ACS compiler in zdoom?

[Slasher]

Oh yeah, you don't need an extra compiler for Strife dialogs. My editor will decompile and compile dialogs natively (so the KSSC support is just for compatibility at this point).

I've never tried your editor because last I heard, it was a .jar file, and I don't even know what that is, or how to run it.

The "uncompiled form" of the Strife dialog scripts is unknown. All we have is their compiled form. Kaiser's tools were using a grammar of his own; but any other grammar could have been used... In fact, IIRC, the language has changed between two versions of KSSC.

ZDoom's internal representation of Strife dialogs is quite different from the compiled format. As such, if there was a text based format for it, it would not be KSSC. For example, internally actors are represented by their name, not their conversation ID.

Correct me if I'm wrong, but this stuff sounds to me as though it doesn't matter what type of syntax is used, as long as the compiler can convert it into something the engine can understand. If that's the case, it would be even easier to abandon KSSC syntax entirely and move to something more similar to ACS or DECORATE or whatever the devs might want to create.

@ Graf: Is it at all feasible for someone to create a syntax that is not only easier to understand, but also with the ability to add more features to the Dialog Scripts, such as the way ACS has been significantly enhanced? Is it possible to perhaps remove some of the awful limitations of the current Dialog Scripts to make it more powerful and useful?

[Macil]

Oh yeah, you don't need an extra compiler for Strife dialogs. My editor will decompile and compile dialogs natively (so the KSSC support is just for compatibility at this point).

I've never tried your editor because last I heard, it was a .jar file, and I don't even know what that is, or how to run it.

More than likely its java. If you have java installed (almost anyone with a browser) then I'd think you could just double click it to run it.

[bagheadspidey]

It's something like java -jar something.jar... but I think blzut is talking about an unfinished project?

Anyway, glad to hear this might be possible eventually. Seems like it could open up all kinds of possibilities.

[Blzut3]

My visual dialog editor has been around for over a year now! See here.

It's still said to be alpha, but it should work fine. Assuming you installed the JRE by normal methods it should take over the jar extension so you can run it just like an exe. But what bagheadspidey said is correct (you might need to do "c:\path\to\jre\bin\java.exe" instead of just java though). I'll be making a C++ version eventually (hopefully soon), but if a new script format comes out before then I'll definitely update the Java version.

[bagheadspidey]

wait, you are GhostlyDeath? Damn that took me a minute.

edit - oh. :yup:

[Blzut3]

No, GhostlyDeath handed SDE over to me since he was working on too many projects (the logo is still his property). GhostlyDeath doesn't care for Qt.

[Graf Zahl]

@ Graf: Is it at all feasible for someone to create a syntax that is not only easier to understand, but also with the ability to add more features to the Dialog Scripts, such as the way ACS has been significantly enhanced? Is it possible to perhaps remove some of the awful limitations of the current Dialog Scripts to make it more powerful and useful?

Yes, it is. However, to define proper dialog actions ZDoom is currently lacking a proper scripting language. It'd had to be hacked in. So here's another situation where I have been waiting (and waiting and waiting and waiting and...) for Randy to get somewhere with Doomscript. Too bad that he doesn't get anywhere.

[MartinHowe]

At this rate, Halloween TC or DNF will be out before DoomScript

Ahhh, much better! (Haven't had a pop at DoomScript for ages )

[Slasher]

Once again, progress stalled by the absence of DoomScript...

[Slasher]

Bump.

Integrating the ACS compiler into ZDoom is something I've been thinking about before. The problem lies more on the editor side of things. While some use a lump named 'scripts' to store the ACS source not all do. Furthermore, this may cause issues with ACS libraries.

Personally I would love to dump the external compiler for good. The only thing it is good for is to allow certain people to release their mods without ACS source. I think when DB2 gets released and UDMF can be used I'll go ahead and make source-based ACS distribution mandatory for that. This of course should mean that development on the external compiler has to stop.

On the plus side, integrating ACC into the main executable would make many, many things much easier, most notably adding new functions to ACS.

DB2 with full UMDF support has been officially released. And then for whatever reason, I suddenly remembered this thread. Perhaps this can be considered now?
(Both the integration of ACS into Zdoom, and perhaps the discussion of a new syntax for Strife Scripts?)

[Jimmy]

This feature has my support.

[Graf Zahl]

I will not integrate Raven's ACC into ZDoom. The compiler is just a piece of garbage that's better retired. Better write a new compiler from scratch than keep a dinosaur alive.

Wait a moment: Wouldn't that be Doomscript?

[Slasher]

Wait a moment: Wouldn't that be Doomscript?

Is Doomscript any closer to reality than it was when this topic was first being discussed? Just curious.

[Gez]

Wait a moment: Wouldn't that be Doomscript?

I'd hope that DoomScript would be something else than ACS++... Because ACS++ would mean either keeping most of the same limitations ACS has (such as, just for example, treating strings as just a weird type of integer) or introducing a lot of kludges to account for the difference in expected behavior between the ACS scripts (written with these limitations in mind and possibly dependent on them) and the DoomScript scripts.

Is Doomscript any closer to reality than it was when this topic was first being discussed? Just curious.

You tell me.

[Rachael]

I will say... it is. By 7 months.

Of course, that's assuming it ever will be implemented. But the chances of that are very high - it's the when that's so uncertain.