Ambrosia Software Web Board: ARES X - Ambrosia Software Web Board

redsteven, on Oct 22 2008, 09:14 AM, said:

Am i allowed to compete? after EV Nova * is easy.

rudy, on Oct 22 2008, 03:44 PM, said:

I'm not totally sure what you mean by that... Is * supposed to be "Ares?"

And there's really no competition here. It's a cooperative effort to do something awesome

rudy, on Oct 22 2008, 03:44 PM, said:

Very much yes!

Rebelious, on Oct 22 2008, 09:09 AM, said:

I just sent an email with a zipped version of the Ares Data folder to the email address in your profile.

But those files will be useless unless we figure out how to use that uncompiled SMIV decoder. I looked in the folder, and wasn't greated by a nice, friendly application.

I went through those five topics mentioning SMIV, and found a post mentioning putting that code into Quickdraw. Que?

This isn't something we can brute force do by hand using some image converter and Photoshop, Nathan put this to us as a test to see if we are truly worth his time. You can't trick him by handing him a file with a png of the sprites.

These files are C files. The .C means that is a C programming class file. That file will have the code for 'handling the sprites.' The .H is a library-type file in C. It has common declarations and constant values that get used a lot.

So, the guess is that the SMIV resource is the code wrapper that he *made up* to keep track of where on the sprite the different views are, and where on the sprite the weapons are fired from.

Therefore, what we *really* need is a C programmer who can take these files and and create the code wrapper which will then fetch the png files (that we convert) for the ships.

If we can do that, then Nathan will continue to help us - and only that. Sorry, but no brute force will win this one.

Having said that, I am not saying we give up. There are at least a dozen of us here, right? One of use has to have a friend who knows C. All the C I did was way back in the day, and I don't have the time to *really* re-teach myself. I can try, but it will take me some time - so if someone else knows C, that would be super.

-eveningtoast.

Alright, so our main mission now is to find someone who has at least basic knowledge of C

Glad to see so many people interested so far

I'm a fan of Ares (I bought it ages ago, it was great fun), and I know some C (but I'm by no means a programmer).

This code *LOOKS* like its the code that handled drawing things on the screen from the data files. Which means it is (somewhere) pulling data out of the resource files. Which means it SHOULD be possible to make it "run" and read all the sprite resources and dump them to a "screen" which is in fact a bitmap grid, which can then be saved as a file, and converted to a png [the bmp -> png should be trivial].

I've got no idea how OS9 handled multiple resources by IDs or whatever in a single file. For all I know the code that I think is picking which sprite to load is in fact directly accessing a memory mapped mirror of the game's sprite data.

As it is, I have no idea what the vast majority of the code does, its literally completely void of useful comments (ok, that's a line, there's a line here and there), given much of it I can take a wild guess based on the function names, and not be too far off--but that's hardly helpful for getting it to work. There's alot of other stuff there, and code pulled in from includes, and at lease one large block of assembly that may/may not be remotely meaningful to a non-ppc chip.

I don't have a machine that can run OS9 (ok, that's a lie, I have several, most of them are broken or missing too many pieces to turn on--or they're running linux). I don't know if its possible, but it *MIGHT* be useful to somehow get gdb to run on the ares game itself and attempt to see if we can track how things are being called. Though, the only time I've had that work was on very simple things, and only in a capacity to verify something I already suspected.

Someone might want to look a bit harder at how the code is displaying stuff to screen (if it is) it is most certainly accessing positioning and scaling graphics, I strongly doubt nathan would have given us code that didn't at LEAST read/decode the stuff we're after.

For starters, I assume that IF the sprites are stored (however SMIV stores them) in a grid like EV's PICTs; which actually I think they have to be, because I think some of the carriers had sprites that weren't exactly top down--unless they're 3dmodels, which they aren't, then it would likely be most productive to figure out how to get a SINGLE position [IE like the straight-up index position zero in EV in the upper left corner] non-animated image out of ANY SINGLE sprite into SOME other image format, I think BMPs would work well for this, and worry about generating pngs/masks/whatever afterwards.

From being able to pull out a single image, it should be (far less complicated) to procedurally pull out every other and reconstruct whatever we want from them.


At any rate, until someone (that isn't me--not that I wouldn't try, but I don't have time these days) can figure out how to access resources in an OS9-type program file, everything else is speculation.

Somebody should also figure out what the compile environment on OS9 is/was like, IE, what headers belong to the compiler, what/where they are/comefrom/do. I personally have no idea.

Sorry this is so long-winded circular and probably ~51% wrong, I also think I repeated myself numerous times, oh well--Like I said, its a guess based on a first pass go-over of the code, no prior knolage of what it does.

I should also say *NOW* that I'm not currently able to commit to helping on anything in a serious capacity. In the kind of capacity where I can jump in do something and vanish without warning I'll gladly help when/where I can.

-TLTMAA

ToLazyToMakeAnAccount, on Oct 22 2008, 05:58 PM, said:

Actually, looking at the compiler commands around it, it's meaningless to a PPC chip, too- that's apparently 68K assembly, and the C version of it is the function immediately preceding it. Which is good, as assembly is a pain to read (although that's actually the best-documented code he gave us, so it may well be worth reading through anyway).

I kind of wish I had the time to spend working on this, as it looks like an interesting challenge, but unfortunately I have too much stuff to do this semester (I'm just taking a short break from staring at rocks through a microscope for hours on end right now).

- Edwards

Edwards, on Oct 22 2008, 09:44 PM, said:

I've got the code compiling, i did have to stub out 10 functions that may or may not be needed in order to accomplish NL's task. I also threw together some code to generate the png files, its rough and doesn't do any of the simple calculations for determining how big to make the context based on how many sprites there are and their dimensions. /me now goes to sleep so he can fix Defcon and relax by working on EV Nova tomorrow.

rudy, on Oct 22 2008, 10:42 PM, said:

Rudy that's awesome.

If you don't mind posting what you compiled that would be great. I imagine it could be useful as a learning aid to some of us, if not just interesting.


It sounds like you're pretty confident you're on the right track, correct?

redsteven, on Oct 22 2008, 09:17 PM, said:

Fixed.

I could get going on the port as soon as this Saturday if you guys can get me what I need.

@ Captain Bob - I didn't get the email.

Edited for clarity.

This post has been edited by Rebelious: 23 October 2008 - 08:58 AM

Actually, since Kepi asked for it too in the Just Games thread, I decided to upload it:

Ares Data ƒ (7.9 MB zip file)

I'm about to go to bed, but here's something: SMIV resource 1001 ("Planet Marslike") looks something like this:

 Planet_Marslike.png (97.03K)
Number of downloads: 82

I didn't read the code that Nathan gave for the sprites, but my understanding of the format is that there is a (roughly) 80-byte header with metadata about the sprite, followed by a sequence of bytes which index into the standard Ares CLUT. I don't have a copy of the CLUT on hand, but my recollection is that it consists of a block of 16 "pure" colors, followed by 15 blocks of colors, with 16 shades of brightness.

The "specialness" of this CLUT is why Ares is able to recolor sprites in net games so easily; it simply ignores the four high-order bits of the CLUT index and maps them to another block. Since brightness is consistent across them, you get the same shading, with a different color.

Anyway, I need to sleep too, but I will definitely be continuing with this tomorrow.

Closer.

This post has been edited by Pallas Athene: 23 October 2008 - 11:44 PM

Well, certainly looks like the clock is in motion on this one. Keep up the great work guys.....

Is the intention a direct port? or an upgrade?

Or a combination?

IE, Port the game and all its resources as-is, so the classic version of ares Runs on OSX--with the same look (but with newer code, and less vauge easier to work with media formats), but then also make it so it can take full-color graphics and longer effect animations as well as high(er) resolution ship sprites.

I guess kinda like Marathon/Aleph One.


Though either way, hopefully it can be upgraded to support, for example fighters that can be launched and recovered to their mothership. And maybe weapons that do negative energy reserve damage, or perhaps negative shield damage [recharge, since in prolonged games it always irritated me when I'd end up with cruisers without enough reserve power to fight anymore, but that's not really a gameplay issue--its more of a style].

The energy bit I don't think would be to hard, but the fighters would probably be harder, ~ONLY~ because I don't think the game keeps track of which fighter came from what ship (other than who its orders say to follow).

I'm hoping for a direct port.

At this point the plan is for a direct port. It's not that I'm opposed to the idea of improvements, but I wouldn't want to do anything like that without Nathan Lamont's approval. An excerpt from one of his emails (I added the boldfacing on my own)

Quote

Since he's protective of the universe, that would make a sequal totally out of the question... for now.

He might be receptive to upgrades, but let's get this up and running first

And Pallas... you deserve a cookie... or two.

This post has been edited by redsteven: 23 October 2008 - 12:25 PM

Definitely ou should go for a direct port only.
Add-ons can be added on later.

A GUI change could be in order but if he is only protective of the universe then a great port could convince him to write a second one.


FYI you will all be My life long heros if you pull this off

Lord Commander Anic, on Oct 23 2008, 04:23 PM, said:

I'm inclined to agree with the first part, and disagree with the second.

Only because if you need to do a significant amount of reworking/tweaking/fixing for any new features, it'd be much easier to do during the port, not as a hatchet-job afterwards.

The thought being to end up with something like an EVC TC plugin for EVN (not that I ever found one that really did a good job), same game/missions/graphics, but new engine that the default scenario makes no use of the new features of (other than the ones that it might need to appear the same).

Of course, I'm still speaking with little idea of what needs to be done for either, and assuming that much will need to be remade from scratch.

Though as an unrelated note, would it be worth trying to use something like SDL, so it could be ported beyond OSX?

Back when I first talked to Gavin, the OP, we discussed continuing the story line. However, as it seems Nathan is protective of the actual Ares Universe, I would agree that it would be better to port first, staying within Nathan's good graces so that this project isn't boxed by either him, or Ambrosia.

Having said that, I think we should consider a 'neo-port' option.
This 'neo-port' would include the following;
• The first and foremost goal would be to port the code over to OSX, allowing it to be played natively on a modern computer.
• On top of that I would suggest - provided Nathan allows it - an upgrade to the graphics. It seems to me we have a number of artistically inclined people here, I myself am a professional print and graphic artist, so the world of graphics is no stranger to me. And we can make all the vessels 100% the same as the legacy ones, but with greater detail, the same with the planets, the marines, etc.
• Additionally, I would recommend some game tweaks, like ToLazyToMakeAnAccount stated. I too was annoyed when my Cruisers ran out of energy, and I swore there had to be a way to refuel them.
• Minor eye-candy upgrades, things like when you board a station, have some obvious battle happen on the inside (like compartments decompressing, and bodies ejecting into space), planets slowly rotating, visual damage taken to the ships, etc.

*Another idea would be to modernise the games ability to play multiplayer. GameRanger is still alive, and made the transition to OSX, so if Ares also makes the transition, we should see if we can maintain the games ability to utilise the GameRanger service.

These would be my suggestions. If we have enough people who are graphically inclined - and the permission of Nathan - then I don't see at least a couple of my ideas being too far out of the realm of possibility. As for modifications to the game engine, whoever is doing our port for us would need to be willing to take on this undertaking.

Finally, as Lord Commander Anic said, we should also leave the possibly open for add-ons to be added on later.

-eveningtoast

Here's what I, the dude who's actually going to do the porting at this point, say:

First, we port the thing. Yay, everybody's happy.

EDIT: [And if you're very nice and say "please and thank you" (and, more importantly, I get permission from Nathan Lamont) I might do a Windows port, if everything works well the first go-around.]

Second, if you guys really want upgraded graphics, I can go for that. You make 'em, I stick 'em in the game (as a plug-in, of course).

Third, I might try to port the map editor. If there's a ton of demand.

Fourth, I will personally talk to Nathan Lamont, explain my position, and lobby for a sequel.

But in that order, and in no other way, do I agree to do this.

Understood?

Now, if I can get all the .pngs and the source code by this weekend, I might be able to begin work over the weekend on this. I'll try to set up a sourceforge site for it, but it all depends on how fast you get this going.

This post has been edited by Rebelious: 23 October 2008 - 08:37 PM

Rebelious, on Oct 23 2008, 08:28 PM, said:

Now we have an agenda. So, if I understand this correctly, first we need to make these sprites pngs, then Nathan will give us the source code to be converted, correct?
I can convert files any files we have to png, I just need the files. If whoever gets pulls them out can attach them to a post, or contact me, I can convert them.

Is that where we stand?

@Rebelious, are you able to get the sprites, or is that something someone else needs to do?