
I was wondering the same thing. Nvidia's developer zone isn't any better either. The only vendor I'm usually able to get a reply from these days is Apple.
Originally Posted by Nowhere-01
I completely understand the time issue. But I think you misunderstood the point.
Originally Posted by Graham Sellers
What I'm trying to get at is that the database looks like it isn't being used. Almost all of the bugs are still marked "new". To the reader, this status means, "Nobody's looked at this bug yet." And that's the main problem.
If you've taken the time to look at a bug and evaluate whether it is a legitimate bug or not, then you should also change the status of it from "new" to something appropriate. That way, people can see that yes, the owners of the bug DB do indeed see bugs. They may never be fixed, but at least a human being saw it, thought about it, and made some kind of decision based on it.
Even if you don't have time to fix the bug right then, the first step is to at least acknowledge the bug.