This might be a stupid question -- I'm sort of new to this -- but will anyone ever release an actual level editor for either SMB3 or SMW, one that allows you full custom control over editing maps, levels, "exits" (etc) from scratch, placing as many objects as desired in a level, etc?

Apart from the illegality of copying the actual copyrighted information into an editor program (rather than simply a copyright-uninfringing hacking tool), is it bordering impossible to create a standalone editor that churns out a rom at the end?

SMB3 Workshop

Lunar Magic



The thing with "full custom control" is it only works if you have full custom control over the game engine.  Whatever limits are imposed by the engine, there's nothing the editor can do about it.  Minor asm tweaks are feasable... but often times what seem like silly limits that should be easy to get around actually require massive changes to be made to the engine.  Needless to say that's something that can't really be automated, and thus won't find its way in an editor.  (At least not usually -- but then again Lunar Magic does some pretty insane things)


There isn't a ROM you can just "make".  ROMs are full games... full executables.  For a program to just "chrun out a ROM" it would have to program a game from scratch.  This is something that is quite a complicated task to do by hand let alone something that you can teach a program to do (the AI involved in something like that is in the realm of Sci-fi)

More modern gaming platforms operate very differently from retro cartridge systems.  PCs (and I'm pretty sure PSX, and possibly Saturn) use a filesystem.  That is... game data can be divided into different files.  You can easily have a different file for each level, for example.  And there is no limit on file size, or how many files you have.  The only limit is the size of the medium (IE:  a CD)

Retro cartridge systems like the NES have no such file structure.  The entire game is globbed into one big file that the system sees "at once" with no file parsing or loading.  (The upside to this is that there are no "now loading" screens on cartridge systems like there are on CD systems).  Having the data in one big clump like this means that everything is affected by the size and position of everything before it.  This means you cannot simply add more data for additional maps without shifting other data around (which would break all the pointers/references to all that data).  It's not as easy as just adding a new file.

Additionally, the ROM chips used are minuscule in comparison to CD systems.  NES PRG-ROM chips are typically only about 256 KB.  A CD can hold up to 700 MB.  To put that in perspective, you can have the code/data of 2800 NES games on a single CD.

Now there's no rule that says you can't simulate a file structure in an NES ROM.  Instead of data being stored in a big clump, you could theoretically have a file structure in place instead.  Though this isn't very practical from a game design standpoint.  I can't really explain why -- but if you're familiar with NES programming you might be able to understand.  The one thing I can mention is there would be a considerable speed hit.  A file system involves some overhead in computing time... and the NES (and even the SNES) are slooooow.  The NES's CPU is a whooping 1.79 MHz  (SNES isn't much better, it usually runs at 2.69 MHz, though it can get up to just under 3.58 MHz)

A game like Tomb Raider can get away with a very dynamic engine because it's for a much stronger and faster platform.  The general tradeoff is... the more dynamic something is, the more RAM it uses (plus it tends to slow down a little bit too).  NES/SNES have precious little RAM.  The NES only had 2 KB of on-board RAM.  Though most games had 10 KB by putting an extra 8 KB on the cartridge.  SNES had considerably more, but still puny in comparison to the dozens/hundreds of Megabytes PCs have (these days, even Gigabytes).  Plus PCs have a nice big fat harddrive that can be used for virtual RAM... meaning the memory you have at your disposal is practically limitless.


There have been some homebrew games made with custom engines (Sack of Flour, Heart of Gold comes to mind -- though if memory serves that didn't even work on an NES.  Plus it was pretty lame).  From my experience, though, they tend not to be as good as some of the commercial engines out there.  Plus, while your editing woes are due to the engine... the engine has to be built to work within the boundaries of the system.  Try as you might... there are some things you simply can't get the system to do.

Also note that none of this is due to emulation (I noticed your qualifier there).  Emulators just mimic the behavior of the system.  Anything the system can do, they can do (in fact -- emus tend to be a bit more lenient on what a game can get away with than the actual system -- though more modern emus are pretty strict).


A custom engine would have to be pretty specific in order to work.  Something as generic as a game maker for NES or something like that would just be too big and too slow.


And, I mean, let's face it.  Making a custom engine is making a game from scratch.  And why bother making an NES game from scratch when you can make that same game for the PC?  You have much more resources at your disposal, a more powerful platform, a wider audience, more tools, etc, etc.  I'm not saying the appeal to make an NES homebrew isn't there... but there's a reason there are so few of them.

Of course the engine can be copied, but there will probably always be differences. *summons Nightcrawler*



*points to his last post*

LOL sorry @ creaothceann, when I first read your initial reply I wasn't sure what you meant, but now that I figured it out myself, I forgot you had written the same thing :p


That appears to be EXACTLY what I mean (although I couldn't determine whether that uses a ROM to extract data from)! Awesome.

Wouldn't the point of the "recreation project," as you say, be to create new levels, to start a community of Mario or Zelda fans who want to share their own ideas for levels? I have so many ideas for levels and level maps -- as I'm sure most of us do -- but to face the task of contriving those ideas to fit into a set framework, or the daunting task of learning how to push the boundaries of a hacking editor even slightly, seems a bit unfortunate as, like the Ideas thread shows, so much creativity and awesome custom-made games might be -- heck, probably is -- going to waste (IMO).

Here I am. 


Bowie, there's nothing wrong with your theory. It's just it's simply not that easy to recreate the engine for the PC. Go ahead and try and you'll quickly see the difficulty in doing so. I'm not saying it can't be done. It can and people ahve been successful at such ventures, but it isn't easy at all. I have had personal experience with such an attempt.

As an exercise in teaching myself how to do platform PC game programming, I set out to do a Mario clone with physics based on Super Mario All Stars. SMB3 specifically I think. I can't remember. It wasn't any problem for me to program the game engine given some time to do so. The difficulty came with matching the game play! Sure, I had my physics engine in place. However, getting it to behave and match the fine tuned gameplay that the real Super Mario had was a real challenge.

Honestly, it gave me new found respect for what went into designing those games. It's one thing to program a game or game engine. It's another to tweak it just right for superb game play.

This process can be made easier if you do as Disch suggested and disassemble the original game code to try and get exact physics matches. But that can be pretty time consuming also and not so easy.

So, bottom line. Can it be done? Yes. Absolutely. Is it easy? No. It's for that reason that it's not surprising to me that these types of projects aren't common. It takes a good deal of effort and may or may not be worth it to the person doing it depending on how they feel about recreating games that already exist.

If you ever get things going... do a hierarchical command structure in your group. (Everybody can submit ideas and info, but only very few (competent) people decide how to progress.



If you want to make God laugh, tell him you have plans.

Rockman is disassembled too... A certain TAS sitemaster seems to work on it.