RADV: A Community Open-Source Effort To Get Vulkan Working On Radeon

[duby229]

I don't see a difference. Hardware is hardware. You trying to differentiate openness and closedness. Tell that to the Chinese, they'll laugh at you all the way to the electron microscope.

[starshipeleven]

I don't see a difference.

Because you're stupid.

Hardware is hardware. You trying to differentiate openness and closedness.

Open hardware is hardware where full blueprints and firmwares are opensourced, closed hardware is hardware where blueprints and firmwares are closed.

Low-level firmwares are closed or open as the hardware they run, it makes no sense to do differently.

Also, it's not my own invention, it's here https://en.wikipedia.org/wiki/Open-source_hardware

Tell that to the Chinese, they'll laugh at you all the way to the electron microscope.

It's usually MUCH cheaper to buy such really complex stuff pre-made than making a copycat of hardware itself and then reverse-engineer the firmwares. It's not a matter of possible/impossible, but of cost.

Which is also why for example I go buy chairs from Ikea instead of chopping down a tree and making a look-alike chair myself from its wood, you know.

[duby229]

Because you're stupid.

Open hardware is hardware where full blueprints and firmwares are opensourced, closed hardware is hardware where blueprints and firmwares are closed.

Low-level firmwares are closed or open as the hardware they run, it makes no sense to do differently.

Also, it's not my own invention, it's here https://en.wikipedia.org/wiki/Open-source_hardware

It's usually MUCH cheaper to buy such really complex stuff pre-made than making a copycat of hardware itself and then reverse-engineer the firmwares. It's not a matter of possible/impossible, but of cost.

Which is also why for example I go buy chairs from Ikea instead of chopping down a tree and making a look-alike chair myself from its wood, you know.

Whether you thought up such nonsense or not is still nonsense. Hardware design is done just like any other design. There are application suites specially made to create the design, that application suite must be written with some language, the data it creates and saves must be stored in files. We aren't talking about new concepts here.

The point I'm trying to make is that there is no need to make such a differentiation. It's all the same thing whether you are talking about code that renders a 3d scene or code that operates a cnc machine. The code that renders the 3d scene is not the 3d scene itself, and neither is the code that operates the cnc machine the cnc machine itself. It's because the same terminology and concepts exist where physical or virtual. Openness is openness equally everywhere. Making some sort of artificial differentiation is stupid.

[starshipeleven]

Whether you thought up such nonsense or not is still nonsense. Hardware design is done just like any other design. There are application suites specially made to create the design, that application suite must be written with some language, the data it creates and saves must be stored in files. We aren't talking about new concepts here.

????? I'm talking of openly accessible hardware schematics of the fucking integrated circuits that are your CPU/GPU/whatever, under a license that allows everyone to make their own respin for free (and more importantly to contribute to its development, linux-like).

Open hardware is about that. OPEN SCHEMATICS, as the "source" of hardware is schematics, blueprints.

Just as open source software is OPEN SOURCE because the "schematic" of a binary is its source code.

The point I'm trying to make is that there is no need to make such a differentiation.

Oh no there is, with a open hardware system you know and control the whole system from hardware up, with opensource only you are a passenger on top of 2-4 layers of different closed firmwares that run the hardware and can decide whatever they please.

Openness is openness equally everywhere. Making some sort of artificial differentiation is stupid.

Just as software can be opensource or closed source, hardware can be too, because hardware has a source too, that is called "schematic" or "blueprint".
If this source is available to all then it can be defined open, if it isn't, it can be defined closed.

I'm just adding that low-level firmwares are technically software, but in this day and age are so connected to the hardware that you cannot have open firmwares without also open hardware.

[duby229]

????? I'm talking of openly accessible hardware schematics of the fucking integrated circuits that are your CPU/GPU/whatever, under a license that allows everyone to make their own respin for free (and more importantly to contribute to its development, linux-like).

Open hardware is about that. OPEN SCHEMATICS, as the "source" of hardware is schematics, blueprints.

Just as open source software is OPEN SOURCE because the "schematic" of a binary is its source code.

Oh no there is, with a open hardware system you know and control the whole system from hardware up, with opensource only you are a passenger on top of 2-4 layers of different closed firmwares that run the hardware and can decide whatever they please.

Just as software can be opensource or closed source, hardware can be too, because hardware has a source too, that is called "schematic" or "blueprint".
If this source is available to all then it can be defined open, if it isn't, it can be defined closed.

I'm just adding that low-level firmwares are technically software, but in this day and age are so connected to the hardware that you cannot have open firmwares without also open hardware.

Why do you insist on quoting out of context? Anyway I made my point. Firmware is definitely not hardware just as much as the schematics used to describe it also is not hardware. Openness is the same and artificially complicating things is stupid.

[libv]

*sigh* Don't you guys remember anything?

For Avivo/R500, ATI designed one of the most beautiful display engines i have ever seen, but a different part of ATI did the stupid firmware thing and tried to hide the relatively straightforward and not IP encumbered display/basic board bring up stuff behind firmware (which alex deucher and john bridgman later tried to remarket as "scripts"), with a special interpreter.

AMD had just bought ATI to go a centrino like full platform strategy, and then discovered that ATI was a useless badly managed barren desert with no documentation and really horrible software. When ATI then refused to support either r500 or r600 in their linux catalyst driver, some former SuSE people working at AMD contacted SuSE. I was lucky to be interviewing at SuSE at the time (after my modesetting insights turned out to be correct after all, and then some), and Egbert Eich, Matthias Hopf and I worked out a proposal to wrestle ATI free (which phoronix posted a few years ago).

ATI of course fought back. And lots of kindergarten stories ensued, but we created a proper C code driver nonetheless.

When we did get a driver out in time, ATI changed tac, and found some willing individuals to re-implement everything in the ATI way and then play some nasty power/political games on top. Some names of those who happily sacrificed technical excellence and software freedom for shitthrowing and firmware were: Dave Airlie, Alex Deucher, Daniel Stone, Adam jackson (these two later on vandalized the radeonhd git repository), Matthew Garrett, ...

On the SuSE front, with our mostly C driver (atombios support was demanded by bridgeman, much to my protest), we afterwards noticed a clear correlation between the boldness of the tactics of bridgman and his friends, and the political power of AMD over ATI, which was clearly related to revenue of the two "parts". And when AMD ran out of money due to the financial crisis, it's CPUs were making less revenue than its discrete GPUs, and ATI effectively ran the "reborn" AMD.

That's when C code died, that's when register level documentation died. The only thing that continued were the ISA documentation releases, but that was done by a separate GPGPU team created before the r600 was even released and way before AMD decided that it had enough of the ATI shenanigans and would create an open source driver.

It was almost down to ATI + Red Hat + noisemakers versus AMD + SuSE + big overnight community. Supposed open source "heroes" were actively weakening the major party which was pro open source software and supported that part of the AMD/ATI conglomerate that was always against open source software.

And now one of the nasty powerplayers is surprised that the organization that he helped shape to be far-less-than-free is not releasing code for a vulkan driver?

[artivision]

*sigh* Don't you guys remember anything?

For Avivo/R500, ATI designed one of the most beautiful display engines i have ever seen, but a different part of ATI did the stupid firmware thing and tried to hide the relatively straightforward and not IP encumbered display/basic board bring up stuff behind firmware (which alex deucher and john bridgman later tried to remarket as "scripts"), with a special interpreter.

AMD had just bought ATI to go a centrino like full platform strategy, and then discovered that ATI was a useless badly managed barren desert with no documentation and really horrible software. When ATI then refused to support either r500 or r600 in their linux catalyst driver, some former SuSE people working at AMD contacted SuSE. I was lucky to be interviewing at SuSE at the time (after my modesetting insights turned out to be correct after all, and then some), and Egbert Eich, Matthias Hopf and I worked out a proposal to wrestle ATI free (which phoronix posted a few years ago).

ATI of course fought back. And lots of kindergarten stories ensued, but we created a proper C code driver nonetheless.

When we did get a driver out in time, ATI changed tac, and found some willing individuals to re-implement everything in the ATI way and then play some nasty power/political games on top. Some names of those who happily sacrificed technical excellence and software freedom for shitthrowing and firmware were: Dave Airlie, Alex Deucher, Daniel Stone, Adam jackson (these two later on vandalized the radeonhd git repository), Matthew Garrett, ...

On the SuSE front, with our mostly C driver (atombios support was demanded by bridgeman, much to my protest), we afterwards noticed a clear correlation between the boldness of the tactics of bridgman and his friends, and the political power of AMD over ATI, which was clearly related to revenue of the two "parts". And when AMD ran out of money due to the financial crisis, it's CPUs were making less revenue than its discrete GPUs, and ATI effectively ran the "reborn" AMD.

That's when C code died, that's when register level documentation died. The only thing that continued were the ISA documentation releases, but that was done by a separate GPGPU team created before the r600 was even released and way before AMD decided that it had enough of the ATI shenanigans and would create an open source driver.

It was almost down to ATI + Red Hat + noisemakers versus AMD + SuSE + big overnight community. Supposed open source "heroes" were actively weakening the major party which was pro open source software and supported that part of the AMD/ATI conglomerate that was always against open source software.

And now one of the nasty powerplayers is surprised that the organization that he helped shape to be far-less-than-free is not releasing code for a vulkan driver?

A cake for Dave!

[pal666]

If it's a layout of 0's and 1's, then it's software.

everything is layout of 0's and 1's. now you won't be able to sleep lol

[duby229]

And that is exactly why I say openness isn't the issue, it's bout setting realistic goals that are achievable. One plan was achievable and the other was not even close. History has proven which was which.

[pal666]

Firmware itself is in fact software, it doesn't matter where it was written or how or when.

software running on servers around internet is also software. and most of it is closed, including this forum. somehow this does not preclude you from using internet, i find it very hypocritical