Jump to content

Photo

The future of the Stella emulator


115 replies to this topic

#101 stephena ONLINE  

stephena

    River Patroller

  • Topic Starter
  • 3,408 posts
  • Stella maintainer
  • Location:Newfoundland, Canada

Posted Thu Feb 21, 2013 6:25 PM

OK, I've started a new thread: http://www.atariage....ation-in-stella

Hopefully it will solicit some response, and if I'm lucky, a developer willing to take on the coding task with me.

#102 Thomas Jentzsch OFFLINE  

Thomas Jentzsch

    Thrust, Jammed, SWOOPS!, Boulder Dash, THREE·S, Star Castle

  • 24,081 posts
  • Always left from right here!
  • Location:Düsseldorf, Germany, Europe, Earth

Posted Fri Feb 22, 2013 6:17 AM

Fingers crossed!

#103 Thomas Jentzsch OFFLINE  

Thomas Jentzsch

    Thrust, Jammed, SWOOPS!, Boulder Dash, THREE·S, Star Castle

  • 24,081 posts
  • Always left from right here!
  • Location:Düsseldorf, Germany, Europe, Earth

Posted Mon Feb 25, 2013 5:16 AM

I am not sure if Stephen is prepared to do a full rewrite of the emulation core
...
Else I would prefer focusing on the known issues of the existing core and go for them first.

After further investigation into the existing Stella emulation core, it turned out, that we might have to go that way and completely rewrite the core. Hopefully this can be done step by step and does not require a big-bang.

But if anyone is interested into doing a TIA emulation core from scratch, without having to worry about all the other stuff (output, controllers,GUI, file handling, etc. pp.) this is the chance for you! :)

#104 dualcam OFFLINE  

dualcam

    Chopper Commander

  • 215 posts

Posted Tue Feb 26, 2013 10:55 AM

After further investigation into the existing Stella emulation core, it turned out, that we might have to go that way and completely rewrite the core. Hopefully this can be done step by step and does not require a big-bang.

But if anyone is interested into doing a TIA emulation core from scratch, without having to worry about all the other stuff (output, controllers,GUI, file handling, etc. pp.) this is the chance for you! :)


This is something I am definitely interested in taking on and have been thinking about a lot, but have a week or two of getting some other stuff done first. This could be a parallel effort to patching up the current code, and even if it didn't get used directly, would help in the understanding of how the TIA works (particularly it's quirks).

Tom

Edited by dualcam, Tue Feb 26, 2013 10:56 AM.


#105 Thomas Jentzsch OFFLINE  

Thomas Jentzsch

    Thrust, Jammed, SWOOPS!, Boulder Dash, THREE·S, Star Castle

  • 24,081 posts
  • Always left from right here!
  • Location:Düsseldorf, Germany, Europe, Earth

Posted Wed Feb 27, 2013 5:31 AM

Fingers crossed!

If the patching works well, I think we could maybe use even unfinished code in parallel. So would have one emulator that has all the debugging etc. features and is 99.9% correct and another one, which may not be 100% finished, but the finished parts are 100% correct.

#106 stephena ONLINE  

stephena

    River Patroller

  • Topic Starter
  • 3,408 posts
  • Stella maintainer
  • Location:Newfoundland, Canada

Posted Mon Sep 23, 2013 6:29 AM

Just thought I'd update this thread, since it's sort of acting as my personal blog on Stella development (or lack thereof).  Sorry if I ramble ...

 

Anyway, I've been recently going over bug reports, looking into what's required to fix things, and if some things are even possible.  At the same time, there's a lot going on in 'real life' ATM, so my time is very limited, as usual.

 

I've come to the conclusion that the current release (3.9.2) will be the last in the 3.x series, and I'll be moving forward with creating Stella 4.0 based on SDL2.  This is going to take some time, and it certainly won't be finished this year.  I decided to do this since there are issues to fix where I see that they no longer exist in the new SDL code, and it is a waste of my (limited) time to continue with the current direction.  Also, I've wanted to move things in a slightly different direction for years now, but have been held back by current design, widespread availability of decent hardware, etc.  Some things I plan for the next release are as follows:

  • Move to SDL2.  The biggest end-user effect of this is that all rendering will be hardware accelerated, using Direct3D in Windows, OpenGL in OSX and Linux, OpenGLES on other systems.  So no more software rendering mode, and (code-wise) no more sacrifices in one area to accommodate a diametrically opposed area (in many ways, software vs. hardware rendering are exactly the opposite, and optimizing for one breaks the other, and vice versa).  So that means you'll require a video card with updated drivers.  Don't fear though, any video card from the past 5 years or so will be sufficient.
  • For Windows, XP will be the minimum requirement, and maybe even Vista/7 depending on how Microsoft compilers work in the next few years.  XP is no longer available for newer computers, and long-term support is due to end in April 2014.
  • For OSX, unfortunately no more PPC support, and perhaps of minimum of Intel 10.6 or so.  PPC hardware is getting harder to find, but more importantly SDL2 won't have a PPC build.  And I don't have time to maintain one myself.  This is especially important, since Stella currently doesn't even build properly on the latest versions of Xcode and Mountain Lion :(
  • For Linux, you'll need a working OpenGL implementation.  This shouldn't be a problem, since the Intel integrated video hardware is getting more widespread, and support in Linux is good to excellent.
  • Because of the move to SDL2, Android (and perhaps iOS) will be capable of becoming first-class citizens.  By that I mean that ports to those platforms shouldn't require much extra work to implement.  They use OpenGLES, which is supported by SDL2.

I want to do this now, because of reasons mentioned above, and also I want to take a break from adding new features.  New features are fun, but if I keep working on that I'll never get anything else done.  This means that old issues are also on the backburner for now, unless I get more help long-term.  In particular, that means improved TIA emulation, debugger stuff, improved paddle emulation, Compumate will have to wait post-4.0.

 

Anyway, I hope this doesn't effect people too badly.  I feel that the gains will be more than the losses, although there will definitely be growing pains in the meantime.



#107 Thomas Jentzsch OFFLINE  

Thomas Jentzsch

    Thrust, Jammed, SWOOPS!, Boulder Dash, THREE·S, Star Castle

  • 24,081 posts
  • Always left from right here!
  • Location:Düsseldorf, Germany, Europe, Earth

Posted Mon Sep 23, 2013 7:11 AM

I have (vague) plans to return to my emulation core rework after I got my current projects (Star Castle + Ixion) out of the door. The refactoring is almost completely done, now "only" the individual special rules and timings have to be sorted out.

So there is a chance that we can throw in a better emulation in 4.0 or soon after.

As for XP, I suppose people will use it longer than M$ wants them to use it. MUCH longer!

#108 stephena ONLINE  

stephena

    River Patroller

  • Topic Starter
  • 3,408 posts
  • Stella maintainer
  • Location:Newfoundland, Canada

Posted Mon Sep 23, 2013 7:38 AM

Sure, what I mentioned isn't to the exclusion of everything else.  It's just what I personally will be working on, and something that I can't ask anyone else to do.  Reworking the different classes, SDL2 stuff, etc requires a conceptual view of the entire codebase, and currently I'm the only one that's really prepared to tackle it.

 

All the other features that we want to add could be outsourced to extra help (if we had it :) ).  But the SDL2 stuff falls to me ...

 

EDIT:  I should add that at least I have a healthier outlook on the project than when this thread was started approx 2 years ago.  Assuming I can get other RL stuff taken care of, the future looks a little brighter.



#109 Trebor OFFLINE  

Trebor

    River Patroller

  • 4,744 posts

Posted Mon Sep 23, 2013 8:07 AM

Thanks stephena, for keeping Stella active and well. It's great that you keep moving forward with it and make the necessary changes as hardware and O/S keep evolving. The SDL2 conversion over sounds like a very cool and perhaps very challenging project. Looking forward to the 'future Stella'.

#110 SpiceWare ONLINE  

SpiceWare

    Draconian

  • 12,801 posts
  • Medieval Mayhem
  • Location:Planet Houston

Posted Mon Sep 23, 2013 3:48 PM

Sounds like a good plan!



#111 AtariDude OFFLINE  

AtariDude

    River Patroller

  • 4,302 posts
  • Playing Chess on Lichess.org
  • Location:Kissimmee, Florida

Posted Tue Sep 24, 2013 6:02 AM

Hi Stephan,

 

I hope that you continue to develop Stella. However, if you decide that you have gone as far as you can go with the project, I want to thank you for your Mac OSX support. Without it, I would not be able to run all those 2600 games that I remember playing as a kid without having to reconnect my 7800 all the time.



#112 Vaughan OFFLINE  

Vaughan

    Moonsweeper

  • 373 posts

Posted Wed Sep 25, 2013 6:59 AM

Stella is much appreciated.  Sounds like a "job" that's going to go on forever.  --lol---

 

Still, it's a labor of love, no douibt.  I use Stella regularly.



#113 stephena ONLINE  

stephena

    River Patroller

  • Topic Starter
  • 3,408 posts
  • Stella maintainer
  • Location:Newfoundland, Canada

Posted Wed Sep 25, 2013 10:28 AM

Thanks for the words of support; it really does make all the difference.  That, and any offers to help, of course :)

 

As for the 'job' never ending, yes, this is exactly the case.  I feel that I'll be finished before Stella is.  I already have ideas on the horizon well beyond what I've mentioned here, but realistically I'll probably never get to all of them.



#114 Thomas Jentzsch OFFLINE  

Thomas Jentzsch

    Thrust, Jammed, SWOOPS!, Boulder Dash, THREE·S, Star Castle

  • 24,081 posts
  • Always left from right here!
  • Location:Düsseldorf, Germany, Europe, Earth

Posted Wed Sep 25, 2013 11:28 AM

I feel that I'll be finished before Stella is.  I already have ideas on the horizon well beyond what I've mentioned here, but realistically I'll probably never get to all of them.

Same here. My project list is ever increasing. And with my current project speed I will never finish it. :)

#115 Keatah OFFLINE  

Keatah

    Missile Commander

  • 22,184 posts

Posted Wed Sep 25, 2013 4:00 PM

In that case, it would be prudent to lay down plans on paper. Notes, concepts, all that stuff. With the hopes that someone with time will be inspired to carry the flag or make it real.



#116 MathUser OFFLINE  

MathUser

    Space Invader

  • 49 posts

Posted Fri Dec 13, 2013 8:32 AM

One reason to keep it going is for ports. I use stella on the wii and if there wasn't a stella on pc then there wouldn't have been one on Wii. So your emulator will still be used in places other than pc.






0 user(s) are browsing this forum

0 members, 0 guests, 0 anonymous users