Does Rising Linux Usage Share = Increasing AMD Open Source Driver Development?

[Prescience500]

As I'm sure many other people have noticed, this is an exciting time for Linux. Wayland is rapidly progressing. Open source video drivers are improving at a much quicker pace. Now, apparently Linux usage share among personal desktop computers is increasing. http://netmarketshare.com/operating-...=9&qpcustomb=0

I think I remember Bridgeman once saying that the resources that AMD intends to put into open source drivers will be somewhat proportional to usage share. Does this mean we can expect or at least hope for addditional resources pouring in? Even if not, the extra usage share may bring in new developers, bug testers, software, and market power. Does anyone have any thoughts or hopes?

[Qaridarium]

we write this with Bridgman in multiple places and multiple times.

and the result ist: no amd will not grow there open-source driver team because there team are already bigger than there open-source driver market share.

and they also think if the linux market grow they all use the Catalyst they really think no one use the open-source driver )Read: market share is to low(.

its not only windows vs linux its catalyst vs radeon driver.

in fact the radeon driver is payed by the Server CPU (Opteron) customers only because they want OpenCL on GPUs as a co-processor. because of this they work on OpenCL but not on power-managment. And these linux opteron customers don't want closed source catalyst. *

#Also the forum show the truth: Bridgman don't think Linux grow on the desktop he think only google chrombook OS and andorid grow and because of this the website statistic show more linux systems. '

also amd think andorid (2% market share) is a more important market than GNU/Linux (1,5% market share)

so your news about linux market share grow dosn't matter for AMD!

[Prescience500]

Hmm...well, by early summer, Linux "desktop" market share will be at 2% if this trend continutes and hopefully it will continue after that. Maybe that will help change things.

[bridgman]

Q, you should of course feel free to post your theories and opinions but please don't imply that your information is coming from AMD in general or me in particular, unless it actually is (which is not the case here).

and the result ist: no amd will not grow there open-source driver team because there team are already bigger than there open-source driver market share.

No, my comments in this area were about Catalyst not the open source effort.

On the open source side we are not trying to "write the driver ourselves" and there is no expectation for the team size to scale with market share; the approach is to be *part* of the community and focus on getting information and initial support into developers hands. We also have a couple of developers working on specific areas of functionality which are of interest to Embedded customers (since the Embedded business unit is funding them).

and they also think if the linux market grow they all use the Catalyst they really think no one use the open-source driver )Read: market share is to low(.

Huh ? Number one, no. Number two, what does Catalyst vs radeon have to do with Linux market share ?

its not only windows vs linux its catalyst vs radeon driver.

:confused:

in fact the radeon driver is payed by the Server CPU (Opteron) customers only because they want OpenCL on GPUs as a co-processor. because of this they work on OpenCL but not on power-managment. And these linux opteron customers don't want closed source catalyst. *

Um... no again. Open source OpenCL work is targetted at embedded customers, not server customers.

#Also the forum show the truth: Bridgman don't think Linux grow on the desktop he think only google chrombook OS and andorid grow and because of this the website statistic show more linux systems. '

Sure, but that was what the previous charts said (the ones you sent me) at the time and you also agreed with that view. These charts are showing a slightly different picture (although some other sites are starting to show the same growth).

I never mentioned ChromeOS, and the links you sent showed desktop linux share relatively static plus about 2% new from Android.

also amd think andorid (2% market share) is a more important market than GNU/Linux (1,5% market share)

??? They require fairly different solutions (look at the multimedia stacks, for example), although in the short term there are development tasks that can benefit both and that's where we are focusing. Don't believe I have ever commented on which was more important, just the extent that the same solution would work for both.

so your news about linux market share grow dosn't matter for AMD!

Again, ??? Maybe in the sense that (as I have said here several times) our current staffing level is based on ~2% share for desktop Linux in anticipation of future growth. Historically the "measured" market share seems to have hovered between 1.0 and 1.25% (it went up during the Vista days then slipped back a bit after Win7 came out), so the fact that it's now getting closer to 1.5% is definitely interesting.

[Qaridarium]

Q, you should of course feel free to post your theories and opinions but please don't imply that your information is coming from AMD in general or me in particular, unless it actually is (which is not the case here).

?? i never dit this i only point out what he can find if he read in this forum carefully.
i also can help him with Quotes to find the right stuff.

No, my comments in this area were about Catalyst not the open source effort.

even your writing here in this message prove me right about the market share tropic.

i point out that the market share doesn't matter for AMD because: "On the open source side we are not trying to "write the driver ourselves" and there is no expectation for the team size to scale with market share; the approach is to be *part* of the community and focus on getting information and initial support into developers hands. "

Be a part of the community only means: no we don't increase the team if the market share grow.
you say the same "there is no expectation for the team size to scale with market share" but if i write the same stuff you claim I'm wrong.

but in fact if the market share grow the FGLRX/Catalyst team will get more people.
your words dude. this means catalyst get all the marketshare grow money and radeon don't get the money.

because of this i point out there is a war not only windows vs linux also catalyst vs radeon.

"our current staffing level is based on ~2% share for desktop Linux in anticipation of future growth."

maybe your company think there is a future grows and my own personal opinion is the same. but your personal (bridgman) opinion is difference and i can prove this with quotes.

here my prove:

If the cost of "giving more" is a lot higher than the benefit from selling more, I wouldn't call that a win. This is the one place we never seem to be able to connect.

my answer to this:
"This is just speculation because its a chicken vs egg problem you don't sell well because you don't invest money and because you don't invest money you don't sell well.
Stupid argumentum ad circulum. (Self-fulfilling prophecy bullshit)
But yes i know every stupid economic school teaching and university lecture teach this bullshit. but in fact its just a "Eristic Dialektic pe nefas" -->
"Evil technique with unauthorized sugical and micromanipulative methods "
This bullshit is know over 1000 of years and you are not the first one trying this crap.
o man i read to much old scool "Arthur Schopenhauer - Eristische Dialektik" http://de.wikipedia.org/wiki/Eristische_Dialektik"

If the cost of doing what is being asked is higher than the money we would make from having 100% share of the affected Linux market then I would argue it's not speculation.

my answer to this:
"and you do it AGAIN! because your claimed fix numbers are NOT FIX! '
"than the money we would make from having 100% share of the affected Linux market "
what is 100% of the linux market in the future? this is just Speculation!
this is just bullshit because of the "Grow" if you do have 40% grow every 5 month you have 100% market share after a longer time.
and then the 100% of the linux market in the future is? the HOLE MARKET:
and now you can say not YET but this is also wrong because its only not yet because you do not invest.
ok and you will try it again. "normal people let you go way with this but i will not. "

There is only one logical conclusion in your writing you think the Linux desktop market share don't grow.

"so the fact that it's now getting closer to 1.5% is definitely interesting"

Really? no it isn't because: only the catalyst Linux driver grow in staff because of this but not the radeon driver. this meansironie) its better for the war radeon vs catalyst driver if linux lose market-share (/ironie)

"They require fairly different solutions"

not really wayland+gallium3D fit perfect for desktop and smartphone/touch tabled computers.

"Sure, but that was what the previous charts said (the ones you sent me) at the time and you also agreed with that view. These charts are showing a slightly different picture (although some other sites are starting to show the same growth). "

no i just don't share this viewpoint.
for me andorid is a LINUX- Universal Operating system and in my point of view the market share of desktop Linux grow.
you can have the same driver "radeon+wayland+gallium3D" for desktop+smartphone+tablet pc.

only X does not fit well.

"I never mentioned ChromeOS, and the links you sent showed desktop linux share relatively static plus about 2% new from Android."

my point is: andorid IS A LINUX in the same way you can mark redhat not as a linux because its not debian based. and hey suse linux is not linux because is not debian.
but ubuntu is a true linux because its debian based LOL!

the gnu system is only called linux because of the usage of the linux kernel in fakt andorid is called linux because of the usage of the linux kernel.

"andorid" is just a useless brand of a linux kernel system.

"Um... no again. Open source OpenCL work is targetted at embedded customers, not server customers."

LOL... really its only not target as a server system right now because nvidia+cuda+tesla do have 100% marketshare in this market.

maybe amd should think about this nvidia monopole.

"Huh ? Number one, no. Number two, what does Catalyst vs radeon have to do with Linux market share ?"

simple: the linux market share is so low because of the shitty catalyst driver.

without the radeon driver windows is much better in running catalyst!

i never EVER had any problem with the catalyst on windows in the past but many many problems with the catalyst on linux.

its a matter of fact the linux market share on the desktop only grow because of the radeon driver not because of the shit with the name catalyst.

[Prescience500]

my answer to this:
"This is just speculation because its a chicken vs egg problem you don't sell well because you don't invest money and because you don't invest money you don't sell well.
Stupid argumentum ad circulum. (Self-fulfilling prophecy bullshit)
But yes i know every stupid economic school teaching and university lecture teach this bullshit. but in fact its just a "Eristic Dialektic pe nefas" -->
"Evil technique with unauthorized sugical and micromanipulative methods "
This bullshit is know over 1000 of years and you are not the first one trying this crap.
o man i read to much old scool "Arthur Schopenhauer - Eristische Dialektik" http://de.wikipedia.org/wiki/Eristische_Dialektik"

In school, I majored in economics and I assure you that this was definately not taught. In fact, we were taught that in the long run, it was a fallacy.

So...I'm guessing from the back and forth that Linux desktop marketshare will need to would rise above 2% for a real difference to be made. If Linux spending is not scaling, then maybe 3-10% before more spending on open source drivers? (Essencially meaning that a relatively high market share might cause a policy change or something.) Call me an optimist, but prehaps a higher market share could mean more independant developers working on the open source drivers, too.

[Qaridarium]

In school, I majored in economics and I assure you that this was definately not taught. In fact, we were taught that in the long run, it was a fallacy.

were do you come from? the US people don't act like they get the same education.

US people do never have a fallacy because they are blessed by God !

we read many stuff about the US educating in Germany : "Police in U.S. schools:
Behave yourself or I'll shoot!" http://www.spiegel.de/schulspiegel/a...812885,00.html



they start a war against 10 year old kids with disabilities and learning disorders and put the poor childs in prison a LIFE TIME!.

really the US people FAIL! "Zero Tolerance" only mean: Zero Brain.

So...I'm guessing from the back and forth that Linux desktop marketshare will need to would rise above 2% for a real difference to be made. If Linux spending is not scaling, then maybe 3-10% before more spending on open source drivers? (Essencially meaning that a relatively high market share might cause a policy change or something.) Call me an optimist, but prehaps a higher market share could mean more independant developers working on the open source drivers, too.

you don't get bridgmans words the radeon driver team will relatively get less money compared to the catalyst because the catalyst get more money if the market share grow.

if we get ZERO market share on the Desktop+workstation they close the linux catalyst completely then the radeon driver wins the battle.

after that linux can get 100 market share on the desktop just because: windows is better in playing catalyst.