VIA Releases FOSS Graphics Driver

billybob2 writes "VIA has released a 113,800 line open source graphics driver with full mode-setting support for CRT, LCD, and DVI devices along with 2D, X-Video, and cursor acceleration. Harald Welte, VIA's open source representative, states that the next step is to add 3D (see preview), TV-out, and hardware codec support while integrating this work with existing open source projects. VIA has pre-installed Linux on a significant portion of the company's latest products, including the EVEREX gPC2, 15.4" gBook, and CloudBook. It has also helped port the open source CoreBoot BIOS (previously LinuxBIOS) to several of its motherboards." VIA seems to be making good on the promise of its open source initiative announced last April.

Comments

Obligatory

[Nefarious Wheel]

Yes, but does it support -- World of Warcraft?

Re:Obligatory

[creature124]

You don't even have to RTFA to find that out. it says it right there in the summary - 3D support is the NEXT step, but it isn't there yet.

Re:Obligatory

[Merlin42]

Not yet, but soon hopefully. As stated in the OP.

It is really cool to see more hardware vendors moving to open source. Drivers are one area where more eyes are needed to help make the bugs shallow.

I do hope this pans out...

[fuzzyfuzzyfungus]

I hope that this goes better than VIA's prior activities in this area. VIA has some very, very cute hardware for linux project purposes(loads of small form factor boards, without the restrictions that intel has been putting on atom), some decently interesting netbooks, etc.

If I can trust that VIA video will actually work properly under linux, their boards become considerably more attractive for my purposes. The prospect of coreboot support for such boards would be gravy. I'd love to be able to put together some little linux widgets with linux burned right into the motherboard.

Re:I do hope this pans out...

[Anonymous Coward]

Via really has no choice.

The intel 945G chipset for Atom is fully documented and has quite good open source 3d drivers.

Atom kills VIA in Price/Performance/Power ratio across the board.

Once Intel fixes the problem of their north bridge requiring 6x the power Atom does then via is in really big trouble

It's interesting to see via go from ruling the mini-ITX market to now desparately having to play catchup in such a short time.

Re:I do hope this pans out...

[Jorophose]

Via really has no choice.

Agreed, everyone except for nVidia and maybe Matrox (side note: what a shitty company) is opening their specs.

The intel 945G chipset for Atom is fully documented and has quite good open source 3d drivers.

It sucks up 22W+ by itself though, and is very old. It's nothing compared to the VX800 or CN896.

Atom kills VIA in Price/Performance/Power ratio across the board.

-Price: Maybe. If you just want entry-level options (ie bare to the bone) and don't care about power usage, it's definately cheaper. Normal VIA parts are sold like boutique items. Except, strangely, their mATX boards go for 50$.
-Performance: Definately not, now that Nano has been released (but damnit sell 'em at retail!).
-Power ration: What? Nano desktop parts are what people have been measuring. Typical ULV C7s are like 4W-7W. Considering you get a chipset that ranges in that wattage too, and this is honest counting unlike Intel, VIA certainely has the upper hand.

Not to mention they don't need a P4 connector...

Once Intel fixes the problem of their north bridge requiring 6x the power Atom does then via is in really big trouble

Unlikely. Intel does not want to lose Celeron sales for the Atom. So their miniITX boards remain crap so they can sell whatever 945G boards they have left over that failed their low-voltage tests.

It's interesting to see via go from ruling the mini-ITX market to now desparately having to play catchup in such a short time.

I wouldn't call it catch up, but it's nice to see Intel and VIA compete. The only thing is I hope it drives down the price of VIA parts, at least within the 90$-150$ range, otherwise it's been a waste of time.

Re:I do hope this pans out...

[Anonymous Coward]

Matrox used to have wonderfully open hardware with the specs downloadable directly from their own website and a dedicated developer relations team to help with any queries. They were the perfect model of how to do it.

Then something happened. I don't know what: brain-slugs, possibly. They yanked everything, even the specs for older hardware, and stopped communicating. What a bunch of dicks.

Re:I do hope this pans out...

[antime]

But in the Atom SoCs Intel are using PowerVR, which is definitely not opensource-friendly. Until someone buys ImgTec the most you'll ever get is a binary blob driver.

Re:I do hope this pans out...

[Patrick Georgi]

The intel 945G chipset for Atom is fully documented and has quite good open source 3d drivers.

Our company works with almost a dozen hardware vendors, and none of them are so hard to work with and so open source hostile as intel. Try getting the documentation for the RAM controller of the chipset you mentioned.

Almost unbelievable

[Skinkie]

I really find it hard to accept that a company that around 5 years ago copied GPL code in many of their stuff made such a 180 turn and is now with full commitment in actually supporting the stuff that they have been copying for so long. The motives behind it and better: who was able to make this shift possible from inside the company, hiring an OpenSource devver is one, but the process before that is much more interesting.

Re:Almost unbelievable

[digidave]

Obviously their motive is profit. They went the route of stealing code (although that might not have been management, just some rogue coder taking the easy way out) and it didn't work. VIA understands that there is a large and growing Linux community and that there is money to be made from being Linux-friendly.

Just because their motive isn't selfless doesn't mean Linux supporters shouldn't welcome VIA with open arms. This is the sort of support we've wanted for many years.

Re:Almost unbelievable

[rossz]

Maybe they wronged the open source community in the past, maybe they didn't (I personally don't know). Let's show them that we are forgiving of past mistakes and fully welcome them and their donated code into the FOSS world. They made things right, let's not dwell on the past.

Re:Almost unbelievable

[rtb61]

It is likely they went through a process of discovery. The discovered that keeping the software open source has very little impact upon maintaining competitive advantage on the hardware or making innovative leaps in hardware design and keeping those proprietary. For hardware producers, software is just another overhead and working to minimise that cost makes sense.

There is a real push to achieve low cost ubiquitous computing, UMPC's, smartphone/PDA etc. and every cost saving makes it far more achievable and obviously maintains reasonable profit margins for the hardware manufacturers.

At the moment hardware manufacturers find their profit margins squeezed while their products are carrying closed source proprietary software with 10 times the profit margin, it makes absolutely no business sense as a hard ware manufacturer to put up with this. I am sure most hardware manufacturers thought that M$'s idea of free hardware and 'renting' the software was a load of B$.

Re:Almost unbelievable

[Anonymous Coward]

They made things right, let's not dwell on the past.

I'm trapped in the past, you insensitive clod!

Then for you it's the present, stop complaining.

Re:Almost unbelievable

[bill_mcgonigle]

The discovered that keeping the software open source has very little impact upon maintaining competitive advantage on the hardware or making innovative leaps in hardware design and keeping those proprietary

I've made it a point to mention the open source driver problem in just about every other e-mail to my Via rep. My guess is a few hundred other developers were doing the same thing. I've also made it a point to express gratitude on each win. Yes, it's good for them, they should have done it anyway, but it's going to make my life a bunch easier too.

Re:Almost unbelievable

[TheLink]

Their business is selling hardware (for now anyway).

If they open source the drivers, there's a chance that they can cut costs - there's a significant chance someone _else_ (redhat, suse, ubuntu, etc) might end up doing the work of keeping the drivers for the _old_ hardware working with the various Linux kernels out there.

Then their in-house coders can do the presumably more "interesting" stuff like write drivers for the newer hardware (esp pre-release hardware - in the initial stages you might end up having to change specs, after release you can send it to the open source bunch).

Re:Uh, Via, makes gfx cards? Why that is NEWS to m

[Super Jamie]

they're mostly for onboard video chipsets, and this is awesome news for integrated devices and lightweight PCs like media centres, internet kiosks, settop boxes, netbooks, etc etc etc

simply the fact that one of the largest video chipset manufacturers in the world is writing open source drivers is huge, and an awesome step forward for linux and foss in general

not everything related to the phase "video card" is about pcie cards in sli and their crysis benchmark

Now show them why OSS is good

[slimjim8094]

As an act of faith, we should build something cool out of this - not to mention promote them to non-gaming computer users.

If we can optimize a graphics driver or do new things with it, they can sell more hardware and everybody wins. God knows ATI isn't making any money off of their drivers.

Hopefully we can use this to drive the point home.

Re:Now show them why OSS is good

[Jorophose]

But the question is, what?

Re:Now show them why OSS is good

[carlmenezes]

Well, I think a simple start would be to come up with a nice polished compiz theme and desktop (like a good avant dock with some nice icons) that uses this driver to its fullest. We are now at the point where a Linux Desktop can look as good as, if not better than, Windows or the Mac.

Give the average Joe Bloggs a PC running Linux that is relatively immune to viruses and auto-updates Firefox, Flash, Java, GNOME/KDE and VLC when its not being used and you have one happy computer user.

Build computers that use VIA chipsets for all the family that you run tech support for and lets start driving Linux adoption up! The drivers are here.

Re:I found VIA sub par!

[Mad Merlin]

The only problem on my part is that I find VIA products mediocre when it comes to gaming.

Well, what do you expect from an integrated video card? They're hardly speed demons.

Arrghhhh

[TheModelEskimo]

So when I bought my Dell Ubuntu laptop last year, I thought, "Intel and nVidia are the LEAST evil of the graphics chipset manufaacturers." Wanting a little more oomph, I went with nVidia.

Now, a year later, nVidia is looking ridiculous by clinging to closed-source binary drivers while the rest of the industry (including ATi, for pete's sake) go open. And the fact that freaking VIA is more open than nVidia really makes me feel...frustrated. Sorry nVidia, but I can't recommend you as long as you lag like this.

Re:Arrghhhh

[Yfrwlf]

Yes, because sharing source code so that you can get extra help from the rest of the world so you can all work together on software is horrible and wasteful and will produce crappy software and is way too expensive for any company to do.

Oh wait, there are several companies doing that already, never mind.

I buy graphics cards for their hardware, and I expect the software to utilize the hardware as best it can, and if anyone can help with that and with fixing bugs etc then all the better.

On the specific point of arguing "IP" politics though, do you honestly think the world has better graphics hardware right now because of the closed nature of graphics drivers? Because guess what, it's usually competition which spurs the development of better technology, competition which drives innovation in the world, so to tell me with a straight face that without the secrecy and closed nature of Nvidia's and ATI's graphics drivers, graphics technology would be further behind than if it were more open and there was more competition for making better hardware instead of screwing around with driver secrecy, that'd be a feat. I believe that most all patents and secrecy now days is nothing but harmful. In a world that's so inter-connected, there are very few examples I can find for justifying monopolies on ideas. They most always serve only to make the rich richer and poor poorer. (See Microsoft's patent FUD, for example, and try to tell me that did any good for the rest of the world.)

Re:Arrghhhh

[Anonymous Coward]

  The whole argument for FOSS 3D video card drivers is just silly in my opinion. Very very very few people have the skills necessary to write good drivers for these chips (others can learn, but that takes months or years to do). The people who write these drivers do it as a full time job, and the drivers are some of the most important IP in a graphics card (if they were released under a gpl like license, it would be much easier for a new competitor to develop a product).

What about the SuSE radeonhd developers? They work full time. You speak as if programming 3D graphics is rocket science. It is rocket science, if you don't have the specs. Otherwise, it would have been done YEARS ago.

Re:Arrghhhh

[shermozle]

And in five years' time, when they've stopped supporting your card in the latest kernel version, you do what?

Re:Arrghhhh

[Jah-Wren Ryel]

Good for you. For me, I dropped $600+ on two cards in order to drive 4 monitors - all based on this supposedly great support they had for linux.
The drivers didn't effing work and the 'support' was completely worthless, little better than "did you plug in the cable" level.
I had to pay another ~$200 for two gefen "dvi doctors" in order to fix an obvious bug in nvidia's driver, a bug I could have fixed myself faster than it would have taken to recompile the drivers if I had source.
Three years later, their drivers still lag without full support for randr.

Your personal experience doesn't mean shit.

Re:VIA stuff doesn't support 720p or 1080i

[Jorophose]

Wait, what?

NTSC ending doesn't mean we'll all be watching 720/1080. It means everything is digital, MPEG2 streams. We're all a looong way off from HDTV-to-the-door.

Their chipsets can certainly do 1080p. Look at the CN400.

never forget quack.exe

[r00t]

It was found that renaming quake.exe to quack.exe
would affect performance. The reason is that the
driver purposely degrades the quality for stuff
that is used in benchmarks. This is dishonest, and
it is a filthy hack. It's damn obvious why video
drivers are a major cause of crashes; they dig
around in kernel memory (totally undocumented) to
enable dirty hacks.

Open Source fixes this problem automatically.

Re:Might already be there, depending on purpose

[DrSkwid]

wtf are you on about, why would a firewall need a fast blit ?

I would guess you don't know shit about what a vga card does.

Re: Matrox

[Lonewolf666]

On top of that, they fell behind badly in terms of performance, and the great signal quality from their cards is mostly meaningless in the age of DVI.

Looks almost like a case of corporate suicide, as in "nobody can be THAT stupid, so it must be intentional" ;-).

Re:Might already be there, depending on purpose

[the_B0fh]

I hate to say it, but do you understand the concept of a firewall? You know, a hardened box running the minimum software necessary to inspect and pass/stop traffic?

Typically, it does not include a gui for a pretty interface.

Just saying.