Nouveau Advances For OpenGL 3.2 Support

[curaga]

1. Buy 1000 Russian identities in bulk, make sure they are spread evenly and unrelated.
2. Buy 1000 signing certs.
3. Even if they close one, no prob.
4. ???
5. PROFIT

[dashcloud]

Well what the fuck. I thought they had an actually secure setup where MS vetted every kernel mode signing cert. If you can just buy one for 200$, that's just asking for rootkits.

I'm sure they do vet it- which is why it's kind of pointless to try getting a cert if you plan to do malicious things. You still have to provide information, and if you're going to try to convince the certificate authority that you're real when you're not, might as well just borrow someone else's or do something that doesn't require a signed cert.

MS can blacklist your cert, and it'll probably be a big news story if an issued code signing cert turns out to be malicious and not stolen, which will further limit the damage that can be done.

[curaga]

Well what the fuck. I thought they had an actually secure setup where MS vetted every kernel mode signing cert. If you can just buy one for 200$, that's just asking for rootkits.

[V10lator]

That's interesting, how did they get a key? What purpose would ReactOS making drivers for genuine Windows have?

They stole it: http://support.microsoft.com/kb/293818

No, seriously, I guess they just paid for it. Verisign/Symantec are selling them.

[curaga]

That's interesting, how did they get a key? What purpose would ReactOS making drivers for genuine Windows have?

[dashcloud]

You can easily make such a Windows driver be GPL by making your changes be GPL. I guess the issue of driver signing is worse, IIRC MS doesn't allow unsigned drivers from Vista onwards without boot code trickery. MS wouldn't sign a driver for free, users wouldn't install something that opens their system to (more) rootkits.

Theoretically, if you had the driver built and working, you could get ReactOS to sign it.

[curaga]

maybe thats a big reason nobody has realy made (opensource) windows support to the current gallium codebase, because is signifigantly more likely it would be used in a propriatary driver? there wouldnt be a lot of motivation to just make that i guess.

You can easily make such a Windows driver be GPL by making your changes be GPL. I guess the issue of driver signing is worse, IIRC MS doesn't allow unsigned drivers from Vista onwards without boot code trickery. MS wouldn't sign a driver for free, users wouldn't install something that opens their system to (more) rootkits.

[bridgman]

i know amd made a propriatary driver for something running windows.

i guess this is allowed because the gallium code is X11 liscence and not gpl.

maybe thats a big reason nobody has realy made (opensource) windows support to the current gallium codebase, because is signifigantly more likely it would be used in a propriatary driver? there wouldnt be a lot of motivation to just make that i guess.

Not AMD -- we defined our hardware layer a few years before Gallium3D was introduced. The open source drivers use Gallium3D but not the proprietary ones...

[benjamin545]

i know amd made a propriatary driver for something running windows.

i guess this is allowed because the gallium code is X11 liscence and not gpl.

maybe thats a big reason nobody has realy made (opensource) windows support to the current gallium codebase, because is signifigantly more likely it would be used in a propriatary driver? there wouldnt be a lot of motivation to just make that i guess.

[smitty3268]

as far a gallium3d goes, what i am really interested in seeing is the other end of the pipe if you will. we have a lot of focus on getting the hardware support tied into gallium, but one of the key selling points when it was first announced was the easeof OS and API interoperability. i know we keep having the discusion of the dx9 and dx11 state tracker come up form time to time, and people keep getting excited about it as a linux api (for some odd reason) but what id realy like to see is a windows driver for a GPU that is entierly Gallium based. when Gallium was first concieved much of the documentation made it seem like the devs were designing it to be super easiy to be implemented as a windows driver to begin with, with the dx9 like IR and the seperation of componants, as if if things had gone differently maybe intel or powerVR would have hired them to write drivers for thier hardware for windows. of course this was all 5 or so years ago when we werent so sure of androids dominance on the non apple phone(and other portable device) market and there was a very good chance windows would have been a relevent mobile os, hell, we didnt even know that netbooks would take a nosedive like they did, there was a lot of speculation that we were going to have an abundance of arm+mobile gpu netbooks around.

so it would be nice to have the infanstructure in place for windows support for gallium3d if only to help entice GPU hardware vendors to select it as a base for thier future drivers. i woult think this would be very beneficial to alll the vendors out there, especialy the mobile gpu vendors who are starting to make some very impressive pieces of hardware that may soon be worthy of a desktop introduction. choosing gallium to base your hardware driver on would largely be a win win as you would get good cross platform integration with your driver and the community would get... a good cross platform integration.

or at the very least i would think reactos would apriciate it.

Windows gallium drivers do exist, they are just all proprietary. They also target Direct3d (10)