Announcement

Collapse
No announcement yet.

Fedora BIOS Boot SIG Launched For Those Wanting To Maintain Legacy BIOS Support

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #21
    Originally posted by jb.1234abcd View Post
    Well, your opinion is as true as mine. Except, we both got it upside down.
    I have no idea what you mean, nor see the relation between my opinion and the article you posted.
    I understand you were sarcastic with the Elvis comment, but other than that :shrug:

    Comment


    • #22
      Originally posted by You- View Post

      I think this point is often missed. The hobby distros and projects can often only function because someone is doing the heavy leg work. If those people or companies stop, their burden will become much higher or even impossible.
      True but at the same time Linux *did* exist in a usable form long before companies piggy backed off the work done by communities and individuals.

      If the individuals stop, the companies will not be able to keep it going themselves.

      If the individuals and community has to pick up the slack and work harder to maintain the old stuff (which they will always do); then it simply means that the companies and consumers will have to wait longer for the monetizable "modern" stuff.
      Last edited by kpedersen; 21 May 2022, 09:51 AM.

      Comment


      • #23
        Originally posted by kpedersen View Post
        True but at the same time Linux *did* exist in a usable form long before companies piggy backed off the work done by communities and individuals.
        Define usable.

        Originally posted by kpedersen View Post
        If the individuals stop, the companies will not be able to keep it going themselves.
        The companies already do most of the leg work, so I doubt it.

        Originally posted by kpedersen View Post
        If the individuals and community has to pick up the slack and work harder to maintain the old stuff (which they will always do)
        Which they most often don't.

        Originally posted by kpedersen View Post
        then it simply means that the companies and consumers will have to wait longer for the monetizable "modern" stuff.
        Such as? Name a few of the big things that didn't have company backing behind, please. Preferably things that were modern monetizable stuff.

        Comment


        • #24
          Originally posted by sinepgib View Post
          Define usable.
          Usability

          More usable than the competition back then. Otherwise we would be discussing Minix / UNIX wouldn't we?


          Originally posted by sinepgib View Post
          The companies already do most of the leg work, so I doubt it.
          Hardly. Reverse engineering their hardware is still commonplace. A great example is OpenHMD having to reverse all the proprietary crap from SteamVR / Oculus. And then the Monado OpenXR runtime uses it as the underpinning.

          OpenBSD is also another great example of an OS project that is almost entirely individual / community built. Very competitive in many ways. What legwork have commercial vendors done in the last 10 years?

          Originally posted by sinepgib View Post
          Which they most often don't.
          Check out the recent work on getting Linux working on an N64. You think a commercial vendor did this work? haha.

          There are many examples like this. Do you see many companies updating existing open-source operating systems just to run on very old systems including SGI's? Not likely.


          Originally posted by sinepgib View Post
          Such as? Name a few of the big things that didn't have company backing behind, please. Preferably things that were modern monetizable stuff.
          Some that I tend to use quite often:
          • Wireguard
          • GCC
          • OpenSSH

          Were all released to users before companies either sponsored or contributed more than the individual developers.

          You probably want to consider taking commercial companies off that pedestal you have put them on. In the long run it is the community and individuals that will take Linux forward (or produce a better open-source replacement once the companies have screwed up Linux).
          Last edited by kpedersen; 21 May 2022, 05:30 PM.

          Comment


          • #25
            Originally posted by kpedersen View Post
            1. Effectiveness—It supports users in completing actions accurately.
            2. Efficiency—Users can perform tasks quickly through the easiest process.
            3. Engagement—Users find it pleasant to use and appropriate for its industry/topic.
            4. Error Tolerance—It supports a range of user actions and only shows an error in genuine erroneous situations. You achieve this by finding out the number, type and severity of common errors users make, as well as how easily users can recover from those errors.
            5. Ease of Learning—New users can accomplish goals easily and even more easily on future visits.
            Just which of these were they meeting long before commercial distros? But even without snark, how do we define the users here? Certainly it was not a regular desktop user back then.

            Originally posted by kpedersen View Post
            More usable than the competition back then. Otherwise we would be discussing Minix / UNIX wouldn't we?
            Hadn't Linux had triumphed we'd still be talking about Windows and MacOS. Let's not forget it killed Solaris long before Red Hat existed, too.

            Originally posted by kpedersen View Post
            Hardly. Reverse engineering their hardware is still commonplace. A great example is OpenHMD having to reverse all the proprietary crap from SteamVR / Oculus. And then the Monado OpenXR runtime uses it as the underpinning.
            Which doesn't change a bit the scenario. Yes, not all companies contribute. No, that doesn't mean development statistics lie.

            Originally posted by kpedersen View Post
            OpenBSD is also another great example of an OS project that is almost entirely individual / community built. Very competitive in many ways. What legwork have commercial vendors done in the last 10 years?
            Fair example. Care to compare the amount of work put into it with that of Linux? Same for widespread usage.

            Originally posted by kpedersen View Post
            Check out the recent work on getting Linux working on an N64. You think a commercial vendor did this work? haha.
            You seriously talking of that as something relevant? It's a hobby, yay for whoever wrote that.
            If we're gonna talk about the community picking up useful stuff, how about PowerVR cards? VIA? Pretty much everything that gets dropped from Mesa due to lack of maintenance? Did the community pick those up?

            Originally posted by kpedersen View Post
            There are many examples like this. Do you see many companies updating existing open-source operating systems just to run on very old systems including SGI's? Not likely.
            No. But how many people does run those systems? I don't expect charity from companies. It's not a defense to companies but a criticism to the unexploited potential of "the community".

            Originally posted by kpedersen View Post
            Some that I tend to use quite often:
            • Wireguard
            • GCC
            • OpenSSH

            Were all released to users before companies either sponsored or contributed more than the individual developers.
            Fair. GCC is technically not monetizable tho, nor is OpenSSH. Both are end products that you'd get for free. But being so universally useful as they are, they count.
            WireGuard, OTOH, has a lot of direct commercial use in providing services and frontends.

            Originally posted by kpedersen View Post
            You probably want to consider taking commercial companies off that pedestal you have put them on. In the long run it is the community and individuals that will take Linux forward (or produce a better open-source replacement once the companies have screwed up Linux).
            There's no pedestal. I see the sheer amount of work, and simply there's not much a few hobbyists can do on their spare time compared to for profit companies. They are not goodwill and I certainly don't align with them.
            Said that, the community is, let's say, mostly less than eager to actually do work. Those who do, kudos to them, are really few compared to the most vocal idiots. You see how when someone says "let's drop this, I don't want to maintain it" in most cases _nobody_ steps up?
            All the scrollback buffer fiasco and what not. That's representative of most of the community.
            Say, again, VIA hardware. "The community" was about... Two guys? Three if we count libv, but he got fed up long before VIA was a figment of the past. That doesn't mean VIA wasn't a shitty company, it was and all we got from them was docs, which is still more than some other companies provide.
            But it's evidence the community isn't that all powerful thing you seem to believe it is. The reality is that everyone scratches their own itches. It just so happens that, for some companies, the resources they can spend to scratch theirs are just much greater than we can use our free time for, and most of the community isn't too eager to fund those who have the skills but no allegiances to companies. And some of those companies, having their businesses based on open source (specially Linux) or on selling hardware you can use with it, tend to itch a damn lot.

            In the end, I will do a tiny defense of companies in a way, tho: I'd much rather not be able to run Linux on the N64 than not be able to run it properly on my Intel laptop. Sorry, but I work on the latter.

            Comment


            • #26
              Originally posted by sinepgib View Post
              Just which of these were they meeting long before commercial distros? But even without snark, how do we define the users here? Certainly it was not a regular desktop user back then.
              It could have been. It was a clone of UNIX to an extent (Stallman and other individuals did a fantastic task of creating all the bits we see in coreutils today).
              It was "user-friendly" because it adhered semi-closely to the userland of UNIX. Compared to DOS, it was a far more pleasant system to develop on.


              Originally posted by sinepgib View Post
              Hadn't Linux had triumphed we'd still be talking about Windows and MacOS. Let's not forget it killed Solaris long before Red Hat existed, too.
              Exactly. It was doing well. Yes, there were some commercially backed things like SLS but it was the individual developers that were combining their skills and competing very well against the commercial products.


              Originally posted by sinepgib View Post
              Fair example. Care to compare the amount of work put into it with that of Linux? Same for widespread usage.
              These days it is a different matter. In many ways commercial companies are pushing the little guys out from contributing. No longer would some (very talented) 19 year old be able to start tinkering with the Intel vendor specific code and expect to have their contribution accepted. Back in the day this was obviously different because there was no Intel involvement.

              Originally posted by sinepgib View Post
              You seriously talking of that as something relevant? It's a hobby, yay for whoever wrote that.
              If we're gonna talk about the community picking up useful stuff, how about PowerVR cards? VIA? Pretty much everything that gets dropped from Mesa due to lack of maintenance? Did the community pick those up?
              The fact that the i.e older Intel GMA 965 and things like that works is due to individuals. Intel have been quite happy to drop support for these cards. They even broke OpenGL 2.1 for the older cards for a while.
              https://wiki.archlinux.org/title/int...L_driver_(i965)
              It is indeed the community that is keeping that stuff working.

              PowerVR and VIA are not massively common outside server-space (or shite phones). Think common hardware that the masses have/had.
              Such as things like the Voodoo (https://www.phoronix.com/scan.php?pa...item&px=ODQxMw). Yes, they did ultimately fall out of Mesa around 2011 but then projects like (https://www.phoronix.com/scan.php?pa...22-Mesa-Grover) pick up similar things like API. Again, no commercial entity is going to do this. Intel GPUs are far more common than the Voodoos. These things are going to community supported for long past our lifespan. Didn't they just recently add a new driver specifically for aging Intel cards?

              Originally posted by sinepgib View Post
              No. But how many people does run those systems? I don't expect charity from companies.
              Thats fair. Most companies don't even offer what I would consider the base commercial support of products I buy from them so there is no way I would expect charity haha. You are right; numbers (in particular monetizable numbers) is all that companies will care about. This isn't sustainable; the community does do better here. Some lone developer keeping some old kit working is better than a company failing to keep some old kit working surely?

              Originally posted by sinepgib View Post
              It's not a defense to companies but a criticism to the unexploited potential of "the community".
              I do have more optimism. Mostly because I have seen Linux be created back then. It was impressive. These days there is a lot of time wasters and companies strong arming their way in; so you might be right that the potential of the community has deminished somewhat. However I feel it is a defeatist attitude to rely on the "almighty powerful commercial gods". They can't (by definition) have peoples best interests at heart so effectively we do want to be reducing our reliance on them as we have proven we have done before with early Linux.


              Originally posted by sinepgib View Post
              But it's evidence the community isn't that all powerful thing you seem to believe it is. The reality is that everyone scratches their own itches.
              I think time will tell. I don't think the direction that Linux going is fantastic. I think it is unlikely that a Linux2 will appear (with a slightly more anti-commercial license). However if companies start restricting individual developers again, we might see either them favoring the older more open stuff or we will see gravitation towards a different system. Perhaps even GNU Hurd (who knows!!! ).

              Originally posted by sinepgib View Post
              In the end, I will do a tiny defense of companies in a way, tho: I'd much rather not be able to run Linux on the N64 than not be able to run it properly on my Intel laptop. Sorry, but I work on the latter.
              Linux was running on my Intel PC long before the company started "helping out".
              Last edited by kpedersen; 22 May 2022, 06:09 AM.

              Comment


              • #27
                Originally posted by kpedersen View Post
                It could have been. It was a clone of UNIX to an extent (Stallman and other individuals did a fantastic task of creating all the bits we see in coreutils today).
                It was "user-friendly" because it adhered semi-closely to the userland of UNIX. Compared to DOS, it was a far more pleasant system to develop on.
                And you just pretty much nailed the implicit here: usable, for who. Usable, for devs. Not enough apparently tho, as it never really took of beyond rounding error outside of embedded and servers. But to be fair, it had more to do with having no really usable toolchain to target end-user devices than any lack of merit of the system. When you develop for Windows, as was the case of most paid developers at the time AFAICT, you're expected to run your stuff on Windows, and developing on it only comes naturally.

                Originally posted by kpedersen View Post
                Exactly. It was doing well. Yes, there were some commercially backed things like SLS but it was the individual developers that were combining their skills and competing very well against the commercial products.
                Individuals only got so far. One problem with community based development is that there's a stronger separation between consumers and development focus. When you develop for yourself, you end up tailoring your system to developers only. Indeed, all the examples you gave on the prior post were useful only for technical people.

                Originally posted by kpedersen View Post
                These days it is a different matter. In many ways commercial companies are pushing the little guys out from contributing. No longer would some (very talented) 19 year old be able to start tinkering with the Intel vendor specific code and expect to have their contribution accepted. Back in the day this was obviously different because there was no Intel involvement.
                Wasn't NIR actually developed by a high schooler tinkering? The ability, IMO, is still there. But as the user base grew, many people felt entitled to demand things. And that's a huge problem. When you do something because you love doing it, in exchange for nothing, you can't really accept demands. Companies can and should to some degree, because people are paying them for a product. Individuals get burned out.

                Originally posted by kpedersen View Post
                The fact that the i.e older Intel GMA 965 and things like that works is due to individuals. Intel have been quite happy to drop support for these cards. They even broke OpenGL 2.1 for the older cards for a while.
                https://wiki.archlinux.org/title/int...L_driver_(i965)
                It is indeed the community that is keeping that stuff working.
                And it's also Intel that put the initial work there AFAIR.

                Originally posted by kpedersen View Post
                PowerVR and VIA are not massively common outside server-space (or shite phones). Think common hardware that the masses have/had.
                Both were massively common in budget netbooks and laptops, respectively. They aren't now, but they didn't get support by the community (again, I mean enough manpower to actually come through, not to minimize the effort of the few that worked on those) when they were common.

                Originally posted by kpedersen View Post
                Such as things like the Voodoo (https://www.phoronix.com/scan.php?pa...item&px=ODQxMw). Yes, they did ultimately fall out of Mesa around 2011 but then projects like (https://www.phoronix.com/scan.php?pa...22-Mesa-Grover) pick up similar things like API.
                Again, no commercial entity is going to do this. Intel GPUs are far more common than the Voodoos.
                Indeed, no company will do so. But that's because it doesn't hit a margin. I'm not painting companies as saints that will do stuff for us. It's not their role really. The claim I'm contending is how much actually gets done by the community, and how much gets done by companies.
                Maybe I just became too cynical by lingering on the internet too much. I've seen very vocal members of the "community" (which I'd rather divide in two: people who actually behave as a community and does stuff, and assholes that demand stuff but aren't willing to do anything) being a bunch of shitheads towards actual developers, company sponsored and individuals alike.

                Originally posted by kpedersen View Post
                These things are going to community supported for long past our lifespan. Didn't they just recently add a new driver specifically for aging Intel cards?
                I seem to recall the same, but can't remember which one. I don't entirely remember who did it either.

                Originally posted by kpedersen View Post
                Thats fair. Most companies don't even offer what I would consider the base commercial support of products I buy from them so there is no way I would expect charity haha. You are right; numbers (in particular monetizable numbers) is all that companies will care about. This isn't sustainable; the community does do better here. Some lone developer keeping some old kit working is better than a company failing to keep some old kit working surely?
                I think there's a dilemma there, clearly. While I agree most companies don't even offer the base commercial support (most hardware ones really do one blob drop at most and forget they even released those chips afterwards), I really, really don't see the community as sustainable right now. I need to emphasize that I don't mean to shit on people giving us stuff, quite the contrary. I mean to shit on most of the users who will demand stuff but be unwilling to either put time or money for those individuals doing work. Surely the bus factor of that model doesn't help either.
                If we talk about communities, we should be worried about the fact most of those longer support cycles offered by the community are actually the effort of a single individual instead of, you know, a group of people, which is what conforms a community. Again, maybe I'm just biased and bitter, but it is what it is.

                Originally posted by kpedersen View Post
                I do have more optimism. Mostly because I have seen Linux be created back then. It was impressive. These days there is a lot of time wasters and companies strong arming their way in; so you might be right that the potential of the community has deminished somewhat. However I feel it is a defeatist attitude to rely on the "almighty powerful commercial gods". They can't (by definition) have peoples best interests at heart so effectively we do want to be reducing our reliance on them as we have proven we have done before with early Linux.
                Don't get me wrong, I don't think we should rely on the "almighty powerful commercial gods" in the slightest. I think we, as a community, should start seeing that as a symptom that we're not committed enough. Precisely because companies will never have people's best interests at heart. Nor do they owe us that.

                Originally posted by kpedersen View Post
                I think time will tell. I don't think the direction that Linux going is fantastic. I think it is unlikely that a Linux2 will appear (with a slightly more anti-commercial license). However if companies start restricting individual developers again, we might see either them favoring the older more open stuff or we will see gravitation towards a different system. Perhaps even GNU Hurd (who knows!!! ).
                I'm not sure how companies are restricting individual developers. Can you give me an example of that? What I see is that, simply put, the sheer amount of resources companies can put naturally offsets what unpaid individuals can send in their spare time. I don't think we should rely on companies, but we should really not rely on unpaid individuals. In a different context we would call that slave labor, it doesn't sound so good put that way, right? If we want an ecosystem to be by the community for the community, we should seriously consider giving back to the part of the community that puts the effort, and get more involved if we have the skills as well.
                Regarding GNU Hurd, I don't expect that to be a replacement for Linux ever. It'll probably remain a project testing interesting ideas, but never to be ready for the general public. I'm not skeptic that something else will replace it at some point, tho. Software always has its cycles and all software eventually grows to find some fundamental assumption about the world stops being true and needs a whole replacement.

                Originally posted by kpedersen View Post
                Linux was running on my Intel PC long before the company started "helping out".
                Yes, but how well? From what I've been told (I'm too young to know but too old to have the energy to try it now), in the Slackware times you would need to build your kernel to have something barely bootable. And I've experienced poor hardware support in several fronts as well (granted, not the CPU, but when I mention "Intel PC" I mean the whole box, the CPU alone is useless) not too long ago. Maybe about a decade it was actually common, and I still see it quite infrequently today.
                People supporting N64 is not helping fix that in particular, but some hardware and commercial distro companies are. Again, I don't intend to shit on whoever supports N64, it's neat, but if the premise is "Intel is leeching of them", well, not really, and if I had to choose what either gives me, one gives me what I need and the other doesn't.

                Comment

                Working...
                X