Eric Anholt Leaves Intel's Linux Graphics Team For Broadcom

[_ONH_]

IN theory they would be kind of comparable to ati, becuase they also depend on loading a driver firmware blob at boot which is not opensource. If you ask richard stallman thats no full free driver. I dont know if you could enforce that with the gpl and a law suit, I am no laywer and I dont know every single word of the gpl out of my mind.

Jea, as I understand his words on firmware, it is complete bs. Firmware on hw rom is ok, but the same fw from hdd into ram it isn't ok. But in reality as long as the fw blob is available on a independent server licensed appropriate, that it doesn't vanishes from availability. There is no difference at all for the dumb user, and therefor should not be treated differently.

[blackiwid]

Jea, as I understand his words on firmware, it is complete bs. Firmware on hw rom is ok, but the same fw from hdd into ram it isn't ok. But in reality as long as the fw blob is available on a independent server licensed appropriate, that it doesn't vanishes from availability. There is no difference at all for the dumb user, and therefor should not be treated differently.

I am tired of arguing pro or contra to that point, I am undesided. I buy myself amd hardware, so I have no real problem with it, at least till nobody else has so good drivers. And intels solutions are not always better maybe the driver is 1-2% better but then the amd gpu hardware is 100% better so even with a not so good driver maybe (at least they have gallium3d driver as their main driver so here they gone further done the road to go), it still is a better package. And I kind of not buy to much intel stuff (even I consider a intel-tablet now) has also to do with intel beeing nearly a x86-cpu monopolist and often also intel is much more expensive.


So I dont see that as such big problem too. But I am slow to say rms has not good reasons to see that that way. And even if he would be wrong at that small point, is that very relevant, is everything else he says wrong because of that?

Generally I think even all firmwares and bioses should be "forced" to opened up, and only that is true freedom of software-users. So if you make small compromises to that pure idea, its hard to draw the line clearly.

But that rPi situation makes me more likely to agree to his point. Because if you think prakticaly, if they would have put the firmware on a rom, they would never had put all the functionality into the firmware. They would have made it as small as possible most users dont do firmware updates.

So in a way it at least allows companies to do more evil stuff, and take from users more control, when u allow them to flash the firmware. And that again is all what the gpl is about.

It never said that every proprietary developer to 100% do implement evil antifeatures against u, maybe they just go bankrupt and nobody can fix the stuff in that code. So having a peace of software that needs much patches as proprietary software on your machine is more problematic than a rom that the vendor must be shure that it will work at least for most people good enough so they will buy someday another peace of hardware from this company, and they dont get extremly bad press.

its hard to put it into a clear reasoning short sentence why loadable firmware blobs vs fixed firmwares are bad, but there are some reasons that kind of make a bit sence.

Another thing is u have to burn the proprietary blob onto your distro-cd and bundle here proprietary and free software. So a gnu/linux group have to more or less advertise non-free software.

that are at least all differences to having a firmware in a rom or something, even if its flashable.


The point is, it does not really matter here, because every hyper-pragmatic opensource gpu-driver developer called THIS driver/firmware bullshit. Because they used that "trick" to its worst possbile version, by protecting as much code as possibel to openend up by hiding it in a place where it technicaly makes no sense.

[curaga]

Well for starters more companies hiring Mesa dev's shows the health of the project and represents more financial backing, this is generally always a good thing. I totally disagree that this creates low morale or an unhealthy environment. As a developer my morale would be boosted knowing that there are more job opportunities around, that I could get a pay raise if my current company wasn't going to give me one, or that in fact they might give me one to keep me. Also the comfort in knowing that if intel were to cut back on Mesa development that there are other companies that I could go work for is great peace of mind.

As far as poaching goes its a good thing to allow developers to be paid what they are worth. Not allowing poaching is an antitrust issue which google, apple, ebay, etc are all in trouble over currently.

Active poaching incentivizes people to only work on the highly shiny, externally visible parts. Once it grows big enough, there ceases to be any loyalty to the employer no matter what they pay, causing project quality to suffer. It leads to a situation like that in India, where in a one-year project at the end there is nobody present who was there at the beginning. Nobody knows the code, they all left, and quality is in the shitter.

If person X happened to be the only one who knows a specific area (a very common thing in FOSS), the users can say goodbye to any work in that area if they switch companies. Any new person will take a while to get to know the codebase to work on it in sufficient quality, and once they leave, the cycle repeats.

Active poaching makes that cycle faster, a lot faster. It's in nobody's long-term interest, either, not even the switcher's, even though they may get a higher salary. Should the situation grow to the extreme, as in India, a lot of companies will simply quit. You can only game the system so far before the only winning move becomes not to play.

[tarceri]

Active poaching incentivizes people to only work on the highly shiny, externally visible parts.

This will generally happen regardless of poaching.

Once it grows big enough, there ceases to be any loyalty to the employer no matter what they pay, causing project quality to suffer. It leads to a situation like that in India, where in a one-year project at the end there is nobody present who was there at the beginning. Nobody knows the code, they all left, and quality is in the shitter.

Open Source devs are generally passionate about the code they write, more so on average than those working on closed projects. Pride in ones work will always be trump loyalty to any employer. General advice for anyone in IT is to work 2-3 years in one job unless you want to be doing it for the rest of your life.

If person X happened to be the only one who knows a specific area (a very common thing in FOSS), the users can say goodbye to any work in that area if they switch companies. Any new person will take a while to get to know the codebase to work on it in sufficient quality, and once they leave, the cycle repeats.

Having someone continualy work in a specific area is a good way to close the mind, and decrease morale.

Active poaching makes that cycle faster, a lot faster. It's in nobody's long-term interest, either, not even the switcher's, even though they may get a higher salary. Should the situation grow to the extreme, as in India, a lot of companies will simply quit. You can only game the system so far before the only winning move becomes not to play.

How is higher pay, the chance to lead a project and the opportunity to increase ones knowledge and skills via working with new people and systems not in the long term interest of the switcher? Your talking about programmers as if they are purpose build robots, and lets not get carried away we are talking about one developer "maybe" poached after many years at the same company, to get back to my original point an increased backing of Mesa work is a good thing in my opinion.

[tarceri]

I hope he will use Gallium this time.

Looks like it will be. At least to begin with.

I'm probably going to start out doing a gallium driver for simplicity,
to avoid having to do all the DRI crap we've got in brw_context.c and
texture miptree validation and getting user data into VBOs and all that
other almost-boilerplate. Long term I may end up switching to classic
so I can get swrast fallbacks and the ability to implement single-copy
manually-tiled TexImage uploads like. For now I want to get to drawing
triangles as soon as I can.

http://lists.freedesktop.org/archives/mesa-dev/2014-June/061692.html

[curaga]

This will generally happen regardless of poaching.

You miss my point. It will happen even if the employer orders to work on the plumbing.

How is higher pay, the chance to lead a project and the opportunity to increase ones knowledge and skills via working with new people and systems not in the long term interest of the switcher? Your talking about programmers as if they are purpose build robots, and lets not get carried away we are talking about one developer "maybe" poached after many years at the same company, to get back to my original point an increased backing of Mesa work is a good thing in my opinion.

You again miss my point. If poaching gets too high, companies will quit. Do you not understand where that leads?

If all companies around Mesa quit, that means that the switcher can no longer find work in the area they are qualified in. That's bad for them. That's also bad for the economy, the users, and Mesa.

[tarceri]

You miss my point. It will happen even if the employer orders to work on the plumbing.

You again miss my point. If poaching gets too high, companies will quit. Do you not understand where that leads?

If all companies around Mesa quit, that means that the switcher can no longer find work in the area they are qualified in. That's bad for them. That's also bad for the economy, the users, and Mesa.

I do get your point, I just dont agree with you. Your looking at the absolute worst case scenario, I'm a glass half full kind of guy myself.

[daniels]

1. the reason for the workplace change, he not even claimed but it could look like he searched a new interesting challenge.
I woul doubt that, the reason I would bet my money on is that, most of the time you cant get a higher salary without changing the employer, so he switched to broadcom.

I find it strange that intel really pays not as good as broadcom, but it seems thats the case.

speculation, much?

4. and even worse, devoloper have to invest even more time to make a good driver with this hardware, becuase they have to make workarounds around that bad gpu design, to make suff like gallium3d or even stupid normal x-drivers work with that hardware.

and again - you're just making stuff up.

[daniels]

How so? If there indeed is dev poaching between mesa-contributing companies, that affects morale and creates an unhealthy environment. I don't see how it's a good sign for the future.

which is more speculation ...

[blackiwid]

speculation, much?

No that is rational thinking. Are open-source-driver-developers better persons than all the other developers? I doubt it. So thats the normal game u learn u cant raise your income by much if you are not willing to change your employee. You have to at least threaten to do so. And even if you self dont search active for such job, they call u on the workplace if you are interesting and allure u with higher wages.

Thats just how it is. ok another posibility would be that intel is a toxic workplace and he did flee.

In reality searching a specific other work-place because u like the task there better happens not very often. Especialy if he does the exact same thing he did before.

Another point could be that he was on intel only one of many developers and want by broadcom to be the lead-developer but again that has something to do with money.

Of course thats speculation, but the changes that I am right here are very high, if anything I learned in my live about how business works is true.

and again - you're just making stuff up.

maybe u should not anwer to such old comments then, I dont make stuff up I speculate with the information I have. But here my information I had saved in my brain was wrong. I thought that their gpu only understands some high level commands like opengl and its the gpu design, but it seems they just made a horrible firmware on top of a normal gpu. So it should be fixable.

Point is, I got this information becuase I wrote here down my wrong thesis/understanding. So beeing wrong doesnt mean making something up, making something up would mean that I have no information and just claim something. But in fact I did read all news about the rpi and I even regerged about it because I even thought to buy one at some point, so I thought I had the real understanding so I made a total right statement on my information. The problem was only that I did understand this architecture a bit false.

So showing my thoughts and what information it is based + some people that corrected me made me learn something. So I think thats a very effective way to get your facts straight.

BTW, at least the video-encoder seems to be closed up, you need to pay them money to unlock other codecs and I guess they want to keep it that way, so again I feel their woudl be better socs/gpus that would be better targets. Like I said mali and so on are way more used gpus.