Nope. You're focusing on NVidia and still missing the key part of the analogy, which is that if you use Intel HW (or MIPS CPUs) you get the binary microcode (yoghurt dressing) automatically and you don't have to add it yourself. If dependency on binary microcode was the problem then FSF would have to remove their recommendations for most of the hardware out there -- so it's only the "adding dressing yourself" (including microcode in the distro images) that is considered a problem.
If you look at Intel or NVidia separately it's possible to convince yourself there is a logical explanation, but when you look at them together the logic falls down.
Anyways, there's a reason people don't use dietary analogies when discussing technical issues.
Announcement
Collapse
No announcement yet.
Nouveau: NVIDIA's New Hardware Is "VERY Open-Source Unfriendly"
Collapse
X
-
Originally posted by bridgman View PostI guess the most accurate analogy would be :
- our vegetarian restaurant serves salads with yoghurt dressing on the side, so you can add it to the salad yourself
- a second restaurant serves the same salad with the same yoghurt dressing, but the salad and dressing are pre-tossed
- the steakhouse down the road has the oil and vinegar dressing you prefer, on the side, but you have to bring your own salad greens (or, as you say, this could be the "bland salad")
An even more accurate analogy could be:- A salad bar which sells two things: 1. lettuce 2. a mix of other salad ingredients and yoghurt dressing
- A steakhouse which sells lettuce (and non-vegetarian dressing), but community members walk around the place giving you a mix of vegetables and vegan dressing (reverse engineered from the non-vegetarian dressing) to go with that lettuce. However there is a recently opened new section of the steak house which the owner has locked down so community members cannot access it (what this Phoronix article is about).
So as a vegan you have the choice: Go eat only lettuce at the salad bar, or go to the steakhouse and enjoy a salad with a variety of ingredients.
If we now put
Unaccelerated graphics = lettuce
3D acceleration, power management, etc. = other salad ingredients
People insisting on free firmware = vegans,
People ok with proprietary firmware = lacto-ovo vegetarians,
FSF = the vegan association,
AMD = salad bar owner,
NVidia = steakhouse owner,
Putting firmware in ROM = growing meat on trees
Then the analogy mostly fits and explains nicely why the FSF cannot endorse AMD graphics.
Leave a comment:
-
Originally posted by bridgman View PostYep. I keep coming back to the conclusion that there was a necessary clean-up because a lot of the microcode did not have licenses with appropriate redistribution rights, and that we just got caught up in the enthusiasm. Collateral damage, if you will.
The NVidia microcode wasn't RE'ed because of an overwhelming desire for freedom, it was RE'ed because NVidia did not provide the microcode with a license allowing redistribution. On the other hand I spent time with distro & AMD legal folks to work up a license that *did* meet redistribution requirements, so NVidia HW gets recommended and ours does not.
Madness.
However, I do not like all this DRM/corporate bullshit (unless DRM stands for Direct Rendering Manager :P). So, AMD looks like "least evil" option for those who wants powerful GPU wthout huge proprietary BLOBs on main system CPU. However, what you've told basically means AMD still considers users as their enemies. And while AMD is least evil option, there is heck a lot of room for improvement. And one shouldn't be surprised AMD doing something nasty, because it turns out they're still DRM-inclined and would not not mind screwing users up. That's what I can see from your words.
In fact hardware these days is something similar to what software has been in 90s - I can only choose from proprietary, hostile things, where vendor dares to apply DRM dictatorshit after getting my money. I consider it fundamentally wrong and would warmly welcome new competing forces and some REALLY open hardware (where full documentation provided, no blobs without sources, no treachery and no DRM shit). And it seems to be the only proper long term solution. I'm really fed up with all this DRM, locked boot loaders & other anti-user shit.
Leave a comment:
-
I've actually seen this during animal rights campaigns
Originally posted by boltronics View PostYou people have absolutely no idea what you are talking about. I'm a purist, but don't force my views on anyone. That happens automatically out of convenience as per my above example (just like the people who use proprietary software encourage other people to use it out of convenience). Just drop the whole vegan comparison if you don't comprehend it.
The FSF similarily needs to encourage people to use the free alternative whenever it exists, without taking the risk of hobbling free software to the point that people say "it doesn't work, so I am going to install Windows." THIS is the sort of purism being discussed here. Keep in mind, the freest hardware from the "free beer" perspective is that which people already own, and that will be random for most people installing Linux for the first time. If someone has a laptop that cannot connect to wifi without a proprietary driver blob and ndiswrapper, they are still MUCH better off than if they ran all of Windows instead of just the Windows wifi driver. For instance, the FBI CIPAV used in the Freedom Roads case would not run on a laptop with Ubuntu, the Nvidia blob, and a Windows wifi driver under ndiswrapper. It required a full Windows stack to run, because that's what most of the targets were presumably running.
The Guardian journalists communicating with Mr Snowden, by comparison, required an air-gapped machine that had been bought with cash and never been connected to any network for decrypting the emails, and a separate networked machine for receiving them. In this kind of situation, Trisquel over Coreboot for both machines reduces the attack surface to just the unmodified firmware blobs in permanent flash and direct hardware exploits.
Finally, in the cases of machines with no known firmware blob on disk requirements, it can be interesting to strip them all out and see what happens. Serious efforts should be made to grow the list of machines that will work this way.
Leave a comment:
-
EDIT: I think AMD identifying undocumentable technologies in use currently and developing documentable OSS friendly alternatives would give them a remarkable opportunity. That is, they can scale down. If they can scale it down well, it would have more room for scaling up.
I think IC development history has proved that time and time again.
Leave a comment:
-
Originally posted by Veto View PostBeing an engineer, I must agree with Bridgman, that the lines between software and hardware are so blurred that the "vegan" idealist view of Free Software does not make sense. There is currently no such thing as a truly free and open computer platform! Sorry to say: it is not "turtles all the way down"...
It basically boils down to:
- How well defined and documented is the "Closed Hardware" and the boundary to our "Free software".
- How far down the stack is this boundary.
- How much are we (the users) in control of changes to the underlying "hardware" (e.g. in the form of firmware updates).
- How well documented are the changes to the underlying "hardware".
IMHO AMD fares much better than most comparative vendors in these criteria. The ROM/RAM distinction is ridiculous and I would personally rather have a distributable firmware blob than a costly Flash chip.
However, most vendors are seriously bad at documenting changes to their firmwares, effectively depriving the users of the capability of making an informed choice with regard to firmware updates. Furthermore we have too little control to effectively ensure the firmwares are not updated against our will, making these firmwares a security nightmare.
The thing is reality and ideology don't match in this case. The way I understand things is like this; ATi and then AMD have licensed technology implemented in there products. Such that they can't release documentation for it. What they are able to do is release abstractions that expose that licensed capability through microcode firmware.
So this licensed technology usually has a hardware component and a software component. The hardware can be a processor or microcontroller or a fixed function controller. The software can be an interface of some kind, like a compiler or interpreter or library interface or whatever. GPUs are complex devices and there are tons of instances where documentation isn't possible. Ideologically speaking, you'd need documentation for the hardware implementation and the software implementation needed to use it.
Well, that's just not gonna happen unfortunately.
To bring reality closer to ideology AMD would need to design it's future GPUs with OSS support in mind for both the hardware and software.
Leave a comment:
-
Being an engineer, I must agree with Bridgman, that the lines between software and hardware are so blurred that the "vegan" idealist view of Free Software does not make sense. There is currently no such thing as a truly free and open computer platform! Sorry to say: it is not "turtles all the way down"...
It basically boils down to:
- How well defined and documented is the "Closed Hardware" and the boundary to our "Free software".
- How far down the stack is this boundary.
- How much are we (the users) in control of changes to the underlying "hardware" (e.g. in the form of firmware updates).
- How well documented are the changes to the underlying "hardware".
IMHO AMD fares much better than most comparative vendors in these criteria. The ROM/RAM distinction is ridiculous and I would personally rather have a distributable firmware blob than a costly Flash chip.
However, most vendors are seriously bad at documenting changes to their firmwares, effectively depriving the users of the capability of making an informed choice with regard to firmware updates. Furthermore we have too little control to effectively ensure the firmwares are not updated against our will, making these firmwares a security nightmare.Last edited by Veto; 03 May 2015, 11:20 AM.
Leave a comment:
-
-
Originally posted by boltronics View PostAs somebody who's been vegan for years, that sir is utter BS. I have had to go to a pub with colleagues for a work thing, and grilled the waiter about the sauce (which after the waiter checked turned out to not be vegan).
So +1 for finding another analogy. Maybe we can do cars next, that never works eitherLast edited by bridgman; 02 May 2015, 11:30 PM.
Leave a comment:
-
Originally posted by Luke View PostIn this way the FSF is like an "Vegan purist" (used to theory books instead of front lines at protests) who hassles a semi-vegetarian to the point that they throw in the towel and go back to eating meat
Leave a comment:
Leave a comment: