i think those two statements are related   i downloaded a lot of stuff needed to build the source but never finished with my PHP based extractor, and it's just easier to let you build it anyways

during play testing i saw quite a few things in the worlds 5-8 rom which need fixing so i will fix those things

so uh, where might one find an windows X1 emu?  i could only find a gameboy advance one  and a macOS one :\\

And also, the area music sounds faster than on the PC8801 version.

Well, honestly, I didn't try to make it that way, it just sorta happened. When it started out, basically, the only requirement was the assembler. But, of course, a way was needed to combine the levels, so I wrote one, which added Python to the list. Plus, there was a need to extract some other binary data from your versions (CHR-ROM, scenery, the demo) so dd or a similar tool was needed. (I suppose I could have just put most of them into the source code, but then I'd need to redo it if they changed.)

Make and the file manipulation commands where just gravy. I'd been using make since my first build, but it's not needed, just handy. I guess the real issue is that I had all this stuff on hand, whereas most (non un*x using) people don't. Hmm.

Speaking of source releases, you can get the latest here and the data pack (if you want it) here.

Changes since the last source release are pretty signifigant and I'm not going to go into all the details. (Read the various change lists that have been posted for the patch releases if you really want to know.) The automated build process has been streamlined (provided you've got the deps), and there are now detailed instructions for building the ROM image manually.

I've decided on a license for my original code containded in the package. Everything other than the level extractor is released under the LGPL. The level editor is realeased under the GPL. The documenation (ok, just the readme) is realeased under the FDL. All the bits derived from code belonging to Nintendo or Hudsonsoft are (of course) owned by them and used without permission.

In regards to making a Windows friendly package, at some point I am going to build a stand alone Win32 binary of the level extractor, so that should help.

Those with no interest in building the ROM themselves can get the latest patch here. Changes in this version are minor and include updates to the first half ot the levels.

Well, I'm off to mess around with the X1 version. Hack and enjoy.

Edit:

I took a look at the Sharp X1 disk image and it's very similar to the PC88 version as far as the levels go. The disk format is different (it's got no headers, just data), but once you account for that, the level data even occupies the same sectors. I was able to dump the levels and get my level renderer to generate images of the levels without too much difficulty.

The level format changed a little bit. All the data present in the PC88 levels are in the X1 levels in the same locations. However, the spaces inbetween are no longer filled with zeros and are filled with something else instead. They added a few more tiles and put them in the middle of the tileset, so any tiles beyond the addition points had their number changed. I created a new tileset image (using the X1 graphics) and almost everything seems to be working so far. I don't know if they changed any of the sprite data format or not. They don't appear to have.

I've still got a few things to check out, but I should have a set of X1 level renders availible sometime soon.

For some reason, all the X1 emu's I've tried have signifiganty worse performance than the PC88 emus I've found. I wonder why that is. They seem to have very similar hardware.

100% LevelWise (I Dunno, Depends on the level of ASM hacks will be implented in future versions)
ASM levels are unknown (about 50%-25%?)
Memory (70%-80%?)
Music does not need any modification what-so-ever.
Graphics - 75% (May need work due to NES Graphics Limits)

If you need someone else for that stuff, Might have to wait,

Um, unless I missed something, the bug with the hidden blocks was fixed in the release prior to the latest one. If it's still present, I have some sort of regression and will have to look into it.