Unigine Engine Advances, But No Linux Heaven Yet

[Melcar]

Did you not read the "way to broken" quote from Denis?

Were are on 10.2 and on to 10.3. The guy was talking about 9.11/9.12 (and most likely referring to support found on earlier releases still). The drivers could have gotten better by now, or not. Point is that the quote is old and you really can't use it now; a lot in the driver has been fixed and even broken since those releases. I want a statement from either AMD or Unigine that says that as of now the drivers are too broken so Heaven on Linux has been delayed until further notice.

[deanjo]

Were are on 10.2 and on to 10.3. The guy was talking about 9.11/9.12 (and most likely referring to support found on earlier releases still). The drivers could have gotten better by now, or not. Point is that the quote is old and you really can't use it now; a lot in the driver has been fixed and even broken since those releases. I want a statement from either AMD or Unigine that says that as of now the drivers are too broken so Heaven on Linux has been delayed until further notice.

Considering that Phoronix is running the engine internally, makes mention of "Unigine Heaven on Linux internally and it's a beautiful tech demo / benchmark to say the least even without a bug-free tessellation experience. " and the only hardware out there right now is ATI's offerings then the mention of "We don't yet know whether the Linux build will make its public debut at that time or whether ATI Catalyst 10.3 for Linux will finally be the golden egg for Unigine Corp, but the Unigine Engine does continue picking up new features. " It doesn't take a rocket scientist to figure out what the delay is. To conclude anything but ATI holding up the release is wishful blind thinking.

[mirv]

Considering that Phoronix is running the engine internally, makes mention of "Unigine Heaven on Linux internally and it's a beautiful tech demo / benchmark to say the least even without a bug-free tessellation experience. " and the only hardware out there right now is ATI's offerings then the mention of "We don't yet know whether the Linux build will make its public debut at that time or whether ATI Catalyst 10.3 for Linux will finally be the golden egg for Unigine Corp, but the Unigine Engine does continue picking up new features. " It doesn't take a rocket scientist to figure out what the delay is. To conclude anything but ATI holding up the release is wishful blind thinking.

But that's just the point - all of that information comes from Phoronix, not from Unigine.
I actually find it strange that AMD would try to hold up releasing the demo based on that they are the ones with a hardware tesselator out right now - and are the ones with an OpenGL extension for it. This particular extension has been around for some time now, and I find it also strange that Unigine are having issues with it - maybe they don't use it, as then it becomes more of an AMD specific demo (geometry shaders could be used instead). Or maybe they really are working with nvidia to have both AMD and nvidia code paths (for however nvidia are going to do things).
I remark upon the tesselator as it would appear that the demo would otherwise be ok to be released - but again information there is vague at best, so who knows.
AMD could well be holding it up for a future driver release, but I wouldn't blindly take the phoronix article's word for it. I'd rather draw my own conclusion from something a bit more official.

[deanjo]

But that's just the point - all of that information comes from Phoronix, not from Unigine.

So what happens when you read a newspaper? Do you believe it's all rubbish until the first party admits it first hand?

I actually find it strange that AMD would try to hold up releasing the demo based on that they are the ones with a hardware tesselator out right now - and are the ones with an OpenGL extension for it. This particular extension has been around for some time now, and I find it also strange that Unigine are having issues with it - maybe they don't use it, as then it becomes more of an AMD specific demo (geometry shaders could be used instead).

Given AMD/ATI's openGL history there really isn't anything surprising here at all. If you need another example take a look how long it took them to show off UVD in linux.

[mirv]

So what happens when you read a newspaper? Do you believe it's all rubbish until the first party admits it first hand?

Newspapers often do quote people first hand. Newspapers also have hot gossip sections where they make stuff up or just like to push their own opinion.

Given AMD/ATI's openGL history there really isn't anything surprising here at all. If you need another example take a look how long it took them to show off UVD in linux.

Actually AMD's history with OpenGL is pretty good (since the catalyst program started up) - as long as you stick to spec. And they've been getting remarkably better for when you don't. Which has absolutely nothing to do with UVD.

Really, I'm all for things pointing at AMD's drivers, but the article was an opinion dressed up as news; if that's the new direction of Phoronix, I'd suggest looking a bit more at how semiaccurate presents such things.

[deanjo]

Newspapers often do quote people first hand. Newspapers also have hot gossip sections where they make stuff up or just like to push their own opinion.

There is also something called investigative journalism which rarely has the co-operation of the parties involved.

Actually AMD's history with OpenGL is pretty good (since the catalyst program started up) - as long as you stick to spec. And they've been getting remarkably better for when you don't.

Not sure what time line or universe you are looking at but ATI's history of ogl support has always lagged well behind the official released spec.

Which has absolutely nothing to do with UVD.

Never said it had anything with ogl. Just used it as another example of an item that has been present in the drivers for a while but took them ages to actually introduce.

Really, I'm all for things pointing at AMD's drivers, but the article was an opinion dressed up as news; if that's the new direction of Phoronix, I'd suggest looking a bit more at how semiaccurate presents such things.

The only thing presented as an opinion is your speculation. The article goes by past and present issues that are verified by the reporter first hand.

[Melcar]

Still, there are very little facts given to the story. It's mostly conjecture. The delay could be AMD, but it could very well be due to something else.

[deanjo]

Still, there are very little facts given to the story. It's mostly conjecture. The delay could be AMD, but it could very well be due to something else.

Official word.

Hello Dean,

Right now no, it isn't.

--
Denis Shergin
CEO / Unigine Corp.
binstream@unigine.com
tel.: +73822553458 (office)
tel.: +79138250566 (mobile)

On 08.03.2010, at 1:59, Dean <deanxxxxx@gmail.com> wrote:

Quick question, is it the ATI drivers for linux that is holding up the
release of Heaven on linux?

Dean

[Kano]

Is it Fermi then

[deanjo]

Is it Fermi then

Lol, I'll let you send in that email.