[Openicc] Fedora CM, was: Google Summer of Code . . .

[Chris Murphy]

On May 21, 2010, at 9:25 AM, Hal V. Engel wrote:

> On Friday 21 May 2010 01:57:45 am Richard Hughes wrote:
>> Lets assume there are approximately a billion personal computers, i.e.
>> desktops and laptops. Lets say 0.5% of them are being used with a
>> fairly up to date Linux, which I think is somewhat optimistic. That's
>> 5 million Linux desktops to aim for.
>> 
> 
> I see this type of think fairly often on Linux forums and email lists.  But 
> are only 0.5% of end user machines running Linux?    
> 
> Several years ago I read a report (in InfoWorld I think) that put the 
> installed market share of Linux end user machines at 4%.

Two points related to this persistent statistic of conversation from the color geek who uses Macs (in my defense I do have Fedora running on a separate laptop so that I might only be 98% ignorant rather than 100%):

1.
For all practical purposes Linux forms the basis of Android. Gartner expects Android to take over the iPhone by 2012. It doesn't matter if that happens or not, what does matter is that the number of handheld devices running Android is significant and likely will continue to be significant.

2.
There's every indication that iPhone OS 4.0 has/will have some kind of color management capability in it. Exactly what that looks like and how it will compare to the desktop version of OS X I don't know yet.

So when I think about these two things, along what I've read up to this point, it means there is relevance for color management, on Linux (or variants thereof) in the millions, not thousands. It needs to be really efficient, and I agree with Richard in that it needs to be compact.

What does really basic color management entail? And what would it take for every application to get at least some of those basic capabilities for free?

Now we could talk about a more opt-in approach that works better with Linux culture compared to the "you must sacrifice 10 goats between 23:06:09.001 and 23:06:09:005 on the evening before a full moon in order to opt-out" that is approximately the case on Mac OS X*. But the closer opt-in is merely a developer setting a flag in their app to get some really useful capabilities (and maybe it's just one), the more their application and users benefit from this effort. The more it makes the platform grow.


Chris Murphy

* This would be based on the North American atomic clock managed by NIST, but naturally this detail would not be in the documentation, it would be assumed.



>  This was in Dec. 
> 2004 and at that point the number of Linux end user machines was greater than 
> the number of Macs (about 3.8% of all end user machines were Macs according to 
> the report).  In addition the researchers who wrote the article estimated that 
> by 2010 that the installed base of end user machines would have about a 10% 
> Linux share.    I think based on other reports I have read that the growth 
> rate of Linux has exceeded these estimates.
> 
> I am only pointing this out because it appears to me that many in the Linux 
> community get the number so wrong and end up making estimates that would have 
> been correct in the late 1990s rather than using numbers that are more up to 
> date.   Using outdated numbers makes the Linux community look like a fringe 
> group when in fact Linux is well on it's way to becoming mainstream and is the 
> number two end user OS. 
> 
> Hal  
> _______________________________________________
> openicc mailing list
> openicc at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/openicc