Hyperion Entertainment Message Boards

Nicsoft wrote:I do understand that MiniGL doesn't work (on my system) "now"... But I wan't a PCIe-card, not a PCI-card.

It already has been announced (I think even back at AmiWest) that we are working on bringing Gallium3d to AmigaOS, This will support radeon HD cards among others.

I like that you are working on Gallium, hope to see it soon as possible

kas1e wrote:There is a chance that there maybe will not warp3d support at all for new radeons hd , not directly (what is expected) , not even via gallium-emulation (i was expect that it will be 100% done, without maybe) ? Because imho not give new users with RadeonsHD warp3d/minigl support in any form, will not give them possibility to run many games, demos and apps. I.e. they will have opengl, but will not able to run any opengl stuff from os4depot , will sound a bit bad imho.

Yes, there is a chance. It will all depend on available time and capacity. I think that most stuff will be updated to the full OpenGL anyway.

On aos4 opengl = minigl, and when he have words "cant initi opengl" in any game or app, it mean minigl all the time of course. You know it for sure that everyone who make some ports for opengl, use only minigl on aos4 Just some apps / games write "opengl" words, but its only text, in reality its all minigl based of course.

The bug report is misleading. There's a bunch of screenshots, everyone saying something different. I took the first ome.

@rogue

I think that most stuff will be updated to the full OpenGL anyway.

Imho 50/50 for minigl/opengl. Because some developers already not avail and not interesting in recompiling (but prods are on os4depot already). Some of them just will not want to recompile everything they do (tons of time as well). And some of the apps , which are pure warp3d, will just unpossible to just-recompile and we will loose all those apps Of course, there is not many plain warp3d apps, but still they there .. Few good apps also have in their code warp3d parts , and those apps will not recompiles , and in end, our not so big aos4-software base will be again splitted and we will have problems with running apps on different aos4 machines

Even if by some luck, some prods which done for minigl/warp3d, will be reworked to make them work on mesa, then, we still will have problems, because all the users who have only minigl/warp3d on their old radeons, will not able to run any recompiled app (even, if it works before). Imho, adding of emulation layer are must even with first gallium/mesa release (to avoid users crying and saying "why if we have opengl, all the opengl prods not works for us" ).

And imho warp3d emulation layer over gallium are more than enough for all the times. As warp3d will be no more in develop, will need only one time to make emulation layer, and so all the apps will just works (and apps which want minigl will automatically works as well, via the same minigl libs, which will just works over warp3d emulation over gallium)

Yes, there is a chance. It will all depend on available time and capacity.

In end of all , wazp3d can be used and updated. Alain also have plans to make in wazp3d support of the gallium, so, maybe its all even can be done by 3d party programmer. But sure will be cool if hyperion will do that

@Hans-Joerg, Kas

Maybe i do not fully understand the problem, but will it possible to update Joerg's Wrapper to solve this problem about compatibility from and to Gallium?

@nubechecorre
Joerg's Wrapper is WarpOS/WarpUp and not Warp3D. They only share the "warp" in the name.

i thought it was also for warp3D ..

@kas1e

I agree at 100%, at some point we strongly need such Warp3D layer, there is no alternative IMHO as we can't miss all our current apps

@samo79

I hope that even if Rogue will not do that, then maybe Hanz will do that, or Karlos. I remember when i talk with Hanz half of year ago, he say that "yes, initial plans to have warp3d/minigl wrappers".

In the worse case we can annoy Alain to fix wazp3d and add gallium3d support. But it of course will be pretty bad for all the new users who will buy aos4, and who will not have default warp3d/minigl working: when they will try to run all the demos/apps/games from os4depot , they will be forced to problems , and to solutions like "go somethere and download wazp3d emulation". I, as user, can only hope that hyperion devs understand the importance of having wapr3d/minigl emulation, and they will do all the best. Through, should to add, that "maybe" from Rogue broken all my current motivation to works on new games for aos4 I just can't imagine that all that work which we do in the last few years with all that games, demos, diskmags, will be just dropped out with words "please, recompile it for mesa". I personally do not remember at all what i fix in half of the games which i already release on os4depot (which are not 1-2, but 20-30), and re-compiling of it will be dropped to "soon" => never => we will loose all those games for all the new customers. Even i myself already not have sources of some games which i port, and report them again will be un-interesting and pain.

For me, all looks not _that_ hard. Its just need done 1 time (emulation of warp3d i mean), and that all. It can be then released with every update later just as "petunia" was done one time, and its with us with all the time now. I also sure, that if gallium3d/mesa will be released without warp3d/minigl layer, it will be big moaning and screaming from all the users like : "we have mesa now, why all the games from os4depot not works?".

@Rogue

For now i just do brief check for all the games which we have: i can say now, that almost every more or less interesintg amigaos4 games use minigl. I think now that its just unreal to be in hope that even 50% of them will be recompiled (i not saying about such stuff as demos and diskmags, which also pretty nice to have). If new users will be forced to not have all of this , then what they will have ? They will have to "wait for new games", or to "wait when someone will recompile it over mesa". I not saying also that all the old warp3d games/demos/apps will not works at all for new curtomers, and which ones 100% will never recompiled.

If you want, i can contact with Alain (wazp3d author), and ask him if he want to share his code to use it on aos4. So you all can just get his work, and just add gallium support to it (Allain already have initial parts for it). And everyone will be happy: no need to spend tons of work to it, and no need to make such big problems for users.

We can put that emulation layers on google-page , and so, everyone will contribute it (just like SDL / minigl / etc). And hyperion only need to include it with OS. But what is most important in end, its do it at the same time with gallium release. Just to attract more users, just to make everyone happy, and so, OS will be solds better.

@kas1e @Rogue

Well your opinion sound pretty clear and correct in my opinion but for what i can understand the problem it's not about their intention but real problem are the limited time and resource, they have many things to do altrought as a simple end-user i have the impression that some things has been "scheduled better" aniway

Maybe at this point the best solution can be splitting the work in many small piece, correct me if i'm wrong but while Hans is busy with his Radeon HD drivers and the Friedens are on Mesa3D/Gallium (both of them very busy to finalize) some good developer remain, so in meantime why not ask Karlos directly and set him on this specific task ?

A month ago i talk with him on Amigaworld about WarpOS layer emulation and he more or less say: Ok i have a good idea for what can be done but for now i don't have any access to the Warp3D code so i wait and see

http://amigaworld.net/modules/newbb/vie ... r=0#618026

Well, while Rogue doesn't have time to do that, why don't open the door to the talent of Karlos and working *together* and NOW with him bringing for us finally full 3D functionality (Full Mesa/Gallium --> Warp3D and WarpOS layers) ?

Of course mine is only an end-user point of view, but i fell that it would be a good way