This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Messages - calc84maniac
Pages: 1 ... 3 4 [5] 6 7 ... 197
61
« on: February 16, 2014, 02:02:23 am »
Hmm, I just started up the Pokemon Red app on my calculator and I got the same issue. Well, I guess I only lightly tested Pokemon, after all.  I'll have to see if I can track down this problem. Edit: Heh, I had to think outside the box to figure this one out. I was certain I got my saving code right, and I checked it and rechecked it... but it turns out the problem is actually in the code for loading from archived appvars (which doesn't affect 8KB save files because the appvar gets unarchived first), which was writing the save data to the wrong RAM pages. I'll try to push out a new version soon. Edit 2: In response to Runer's earlier question, I mainly limited the app names because I didn't want people doing things like making two apps called PokemonR and PokemonB, which would happen to use the same save file names. Edit 3: Here's a release of Alpha 0.0.02. This should fix the loading issue, and also, makeapp no longer gives warnings for app pages beginning with FFh.
62
« on: February 15, 2014, 03:37:53 pm »
It's really good! I tried it with Pokemon. But the zoom isn't perfect, it is only sideways, and it doesn't save. Well, I thought I saved it... 
It should be creating four appvars for the save file. (Just to make sure, you selected Save on the in-game menu before quitting?)
63
« on: February 15, 2014, 11:57:05 am »
Great ! 
Is there official incompatibilities with some games like Pokemon or are all games supposed to work (obviously some games don't work, but the question is are there games you are sure they don't work) ?
I don't happen to have a list of non-working games, no. I generally only test the games that are popular or ones I'm interested in, so I guess it's up to the community to find those incompatible games.
64
« on: February 15, 2014, 01:25:31 am »
Since I've had TI-Boy on my mind lately, I decided to finally release a TI-Boy SE Beta 0.2.04 (which will also be uploaded to ticalc.org). I put in a couple of small bugfixes that I found when working on TI-Boy CSE, but more importantly, the makeapp program is now standalone! It doesn't need an external rabbitsign.exe or tiboy.bin, and it also (probably) won't set off any antivirus programs anymore. Oh, and I also included the Adjustris homebrew ROM that I had in the old Alpha version, so I can stick to ticalc's standards of having an actual APP to put in the 84+ games section.
65
« on: February 14, 2014, 11:16:32 pm »
List of releases from newest to oldest:TI-Boy CSE Alpha 0.0.03TI-Boy CSE Alpha 0.0.02TI-Boy CSE Alpha 0.0.01Let's start this off with a release of TI-Boy CSE Alpha 0.0.01! For the uninitiated, this is a Game Boy emulator for the TI-84+ C Silver Edition (Game Boy Color is not supported). At least for now, this will be a forum-only release in case I've made any terrible mistakes. (Hopefully not, but you never know!) I'd greatly appreciate feedback on any issues you may encounter. Advantages over TI-Boy SE: - Displays the full 160x144 resolution of the Game Boy screen
- 12 custom color palettes selectable, based on the palettes available for Game Boy games when played on the Game Boy Color
Disadvantages against TI-Boy SE: - A bit slower due to the 16-bit full-resolution graphics output.
- No audio output support.
Next on the to-do list is probably to implement a proper fullscreen scaling mode. It's currently only half-implemented, with horizontal scaling but no vertical scaling. Edit: Whoops, I forgot to compile makeapp statically. Updated, so reportedly it works now.
66
« on: February 14, 2014, 06:50:04 pm »
It's getting closer...!
67
« on: February 14, 2014, 12:02:10 am »
Nice to hear. Btw for vertical zoom, will you just draw every 2 row of pixel like this to save speed? Not sure how this would look like on the CSE, but on the HP Prime it actually looks fine.
You mean leaving rows blank? I suppose that could be possible, but keep in mind that the image will be scaled by something less than 2x, so it won't be evenly spaced and might not look great. Another thing is that I'm using the window wrapping to move from line to line after outputting each set of 8 pixels from a tile, so it's a choice between outputting the same 8 pixels again (which doesn't have much overhead with my rendering algorithm) or manually changing the position to output after each 8 pixels (which perhaps takes the same amount of effort as outputting 3-4 pixels as a rough estimate).
68
« on: February 13, 2014, 07:55:24 pm »
Nice job on the progress. Has any research been done into the late-era Game Boy games that try to provide their own limited form of palette color data (of course we're talking about Pokemon)? It's entirely possible that the palettization they do only looks simple and is in fact much more complicated under the hood, but I'm pretty sure that a large portion of TI-Boy users use it to play that (especially now with the guaranteed ROM and RAM space).
If you're not that interested in researching it to support it, I could offer to try.
Actually, how it works is that there is a list of games hardcoded into the GBC boot code (or rather, checksums of the names of those games), which indexes a list of predefined palettes. I think I may try to replicate this behavior, but perhaps not for the first release (it would require a good bit of reverse engineering).
69
« on: February 13, 2014, 01:25:54 pm »
So, a small progress update. Since the first post, I've made the following changes/additions: - Changed: App is now built with makeapp
- Changed: cmdPixelShadow is now saved/restored, home screen is redrawn properly on exit
- Changed: Exit sequence now draws the homescreen at the very start (minus the status bar)
- Changed: Crash recovery now seems to actually work properly. It also no longer requires pressing ON twice.
- Added: Power-saving mode (CLEAR button)
- Added: Backlight control (* and / buttons)
- Added: Changeable color palettes (the 12 palettes available on the GBC for GB games)
- Added: Zoomed-in mode (scales horizontally only, still need to implement vertical scaling)
Now I suppose the main things I need to do before a public release are: - Add some sort of splash/loading screen
- Port my "release-build" makeapp which uses librabbitsign and has no external dependencies
- Make a readme!
At some point, I'd like to add proper fullscreen scaling, but that would probably cause too much delay if I want to get this released soon. I think at first I'll release only on the forums, anyway, so I can work out any unforeseen issues.
70
« on: February 11, 2014, 12:06:16 am »
Will custom palettes or Super Gameboy palettes be supported? Also I recall that some late GB games had colors assigned to them so that when you play them on a GBC they use the proper colors. Will such feature be supported?
SGB is not likely, but I am planning on supporting various colorization palettes like on the GBC. I'm not sure if I'll do auto-colorization, but it could be possible on a game-by-game basis (some games were hardcoded into the GBC bootstrap code to use certain palettes).
71
« on: February 10, 2014, 09:38:04 pm »
The I exploit I provided is just the one BrandonW initially gave me last spring; I didn't write it. It has the deficiency that it only works on boot code 4.0 (it depends on the exact placement of certain code), albeit with the upshot that never references any OS code. You should ask BrandonW (or perhaps thepenguin) for a better one.
Well, either way it should probably be boot-code based (or at least not reference ROM pages 0-3, which are destroyed during TI-Boy execution).
72
« on: February 09, 2014, 10:43:52 am »
Now that a CSE-compatible Wabbitemu came out (thanks Buckeye!) I continued working on the ROM-based TI-Boy CSE implementation (by ROM-based I mean I built it as a TI-84+CSE rom image) and finished off the (unscaled) graphics. Then DrDnar supplied me with a flash unlock exploit so I was able to port over the setup/teardown code from TI-Boy Beta, and I'm surprised to say I can actually run it on my TI-84+CSE!
There are still a few issues with this, like it doesn't properly handle the CSE-specific data on the extra RAM pages yet, so there's probably some corruption going on. I also need to put in a proper splash/loading screen and a better exit screen, plus port over Makeapp and modify my paging code accordingly.
As far as performance goes, it's certainly not as fast as the latest TI-Boy SE Beta, but I'd say it's still playable, perhaps on the level of TI-Boy SE Alpha. Unfortunately there's not much I can do about that, without some sort of crazy dynamic recompilation or something (maybe I'll actually implement that one of these days, but I'd rather put out something that works first).
Sadly, it's not likely that I'll be supporting audio in TI-Boy CSE, because it won't play well with my optimized graphics code. But you never know, I also said that for TI-Boy SE Beta, and I proved myself wrong on that. Here's hoping!
73
« on: January 30, 2014, 12:18:15 am »
Learn asm in 28 days actually has a page on the VAT. Using that info and matrefeytontias's trick you should be able to rename it.
Basically take the VAT pointer, subtract 7, and then write the name backwards. Make sure the program is in RAM though, otherwise when you unarchive it, bad things happen. (I don't remember specifically what, but I remember you don't want it to happen.)
It's actually not a problem with unarchiving, but having a disconnect between the real VAT entry and the copy in archive when a Garbage Collect occurs. Unarchiving afterward actually prevents problems in this case, because the incorrect entry in archive is then marked for deletion. The garbage collector takes the name from the archive copy and ChkFindSym's it to update the VAT entry in RAM, but if the name doesn't match, the entry never gets updated and it points to an invalid spot in archive, showing crazy sizes or error in the memory menu. In addition, the OS doesn't check whether it exists or not after searching, so it writes the new pointer to a particularly volatile spot in RAM that causes a crash the next time you get an OS error. Really bad stuff.
74
« on: January 29, 2014, 02:14:11 pm »
I think TI-84+ BASIC does For(Z,0,1000:End in 1.5 or 2 seconds, but there is no sprite nor rectangle command, so sprites have to be drawn pixel by pixel, using line or text commands, and maths are not very fast.
Actually, in the program above, the ?V line is a shorthand for PRINT V. So it's actually looping and printing each value. (Though, that's usually more a test of the speed of the print statement than that of the loop.)
75
« on: January 25, 2014, 09:17:35 pm »
I think I spy an optimization. You can do the AND C after the XOR instead of applying it to both inputs. That should save at least one byte and 12 cycles, or maybe more if that somehow leads to more improvements.
Pages: 1 ... 3 4 [5] 6 7 ... 197
|
Show Posts