I wasn't actually expecting a hotfix for Linux (previously they were Windows only) so I don't have a lot of answers available off the top of my head. I suspect the answer to most of your questions is "because it's our first Linux hotfix" ;)
Presumably the signature is missing because it *is* a hotfix driver. If the driver had been fully QA'ed it probably would have been shipped in place of the standard Cat 9.12 driver. If we are going to continue to provide Linux hotfix drivers then we probably need to find a less annoying way to distinguish between production and hotfix drivers on user systems.
Remember that XvBA hasn't actually been *released* yet, the work-in-process code has simply been *discovered* and some people are using it beneficially.
There used to be two Linux-specific teams - one under Matthew focusing on the core X driver, and one under Larry focusing on customer and distro issues. For the moment I believe Larry is managing the whole group while we figure out whether to replace Matthew's role and leave the org unchanged, or to shift more of the Linux responsiblity into the core component teams which handle cross-OS development.
Well i ask if somebody feels free to update the driver to fix the issues or if there is nobody responsible and nothing will happen (which is more likely). Somebody should responsible who really cares about issues.
A hotfix for Linux
Hi Yall & Bridgman,
kudos guys, for releasing a LINUX hotfix. :-)
Bridgman-makes it a bit hard on you, as you are the much taxed spokesgeek for ATI/AMD, if "....wasn't actually expecting a hotfix." Larry should keep a spokesgeek, in the loop? ;-)
Human error/xmouse rush, etc happens. Is there enough time to release a 2nd hotfix, that fixes the first?
Well i wrote a mod tool for ati installers long ago, basically i could fix those issues myself but i could not upload to ati of course ;)
I don't think we've ever had an official spokes... um.. person for the proprietary driver, although Matthew, Pierre and I all helped out where we could.
My job has always been primarily on the open source sice - I'm not in the loop on proprietary driver releases so there was no reason for me to have known about the hotfix -- I'm just saying that if I was had known I might have asked some questions during the week and so been able to answer questions on the weekend ;)
Well consider nvnews as example (like nvidia handles it) and let the fglrx devs answer in a well known board. it's just that easy.
We do that mostly through the bug tracker right now, although Pierre (OpenGL) comes on here sometimes as well.
the bugtracker is crap when nobody ever gets a response.
Why hasn't the bugtracker been updated to be able to tag bugs for current fglrx releases. Last time I checked it only listed up to 9.7.
But people do get responses (you know that) so it's not crap, right ?
Originally Posted by Kano
If you want to say "not enough tickets get responses" or "even if the submitter insults the developers and provides no useful information the problem should still be fixed instantly" that's OK, but developers *do* respond to tickets when they are trying to diagnose the problem and need more info.
I get copied on every ticket update and I see that happening.
It's not clear whether Matthew or someone outside ATI/AMD was updating the release pick list, or what the account info was if we were doing it.
Originally Posted by Melcar
Last I heard the guys were following up with Matthew.