A First Look At The 2010 Linux Graphics Survey Results

[allquixotic]

Looks like this was thrown together without any proofreading whatsoever, and he didn't tell us how many total respondents there were in the survey. But, to his credit, he got rid of the pie charts and is back to the significantly more useful bar charts.

I still think that a database, allowing people (maybe just paying people, but whatever) to cross-reference and more deeply analyze the data, would be very useful. For example, answering a question like, "out of the people who wanted better 3d / OpenGL performance, what percentage of those were also interested in licensing/open source?" You cannot answer this question with the data the way it is presented.

Basically, even with the bar charts (which are an improvement, nevertheless), this valuable data is largely going to waste, because the potential to draw conclusions from it is rather limited.

[curaga]

Call out for guest articles?

[Jimbo]

The error in the charts is important and misleading. Since some time ago I have the impression that Michael is really overloaded. There are a lot stuff on linux and OSS world, he can not track every software, driver, hardware, etc. correctly in a very professional way.

The amount of articles he writes and the test setups he prepares is a lot for only 1 person. Maybe a change on the methodology could be useful, write less articles, use less computers for the tests, or whatever is necessary.

[pvtcupcakes]

I think it's awesome that the open source Ati drivers have more users than fglrx.

I'm not sure I want the fglrx team working on the open source stuff though given their track record.

[m_gol]

The monitor question is traditionally difficult. Resolution across and expanded desktop, multiple monitors of varying sizes and different aspect ratios, refresh rates, etc etc etc, and then which is the primary display, which is secondary, etc etc etc.

Yeah, but when asked in the way it was, it doesn't even correctly cover the most common single display case.

[energyman]

On the diagram "How many display(s) do you have", page 3, there are several "multiple" columns which probably need to be summed up?

Otherwise it looks like that Single display gets unfair advantage.

emm - no?

people set up their display with randr - people set up multiple displays with randr
peiple set up their display with xinerama - people set up multiple displays with xinerama.

[mendieta]

sorry, i couldn't go through the 7 pages of comments, so this may have been proposed.

But can we have a simpler Survey next year, where most of the questions are the same for all your systems, and then you only add specific information for each of your systems? I was starting to participate, but when I realized I was only half way and I had 3 more full surveys to go I gave up

Actually, it would be even better to run a program in each machine to generate all the hardware info, maybe a port of the ubuntu bug-report tool! This way it's also less likely that someone might cheat (not sure why would folks do that, but better safe than sorry)

[mendieta]

sorry, i couldn't go through the 7 pages of comments, so this may have been proposed.

But can we have a simpler Survey next year, where most of the questions are the same for all your systems, and then you only add specific information for each of your systems? I was starting to participate, but when I realized I was only half way and I had 3 more full surveys to go I gave up

[mirv]

The monitor question is traditionally difficult. Resolution across and expanded desktop, multiple monitors of varying sizes and different aspect ratios, refresh rates, etc etc etc, and then which is the primary display, which is secondary, etc etc etc.

[m_gol]

1440x900=1296000
1280x1024=1310720
1600x1200=1920000

That would be less than 1600x1200.

If You count pixels, it would be less than 1280x1024. But who says there is any linear order amongst resolutions at all? I still think that this question was incorrectly formulated.