You could just make some choices from the list on the wiki. Mainly games that worked poorer, to see if improvements were made.I can test some games right now with my CycloDS. Any requests? I'll update as soon as I'm done.
I should make a list of the error codes. Try at least reformatting your sd with fat32 or somethingCurrently I'm not able to get any of the latest compiled versions to work. It stops before the bios giving me a top red screen and the message DIIF (both ARM7) or TEST (both ARM9). I'll try with the other flash card if I have it here.
Still got the last build that worked on my sd card and it works. It's from July 17th though...
UPDATE: Couldn't get my R4iRTS card running ysmenu to work but that's a problem on my side since I can only make it work on a DSLite, which I don't have right now. Still not able to make it run on my cyclods.
Sure. I'll try that and see if it helps now. If not I'll have to go and get my AK2i and try that some other day. And a DSLite, since I've been using my N3DS and NTR Launcher right now.I should make a list of the error codes. Try at least reformatting your sd with fat32 or something
Actually, I found the error code. It means that dldi initialization failed for some reason.Sure. I'll try that and see if it helps now. If not I'll have to go and get my AK2i and try that some other day. And a DSLite, since I've been using my N3DS and NTR Launcher right now.
UPDATE: Formatting didn't work.
Okay. It's probably a fault of the cyclods then. I noticed that the July 17th version wouldn't create a save on booting whereas the latest would.Actually, I found the error code. It means that dldi initialization failed for some reason.
I wonder why the july 17th version worked though. I might check that out laterOkay. It's probably a fault of the cyclods then. I noticed that the July 17th version wouldn't create a save on booting whereas the latest would.
Anyways, I'll try with akaio and ysmenu later. No use in troubleshooting for a not-so-common flashcard like the cyclods.
If you want I can check later versions between the July 17th version and the latest one.I wonder why the july 17th version worked though. I might check that out later
Thanks for your efforts! It sounds like the cache version really is an improvement for a lot of games. Maybe you could also update the wiki to reflect those results.If you want I can check later versions between the July 17th version and the latest one.
Went and got a DSLite and Ak2i now, so I'm gonna test that and see how things go.
UPDATE:
Here are some good news and, well, different news.
My R4iRTS card with YSmenu on a DSL worked better than AK2i with akaio on both my DSi and DSL. For example Fire Emblem Sacred Stones (E) was fully playable past the prologue chapter (and probably further) with no stuttering in the music and full framerate on R4iRTS, whereas the AK2i would freeze when trying to skip the intro after starting a new game. The R4iRTS only worked with the two arm7dldi versions. The wramicache version might've had less stuttering in the music. I will therefore be testing with the R4iRTS ysmenu from now on. The Arm9 versions both give the TEST error message on launching gbarunner2.
Onimusha Tactics (U)- Tested with DSL R4iRTS Ysmenu and DSi Akaio with same results
Arm7dldi wramicache If you press start quickly together with A, you will be able to press New Game and after a veeery slow fade out of the intro screen you will get a black screen with music from the prologue playing with no stuttering (you might have to hammer start and A. I did this but don't know if it helps). After about 3 mins. the game starts and music plays normally, but the game lags or sort of hiccups. After what becomes a very long battlefield tutorial scene due to the lag, you're able to control your character. Since the game plays in hiccups you have to use long button presses. I managed to get to hitting the first enemy and entery enemy phase before almost killing myself...
The no-wram version just freezes on the start screen.
SMA Super Mario Bros. 3 (U) Sram patched DSi Akaio and DSL Ysmenu with same results
Arm7dldi if you pick save game 1 it goes to map screen of world 1 where the screen just shakes continually. If you pick game 2 you'll be able to play the game with full speed and good sound. Not sure if there actually are sounds or effects missing. Will have to test with emulator later.
Arm7dldiwramicache version freezes after bios with white screen.
Upon further testing I saw that playing game 2 makes the game run as well on the July 17th version. I imagine adding your own save file to the gba folder will fix the save game 1 issue.
Wario Land 4 (UE) DSi Akaio and DSL AK2i Ysmenu
Arm7dldi freezes on portal after first level still. Arm7dldiwramicache gives white screen.
Final Fantasy IV (E) DSL Ysmenu
Still freezes after language selection on either version.
Final Fantasy V (E) DSL Ysmenu
Still freezes after pressing start on start screen on arm7dldi. Freezes after bios on wram version.
Pokemon Sapphire (?) DSL Ysmenu
Arm7wramicache Sound plays perfectly smooth throughout. Arm7dldi still stutters on sound.
Pokemon Leaf Green (U) DSL Ysmenu
Arm7wramicache freezes after first Screen with gamefreak and nintendo credits.
Arm7 Runs as smooth as before.
Pokemon Emerald (U) DSL YSmenu
Still loops the bios in either version.
The Legend of Zelda Minish Cap (?) DSL Ysmenu
Arm7wramicache Barely any background noise. Arm7dldi still has the noise.
Medabots Metabee (U/E) DSL Ysmenu
No change in either version. Sound still stutters. Didn't get to the robattling.
Yeah, from what I noticed the cache version helps in all cases where the game runs. You either have games that run on both no-cache and cache, where cache runs better, or games that run on no-cache but not on cache. In other words the cache version always makes games run better and not worse than the no-cache version.Thanks for your efforts! It sounds like the cache version really is an improvement for a lot of games. Maybe you could also update the wiki to reflect those results.
Yeah, there are a couple of situations that make games work poorly with the caching. This is when they use self-modifying code for example (data is set in memory, but the cache still contains the old value) or when the code is replaced by other code and the old code is still partly cached and not invalidated in time. (I invalidate the i-cache with each interrupt that occurs, which means most of the time that will be at least vblank, sometimes also hblank, and this is enough for most games to work well)Yeah, from what I noticed the cache version helps in all cases where the game runs. You either have games that run on both no-cache and cache, where cache runs better, or games that run on no-cache but not on cache. In other words the cache version always makes games run better and not worse than the no-cache version.
I'll update the wiki in a minute.
Edit:Actually I didn't know you need a wiki account to edit it so I'll have to try and get that first...
Nice work. Could you update the build id date as well please? It is still 17694d4 (2017/05/22).I've started to add the first updates to the wiki. I've been playing around a bit more with the newest versions. Haven't found anything very interesting. Seems that a lot of games take benefit in both sound and framrate from the cache.
Sure thing. So the current build id is?Nice work. Could you update the build id date as well please? It is still 17694d4 (2017/05/22).
Sure, I'll correct that, however I was trying to emphasize that I can't get the arm9 versions to work, for some reason.@Zense Could you please say "wramicache version" instead of "Arm7wramicache version"? Arm7 or arm9 has nothing to do with game compatibility