Better Linux support without allocating any additional resources?
Release documentation. Proper documentation.
How are they going to produce documentation without allocating additional resources in terms of labor costs? Good documentation doesn't generate itself (unless it's been cleverly programmed to do so, which I am guessing is not the case here).
How they can provide better Linux support without allocating any additional resources?
There are many things I would prefer them to do, such as releasing old drivers
as open source, but even that costs money for the vetting process.
The only thing I can think of that comes close to those conditions would be;
Release new drivers for Windows, and for Linux if there are any, for upcomming
cards as open source.
Not ideal in any way, but atleast something more than nothing.
The start about 6 years ago was promising for me.
I thought, well, let that laptop be fine with Gentoo or any other kind of Linux. The price was an issue so I had to go on the low priced region.
Back at that time I think there weren't even those horrible glare screens. But still choice was limited. And then that Elitegroup stuff was there. WITHOUT Windows. Awesome I thought, researched and finally found the ECS G320.
I was fond by the concept of low power processors and the combination with the padlock crypto accellerator. And everything else was told to run with free software, no blobs.
Yep. Besides the softmodem it did (which works but uses somewhere a piece of binary blob).
And the laptop has lots of classic interfaces you would not find anywhere else.
Just that GPU would never get to fly.
But there was still hope. So I got myself in that mini-itx stuff, interesting things, not that cheap but low power... humm, sounded nice. Since there were claims and promises by VIA that development was going on that there had been specs for the devs... I thought everything was going to be okay.
I used VIA in this C3-2 CLE266 and a C7 with CN700 now for years, before I had VIA chipsets here an there on mainboards. Some things were always a little shoddy, but most stuff worked. I was used to take a lot of pain since I had worked with Windows long before (though I took me ages to move from DOS to Windows). So I did not complain too much, still, the stupid GPU issues drove me somewhat crazy, especially since I wanted the C7 to be also a HTPC and not just a fileserver which could operate on low profile (or totally headless).
But that desaster, that Borderline style behavior of VIA!
Pull you close, seeming to hug you, making stuff that is interesting for HTPCs, file servers, small office boxes and so on... but then pushing you away.
Pulling you back with some hints like "Linux" in the supported OS list (and I believed them), then, again pushing you away. Telling to release specs, installing even Harald Welte (and I thought it was finally going to be ok then), pretending to share with openchrome devs (long after Luc was gone from development - and I still remember his angry shouts in the VIA forums)... then again pushing you away.
It really feels like one of my relationships with borderline personalities. A rollercoaster of fly high and fall deep.
You know.... After that you just want to vomit!
I now know to avoid borderline people. And I will sooooo avoid VIA, you can't believe it.
After 6 or 7 years I'm fed up finally. Nearly 2 years ago I decided to stay away from VIA's stuff. Not buy anything new until the situation has cleared.
Oh yes, and that means even avoiding buying a mainboard from e.g. ASUS that has a VIA soundchip or anything like that on it.
And that I will continue to do. I will tell all my friends, relatives, coworkers and other people to stay away from VIA's crap. Be it CPUs, GPUs, NICs, Audio chips and whatsnot (b.t.w. the audio drivers and software they make for windows are also horrible).
VIA could have had a nice market niche with their low power mini-itx but they failed utterly. They came to a time where the x86 folks at intel and AMD only had big, hot, hungry boxes to offer. Transmeta was never going airborne, but at VIA they had some thing for the corner of low power computing. Nothing that any Windows gaming freak would ever buy. They could have had. But they were so utterly stupid I can't believe it.
What chance do they stand when it comes to real CPU/GPU power? What do they just have in production? Now AMD, intel and havin low power x86 options with computing power. From the other side there is ARM design, even nvidia is eyeing in that direction.
That was you chance VIA, I hope you die very soon and quickly. So nobody will ever fall again for your lies.
By the way I want my money back, you ignorant as*hats!
I'm happy with my AMD stuff that just works.
And with the new Bobcat APUs ... I sense something that will make nice HTPCs, file servers and stuff. And the bulldozers for the larger computing demand. As soon as video playback works as good with the free drivers like nvidia's stuff does ... what would be missing then?
So people, buy from the good AMD , buy maybe intel stuff (if it isn't poulsbo -> ImgTec) or if not otherwise possible even nvidia. (Not much FOSS involvement but at least they have working binaries. Though I can't stand that snooty Jen-Hsun Huang.)
To say something productive:
I told years ago what all the others like Greg K-H tell.
Release the specs. You do not need to have the manpower or the Linux knowledge. Release the specs (completely) to the community. There ARE devs out there that will take care of it. But do not frustrate them by NDAs, incomplete specs and other nastyness.
I mean, it's easy. Isn't it? Do nearly nothing but have a driver bringing you hardware to fly on Linux and further on to other OSs and people. But I guess they will die without having understood.
That's why VIA is one of the worst chip makers that I ever seen.
They think that Microsoft will last forever. However they are wrong.
Microsoft will abandon X86 architecture gradually and VIA's investment in X86 will be compromised.
VIA will soon regret of letting Linux as a secondary operating system.
Actually VIA is going to be able to keep on their current path, they've been producing ARM parts for years now, and MS is porting NT and all MS internal apps to ARM, a task they likely started long before they formally announced it the other day.
VIA/S3 shot their own foot off, a few moths ago I'd say they where ripe for a merger, but now I doubt the current politics in the US would let their X86 license be transferred to anyone else since Intel would just lean on the Federal Trade Commission to nix it.
Expect to see VIA drop out of the X86 market completely in the next 18 months to a year to only produce ARM based parts.
VIA => "So much potential and opportunity wasted".
Mini-ITX => They squandered the situation by charging premium prices for their under-performing C3/C7-based crap. Intel and AMD have taken this form factor over with far better products at a decent price.
Nano? => Could have given Intel Atom a run for their money...Had they released more tempting products.
Nano X2? => Aren't we a bit late to the party?
Hardware support? => If you want to know why no one makes VIA based products; that's because their development support is shockingly poor.
I own a VIA Mini-ITX (C3 1Ghz with CLE266 chipset) mobo with dual LAN connections. I never used it anything more than a network gateway for the home network. I don't worry about its specific features it can offer. Mainly because I don't want to wait around for something that won't happen.
My next Mini-ITX mobo will either be Intel or AMD based...This is the price VIA pays in the long term; for focusing on being mediocre.
Releasing windows drivers is not really an option because a lot of the driver infrastructure is provided by and owned by Microsoft. For GL, a lot of vendors base their implementation on the SGI sample implementation which is owned by SGI. This 3rd party IP cannot be released without permission from the owners.