Announcement

Collapse
No announcement yet.

XBMC Prepares Support For Running Emulated Games

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

  • #11
    Originally posted by NeoBrain View Post
    https://github.com/libretro/libretro...upported-cores

    Fwiw, given that these guys haven't made any steps towards us Dolphin devs (AFAIK) I guess they're doing this sort of porting alone and/or are hoping that emu devs will do the work of porting to the new API on their own. So basically, I'd not expect all kinds of different emulator projects to be supported all of a sudden in the near future.
    Yes, we do porting mostly alone, then try to push support for it upstream if feasible. In some cases, porting work has been initiated by the devs themselves, but that's the exception rather than norm.

    I'm looking forward to see how libdolphin turns out at least. At least gamecube emulation should be feasible in libretro.

    Comment


    • #12
      Originally posted by NeoBrain View Post
      https://github.com/libretro/libretro...upported-cores

      Fwiw, given that these guys haven't made any steps towards us Dolphin devs (AFAIK) I guess they're doing this sort of porting alone and/or are hoping that emu devs will do the work of porting to the new API on their own. So basically, I'd not expect all kinds of different emulator projects to be supported all of a sudden in the near future.
      That page is outdated, so I scrubbed the contents and instead redirected to a more up-to-date listing of cores -

      http://www.libretro.com/pages/menubar/ecosystem.html

      Regarding Dolphin -

      we usually contact the emulator authors afterwards after we have done a libretro port. If they are fine with it then the port can be pushed upstream. Many times this has been a successful approach - libretro is now upstream for VBA-M, SNES9x, Final Burn Alpha, bSNES (to a degree), Genesis Plus GX, Picodrive, PCSX ReARMed, FCEUmm and Nestopia. The few times where we first approached the original emu devs to do a port, nothing came out of it and in a few cases there was actually some controversy over it - Virtual Jaguar, MAME - so making the libretro port first and only then contacting the original devs seems to be a more productive approach.

      In some cases pushing upstream doesn't make sense if the original emu had shaky foundations and a lot of edits were required to whip it into shape (couple of things come to mind - file-based I/O which would need to be rewritten for purposes of serialization of savestates/SRAM - Win32-centric codebase that needs to be rewritten so that it plays ball with platforms other than Windows - and more). But I always try to push it upstream in case the dev is interested. I believe in giving back and not creating my own fortress of ports.

      Now as for a hypothetical Dolphin libretro port -

      I know that PCSX2 is a very Win32-centric codebase where the Direct3D plugin is more mature than the OpenGL plugin, there is no x86_64 dynarec and the Linux port in general is way behind the Windows version. In such cases we don't really feel like creating a libretro port.

      Hopefully Dolphin is nothing like that. If OpenGL is treated as a first-class citizen and if the plugin is at feature parity with the Direct3D plugin, and if dynarecs for both x86 and x86_64 have been written, then we could do a Libretro GL port.

      Comment


      • #13
        That wiki page does not list anything. How very useful:
        http://i40.tinypic.com/i74uhi.png

        I recall the retroarch dev causing some drama before, but my memory is hazy on what drama exactly.

        Comment


        • #14
          Originally posted by curaga View Post
          That wiki page does not list anything. How very useful:
          http://i40.tinypic.com/i74uhi.png

          I recall the retroarch dev causing some drama before, but my memory is hazy on what drama exactly.
          The drama that occurred was actually the act of 4chan exasperating everything and them not understanding what libretro was really and refusing to hear anything the libretro project leader said

          Comment


          • #15
            Originally posted by curaga View Post
            That wiki page does not list anything. How very useful:
            http://i40.tinypic.com/i74uhi.png

            I recall the retroarch dev causing some drama before, but my memory is hazy on what drama exactly.
            You've posted this two times now and I still have yet to see how this is relevant to the thread and the subject matter.

            The Wiki page has been pulled because (as i explained in an above post) it was horribly incomplete. This is the replacement page:

            http://www.libretro.com/pages/menubar/ecosystem.html

            Far more extensive I'm sure you will agree.

            Comment


            • #16
              Originally posted by libretro View Post
              That page is outdated, so I scrubbed the contents and instead redirected to a more up-to-date listing of cores -

              http://www.libretro.com/pages/menubar/ecosystem.html

              Regarding Dolphin -

              we usually contact the emulator authors afterwards after we have done a libretro port. If they are fine with it then the port can be pushed upstream. Many times this has been a successful approach - libretro is now upstream for VBA-M, SNES9x, Final Burn Alpha, bSNES (to a degree), Genesis Plus GX, Picodrive, PCSX ReARMed, FCEUmm and Nestopia. The few times where we first approached the original emu devs to do a port, nothing came out of it and in a few cases there was actually some controversy over it - Virtual Jaguar, MAME - so making the libretro port first and only then contacting the original devs seems to be a more productive approach.
              Well, I wasn't really talking about making emu devs do the porting work (sounds like that's what you asked them) - but discussing with developers how to port the emulator properly. But then again, to be fair many emu developers have such a "meh" attitude that still nothing might have come out of it.

              Originally posted by libretro View Post
              In some cases pushing upstream doesn't make sense if the original emu had shaky foundations and a lot of edits were required to whip it into shape (couple of things come to mind - file-based I/O which would need to be rewritten for purposes of serialization of savestates/SRAM - Win32-centric codebase that needs to be rewritten so that it plays ball with platforms other than Windows - and more). But I always try to push it upstream in case the dev is interested. I believe in giving back and not creating my own fortress of ports.
              Completely agreeing with the "pushing upstream doesn't make sense if..." part. I can assure you that Dolphin is behaving well enough that a port is definitely possible. Chances are it'll require serious refactoring, depending on the desired code quality, but anyway:

              Originally posted by libretro View Post
              Now as for a hypothetical Dolphin libretro port -

              I know that PCSX2 is a very Win32-centric codebase where the Direct3D plugin is more mature than the OpenGL plugin, there is no x86_64 dynarec and the Linux port in general is way behind the Windows version. In such cases we don't really feel like creating a libretro port.

              Hopefully Dolphin is nothing like that.If OpenGL is treated as a first-class citizen and if the plugin is at feature parity with the Direct3D plugin, and if dynarecs for both x86 and x86_64 have been written, then we could do a Libretro GL port.
              It absolutely is not. Dolphin has plugins for D3D9, D3D11 and OpenGL, the latter two being the recommended ones (D3D9 is actually about to being dropped). Long story short, the D3D11 and OGL are on about the same level in terms of features and compatibility, OpenGL is actually faster in same aspects (... and people: Do not take this as an argument that OGL is the superior API, that has nothing to do with it). Dolphin has a 64 bit JIT which is usually faster than the 32 bit JIT and the Linux port is on par with the Windows version.

              Anyway, about actually porting things you should really talk to us first. We're currently (well, semi-actively) planning how to refactor Dolphin's core into a library such that other projects can easily write a frontend for it. If you'd go ahead and write some libretro port of Dolphin without us telling, chances are you'd be wasting your time when you could just go the "easy" way once we have finished our refactoring.

              Comment


              • #17
                Originally posted by Dukenukemx View Post
                Or just buy the adapter that allows you to rip roms from your existing games?
                I would be very *VERY* pleased if you could point me to some resource to acquire such adapter.
                I have a bunch of dust gathering old cartridge that I would like to salvage from bit rot. Plus it would make a nice cartridge port for a home-made emulator console.

                The only solution I know of are based around expensive piece of equipement (repurposing EEPROM readers/writer), but I'm sure there's bound to be a few cheap USB readers out there.
                Care to share your source?

                Comment


                • #18
                  Originally posted by DrYak View Post
                  I would be very *VERY* pleased if you could point me to some resource to acquire such adapter.
                  I have a bunch of dust gathering old cartridge that I would like to salvage from bit rot. Plus it would make a nice cartridge port for a home-made emulator console.

                  The only solution I know of are based around expensive piece of equipement (repurposing EEPROM readers/writer), but I'm sure there's bound to be a few cheap USB readers out there.
                  Care to share your source?
                  Again, why not just download the roms from the internet? It's effectively the same thing, only you save the expense of buying the adapter. After all, no one can really tell if the roms were ripped from your own cartridges or not.

                  Comment


                  • #19
                    Originally posted by DrYak View Post
                    I would be very *VERY* pleased if you could point me to some resource to acquire such adapter.
                    I have a bunch of dust gathering old cartridge that I would like to salvage from bit rot. Plus it would make a nice cartridge port for a home-made emulator console.

                    The only solution I know of are based around expensive piece of equipement (repurposing EEPROM readers/writer), but I'm sure there's bound to be a few cheap USB readers out there.
                    Care to share your source?
                    Here you go:
                    Retrode
                    Ars coverage

                    Disclaimer:
                    I don't own one, don't know how well it performs.

                    Comment


                    • #20
                      Originally posted by erendorn View Post
                      Here you go:
                      Retrode
                      Ars coverage
                      Thank you very much, that was exactly the kind of things I was looking for!


                      Originally posted by dee. View Post
                      Again, why not just download the roms from the internet? It's effectively the same thing, only you save the expense of buying the adapter. After all, no one can really tell if the roms were ripped from your own cartridges or not.
                      I don't simply wan to download one of the "1000 ROMs" pack of bittorrent, (nor buy one of the SEGA-repackaged ROM collection on PC Cd-R)
                      I want to be able to put my very own cartridge, that I've bought when I was a kid with my own money that I saved, into my modern machine (And also access the save games that we did play as children, my brother and I).

                      I have enough disposable income to afford a simple cartridge reader like this one (though I would consider a little bit too much the older solution based around full blown professionnal EEPROM reader/writers for 200+)

                      (NOTE: Actually, I have also a few official ROM collections by SEGA on various platforms)

                      Comment

                      Working...
                      X