Page 3 of 4 FirstFirst 1234 LastLast
Results 21 to 30 of 36

Thread: Valve's Steam & Games Coming To Linux

  1. #21
    Join Date
    Nov 2011
    Posts
    263
    Rep Power
    0

    Default

    OK, just skimmed the OpenGL 4 spec and it is hardly a joke. Bur there are some problems that still remain with the original article, and the benchmarks that compare the Windows and Linux versions of Steam.

    • The comparison was made with DirectX 9 (eight years old) and OpenGL 4 (fairly recent), rather than an optimised DirectX 11.Notice that when they used the optimised OpenGL on Windows XP (an old OS), the frame rates were dead even.
    • The specs on the machines for the benchmark are hardly that of the average Steam user, according to Valve's own user surveys. They used a high end gaming box, the average user has a dual core with integrated Intel graphics
    • OpenGL uses a call to OpenCL (the 'glCreateShader()' function) to use Compute Shaders. It does this this very cleverly without having to load libraries, but if I'm not careful I can still get a scheduling stall. With DirectX, the Compute Shader is a part of DirectX itself, so I needn't worry about scheduling issues and resources.


    And lastly, there is something else that really gets to me. See, we can talk about APIs and GPUs and what not, but at the end of the day, the only thing that matters to these people is ROI - Return on Investment. That is why OSS has not had major support from mainstream game studios, or why we have not had the Great Open Source Videogame? Why would I want to make a video game for a small segment of techies, who are philosophically opposed to paying for software?

    Well, if you are a game studio with an App Store , who has just had their app banned from a major competitors App Store (as Microsoft has done to them in Windows 8), and cannot turn to , say Apple, you may to play Knight in Shining Armour to the Open Source crowd. Never mind that Steam is one of the most heavily DRMed software ever released, you know that the penguinistas and the tech media will do anything to get one in on Microsoft. Free advertising and support, without having to spend a single cent. Know that;s what I call Return on Investment.
    Phone: Nokia 1200 with Satantendo ROM v.3.3
    Laptop: (Pentium III 850 MHz, 512 MB, 60 GB)
    Console: Sega Dreamcast
    Check out my blog: http://satanforce.wordpress.com/

  2. #22
    Join Date
    Jul 2002
    Posts
    1,395
    Rep Power
    0

    Default

    Quote Originally Posted by khat17 View Post
    All these young people. I do envy your path still. Mine actually started on Mac and then DOS and Windows. Copy con. Heh. Haven't seen that in a while. Maybe you'll remember things like Blake Stone and SAM and the really epic graphics of CD-MAN and such...
    Nope ... never heard of 'em
    Anyways - Linux is very very good. Free/Open. I should probably abandon Windows but at current here's my thing.

    Windows is for my everyday use. Linux is for troubleshooting and data recovery. Gaming was far from Linux in the past - though some things would work with WINE and some games were ported to Linux. Once STEAM fully goes this way I'll try and streamline some things to get it working.
    Linux had some awesome gaming technology in the past. There were games like Shogo, SiN, Gorky 17 etc.
    Problems in Linux? Have issues with the number of partitions I have - need to use NTFS-CONFIG and MOUNTMANAGER to get everything sorted. After that gets done I realize that sometimes it does not auto-mount the USB drives - FAT/32 or NTFS.
    That problem is relative; maybe even fixed! If you are still using HAL, say hi to Samurai Jack for me.
    Then there's the editing issue. Though there are a lot of alternatives to the Windows software I still firmly believe that there isn't anything that compares with Sony Vegas/Adobe Premiere for video. GIMP and InkScape (among others) are excellent for stills but the learning curve is steep - and the interface isn't as intuitive as those on the Windows platform (familiarity).
    ^^^DIS!^^^
    I totally agree! Though some very creative -- though buggy -- methods showing up. Besides, have you seen "Miniscule", "Sintel" or "Big Buck Rabbit" yet? Next year OSS going to take on multimedia generation seriously (just my hypothesis).

    Back on track though - if they can get this to work with your existing library from Windows then I'll be fine. And no conflicts.
    Two things:
    1. MS has their tentacles deep in the kernel now. What are they doing with them? Extorting companies with secret patents ...
    2. MS has been purposely breaking compatibility with Linux for "years 'n 'ears 'n 'ears"!: Like when MS stalled WINE development over supposed violations through reverse-engineering (similar tactic with ReactOS, though theirs was proven); WINE programmers have to introduce/reimplement bugs! Without them most programs won't work. [Do you know what that does to the "psyche" of a programmer?]


    Reading is Essential
    http://linux.slashdot.org/story/05/0...downloads-site
    http://linux.slashdot.org/story/07/0...ne-windows223/
    http://forums.bit-tech.net/showthread.php?t=95654

    You can only jump through so many hoops till you skim your knee or break your back...
    Last edited by carey; Nov 17, 2012 at 10:00 AM.

  3. #23
    Join Date
    Nov 2004
    Posts
    5,192
    Rep Power
    25

    Default

    The Blender project is just one of the many free/open projects that are around - Big Buck Bunny and the others were nice Sintel was sad though. So am I really one of the oldest members in here? *sigh* so it go.

    Nobody don't remember the old games?
    Knowing the solution doesn't mean knowing the method. Yet answering correctly and regurgitation are considered "learning" and "knowledge".

  4. #24
    Join Date
    Nov 2011
    Posts
    263
    Rep Power
    0

    Default

    Next year OSS going to take on multimedia generation seriously (just my hypothesis).
    That's the problem with Linux. Always next year.... The main problem with Linux adoptation for multimedia is that the ABI is unstable. They seem to change the damn thing every two months, so that you have to relearn it again and again. Compare this to Windows, where I can go into UTech library, borrow a copy of Petzold's 'Programming Windows' from 1999, and everything in it will work just fine. Just ask Miguel de Icaza. And if he can't do it, none of us can!
    Phone: Nokia 1200 with Satantendo ROM v.3.3
    Laptop: (Pentium III 850 MHz, 512 MB, 60 GB)
    Console: Sega Dreamcast
    Check out my blog: http://satanforce.wordpress.com/

  5. #25
    Join Date
    Jul 2002
    Posts
    1,395
    Rep Power
    0

    Default

    Quote Originally Posted by Satanforce View Post
    That's the problem with Linux. Always next year.... The main problem with Linux adoptation for multimedia is that the ABI is unstable. They seem to change the damn thing every two months, so that you have to relearn it again and again. Compare this to Windows, where I can go into UTech library, borrow a copy of Petzold's 'Programming Windows' from 1999, and everything in it will work just fine. Just ask Miguel de Icaza. And if he can't do it, none of us can!
    Then compare that to the guy who just gets the C source of a 32bit linux CLI app from compiles a working app after 12 years. The request that ABI be stable and not change is like saying change does not happen. Hardware changes. The same interface cannot be expected to be efficient/effective with massive changes (unless I err). By the way, I bet everything from Petzold's book somehow crashes after a short run on recent OS without an evident reason.

    I stopped promoting Linux for 1 particular reason. I spend hours tweaking any distro I install to what I want. The typical Windows immigrant wants to have something ready made without any feedback from themself, then they argue how this/that app sucks and it should not be there. The only thing that is a must in a Linux system is the kernel. Everything else is by choice.

    If you compare an inxi report among the months you would see changes in app preferences. The deal is distros sprang from everyone wanting to make '1 true linux'. But then, it is always difficult to agree on a method of doing something. Like, I love ackee, and I love it cooked with coconut oil and as close to raw as possible, but you may prefer to have it cooked with soy or veggie oil and it has to be all crushed and squishy. So distros -- representing the many ways of cooking the food -- abound. Take Foresight. Conary is the method that it pushes. Try testing those ideas in another environment like OSX or Windows. Hence, I say Windows user stick to what you are accustomed to, same OSX user; let Linux be.

    Dying? Au contraire! The evolution never stops!

    Oh. And the line up for next year is Wayland (now >= ver. 1), bootloader evolution (just may have a full blown OS in a bios), self learning and repairing AI, super distributed filesystems ... so much to cover! and who knows? May most likely be seeing new and improved protocols at the network and transport layers.

    If ABI stability is so important, how come Plan9 which seems to never change and has been doing everything the same way from its very beginning is not being given notice?
    Last edited by carey; Nov 18, 2012 at 12:44 AM.

  6. #26
    Join Date
    Nov 2011
    Posts
    263
    Rep Power
    0

    Default

    The request that ABI be stable and not change is like saying change does not happen. Hardware changes.
    If I may respectfully disagree, that does not seem to be a problem for FreeBSD, OpenSolaris, or for that matter, Plan 9. The fact is, having a Stable ABI/API/Kernel is essential for the type of large multimedia project that requires years to develop. Multimedia projects like video and picture editors that need to access Wacom drawing tablets and specialised keyboards or videogames that need to talk to their various APIs. Saying that hardware changes is a red herring, as new hardware, more often than not,comes with old drivers, never mind that upgrading your driver model should have any effect on backward compatibility.

    The problem is as much philosophical as it is technical, as can be seen in your statement

    Oh. And the line up for next year is Wayland (now >= ver. 1), bootloader evolution (just may have a full blown OS in a bios), self learning and repairing AI, super distributed filesystems ... so much to cover! and who knows? May most likely be seeing new and improved protocols at the network and transport layers.
    All that is well and good, but what use is that if every time I close my laptop lid (XPS 15 Sandy Bridge and an old Compaq P4) and reopen it, the keyboard won't work? And this is after I decided to only use Linux on a USB flash drive after a Ubuntu 9.10 upgrade nuked my hard drive on my old machine (GRUB2 epic fail). What is the point of cutting edge gimmicks and gadgets when basic and essential functions like your bootloader and your ACPI drivers are no good?

    Linux needs something like OpenGL's Khronos Group to standardise the kernel/ABI for it to be anything other than a good Server/Datacenter OS. Until then they can forget about games or for that matter mainstream acceptance (Android doesn't count, seeing that Google took a 1.0 series kernel and modified it beyond recognition.) Then you can have one version that is a proper 3 to 4 year LTS hybrid kernel that can be used for mainstream and multimedia, a trimmed down monolithic kernel for virtualization. embedded and server/datacenter and an exokernel/nonstandard hybrid to use as a testbed for all that cutting edge stuff. Until that happens, Linux desktop market share will always be under two percent.

    If ABI stability is so important, how come Plan9 which seems to never change and has been doing everything the same way from its very beginning is not being given notice?
    There are various reasons for the failure of other Unix- and Post-Unix-like OSes in terms of gaining developer/consumer attention.
    Phone: Nokia 1200 with Satantendo ROM v.3.3
    Laptop: (Pentium III 850 MHz, 512 MB, 60 GB)
    Console: Sega Dreamcast
    Check out my blog: http://satanforce.wordpress.com/

  7. #27
    Join Date
    Nov 2004
    Posts
    5,192
    Rep Power
    25

    Default

    From what I've seen with the Unix OSes the main problem is proper support for the hardware devices. Since a lot of the work is done for free it's really up to whether or not the person/s have the time to do the development. Take H.264 for example. Encoding and decoding can be done using software but that puts additional strain on the CPU which could be offloaded to the GPU for example. Until MPLAYER was updated to pass off the decoding to the GPU *nix had no way of getting around software decoding. Until more recent years I haven't seen much where that kind of development is concerned for *nix. Audio and graphics not so bad, but the multimedia focus wasn't/isn't so strong IMO. Yes there are games and such which will run on Linux but when was it that drivers from the manufacturers became available for *nix? That's more on the manufacturers side as a fault still, but earlier drivers had to be compiled which wasn't the easiest thing to do for the average user.
    Knowing the solution doesn't mean knowing the method. Yet answering correctly and regurgitation are considered "learning" and "knowledge".

  8. #28
    Join Date
    Jul 2002
    Posts
    1,395
    Rep Power
    0

    Default

    Quote Originally Posted by Satanforce View Post
    If I may respectfully disagree, that does not seem to be a problem for FreeBSD, OpenSolaris, or for that matter, Plan 9. The fact is, having a Stable ABI/API/Kernel is essential for the type of large multimedia project that requires years to develop. Multimedia projects like video and picture editors that need to access Wacom drawing tablets and specialised keyboards or videogames that need to talk to their various APIs. Saying that hardware changes is a red herring, as new hardware, more often than not,comes with old drivers, never mind that upgrading your driver model should have any effect on backward compatibility.

    The problem is as much philosophical as it is technical, as can be seen in your statement
    Those are why bugs and vulnerabilities proliferate. People do not look at drivers as software apparently. To them it should be a screw with a spiral shape that will always be a screw with a spiral shape.


    All that is well and good, but what use is that if every time I close my laptop lid (XPS 15 Sandy Bridge and an old Compaq P4) and reopen it, the keyboard won't work? And this is after I decided to only use Linux on a USB flash drive after a Ubuntu 9.10 upgrade nuked my hard drive on my old machine (GRUB2 epic fail). What is the point of cutting edge gimmicks and gadgets when basic and essential functions like your bootloader and your ACPI drivers are no good?
    The consensus in the power-user linux community is that Ubuntu is not Linux it's Winnux || Lindows. You get what you pay for. Ubuntu hides everything from you -- including necessary errors and feedback. I bet what is on your USB is actually an Ubuntu variant without proper USB firmware load on boot? Ubuntu contributes a lot open source development. But, that's all it is -- contribution (the life-blood of open source development). Besides, Ubuntu agrees much with including closed source and binary blobs. That is what you get: the end result of not being able to know or modify what is going on inside.
    Linux needs something like OpenGL's Khronos Group to standardise the kernel/ABI for it to be anything other than a good Server/Datacenter OS. Until then they can forget about games or for that matter mainstream acceptance (Android doesn't count, seeing that Google took a 1.0 series kernel and modified it beyond recognition.) Then you can have one version that is a proper 3 to 4 year LTS hybrid kernel that can be used for mainstream and multimedia, a trimmed down monolithic kernel for virtualization. embedded and server/datacenter and an exokernel/nonstandard hybrid to use as a testbed for all that cutting edge stuff. Until that happens, Linux desktop market share will always be under two percent.
    Everything in this paragraph has been written down in security professionals "no-no" books as of now. The info was here at this link. But it was erased from the Internets. What you do not know is harmless to you afterall.


    There are various reasons for the failure of other Unix- and Post-Unix-like OSes in terms of gaining developer/consumer attention.
    I am almost old. What I have learnt so far through my stubbornness is "the customer is always right (on your face; from your mouth); in your mind the customer is always wrong -- just sell them what they want to hear." There is no such thing as a successful, honest salesman. I have been learning to sell for a long time. Honesty has seldom yielded good results. Solve problems honestly, sell suavely. I was angry when BeOS died. I get angry when I get blamed for the software malfunctions on a Windows platform whenever I use it. That is all because of certain developer/consumer choices. I did not make them. Unix-like OS did not fail: developers/consumers did.

  9. #29
    Join Date
    Nov 2004
    Posts
    5,192
    Rep Power
    25

    Default

    Though Ubuntu is as you said the truth is for me (and some others) it's more familiar and a gateway into Debian. I was going to go RedHat but unless I get an older build I not sure how that going go. The mainstream things now AFAIK are Debian and its children, SUSE (OpenSUSE) and those children then BSD and those children. Other than those there are other projects which are independent but not as well supported. The point made above which I completely agree with is that the developers and consumers made things the way they are. If the free/open community had more support - donation or feedback - things would go more smoothly. It would then pivot the manufacturers to support the OS better. The classic case is with video card drivers - for years they've been developed for Windows. Until recently the only ones available were some community hacks - and even so there is very little effort as most of the applications are on Win/Mac.
    Knowing the solution doesn't mean knowing the method. Yet answering correctly and regurgitation are considered "learning" and "knowledge".

  10. #30
    Join Date
    Nov 2011
    Posts
    263
    Rep Power
    0

    Default

    I bet what is on your USB is actually an Ubuntu variant without proper USB firmware load on boot?
    I am now using a FreeBSD that I have made some modifications to. Before that was Puppy Linux.


    Everything in this paragraph has been written down in security professionals "no-no" books as of now.
    I have never seen anything like that in any books that I have read. Could you bring up some other sources? I would find that rather odd seeing that both hybrid/micro and exokernel isolate suspicious processes and files from what is known as the main server (ring 0, IPC, scheduler) and can restore failed programs without the users intervention through the use of a reincarnation server (BSODs drastically reduced).

    But, back to the main topic, or what I think is implied by the main topic -can gaming Open Source gaming be a viable option for developers? The answer, I think, is an emphatic NO. For a platform to be successful, it needs to have an exclusive. The PS2 shot ahead of the pack with MGS2, GTA3 and Devil May Cry. Its predecessor had MGS, FF7 and a whole bunch of Maddens. Even though the PS2 was regarded as a pain to develop for (initially Sony had to fly out engineers to help third party devs), it was still the most successful console ever til the Nintendo Wii came out. That was because third parties had a guaranteed Return On Investment, so that if they sent two years developing a game engine with 250,000 lines of code, a game with another 250,000 lines of code with 10 programmers, 20 artists and a 10 million dollar budget, they can make back enough money to feed their families and pay their bills. How are they going to do that on a platform whose users are ideologically opposed to peopkle owning the rights to their own source code? Not unless they develop on a much more popular platform first.

    As you said above, you get what you pay for. Until then, the phrase 'Open Source Game Industry' will continue to be an oxymoron.
    Phone: Nokia 1200 with Satantendo ROM v.3.3
    Laptop: (Pentium III 850 MHz, 512 MB, 60 GB)
    Console: Sega Dreamcast
    Check out my blog: http://satanforce.wordpress.com/

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •