The article mentions "unaddressable bugs" while in effect the original message only says "unaddressed bugs". Big difference. The first suggests that BulletproofX is just a bad idea that can't be made to work while latter only suggests lack of time or support.
Personally I've always thought this was a pretty basic feature. Drop anyone who is not an X config wizard into a tty and they will have no idea what to do. Given them same basic VESA support and they'll still be able to Google around for solutions or chat with some more knowledgeable friends for support.
Good. Do you know how many users have come into #compiz-fusion on freenode asking for help because their display is driving at 800x600 and compiz won't work. /var/log/Xorg.0.log shows that it's using the failsafe xorg.conf file and, to make matters worse, /var/log/Xorg.0.log.old also ends up showing /etc/x11/xorg.conf.failsafe. So you then have to walk the user through stopping gdm, logging in at the console, trying to start X, copying a useful Xorg.0.log file to their desktop, and then restarting gdm.
This doesn't help anyone, including newbie users.
So a tty console is better?
I guess it would for you, they wouldn't get to #compiz-fusion in the first place.
Yes, dropping BulletProofX is better for a couple of reasons.
A) It forces users to learn something about their system.
B) It forces developers to come up with a better solution that the current piece of crap.
Btw. did somebody look at the new xorg.conf created on current intrepid? It is absolutely clear that it breaks on many systems because of the fbdev override.
Dropping tools that make setup easier accomplishes:
Originally Posted by adamk
A) Keeping new users away
B) Gives MS trolls more food
C) Slows down linux adoption
So if your net goal is to hang on or decrease the already minuscule ~1% marketshare linux has on the desktop then by all means scrap every gui tool that allows the layman to setup their system. Just don't ever bitch when a manufacturer or software vender doesn't do squat for linux because instead of spending resources on ~1% the market they decide to focus and deliver 100% of their product that appeals to 99% of the population.
When you look at the mail then you see that they just recreate the xorg.conf with defaults. Thats usually enough as it would even start without xorg.conf in many cases. Just the current default xorg.conf is not that optimial.
Allowing developers to develop hacky workarounds to major problems just results in a piss poor product.
Originally Posted by deanjo
Develop the correct solution and then deploy it. Seriously, all the developers need to do is make sure they back up the /var/log/Xorg.0.log file that shows the actual problem with the driver before starting X with the failsafe xorg.conf file.
EDIT: As an example of what I'm talking about... There's someone on #compiz-fusion now with a fglrx problem. He starts by saying that he can't get 1920x1200 working with fglrx. He's not even aware that the root of the problem is that BulletProofX kicked in and he's using the failsafe mode, so he doesn't relay that fact to us. It takes a series of questions before I realize that he's running in 800x600 mode and, therefore, fglrx must be failing and BulletProofX has kicked in.
This is what I call an absolutely crappy workaround. I'd rather have it dropped completely, forcing the users to complain to Ubuntu about being left at a TTY, so that the Ubuntu developers actually develop a real solution. In the short term, perhaps linux loses a few newbies... So what? In the long term, we end up with a better product.
Last edited by adamk; 07-07-2008 at 01:38 PM.