Tux

It's an old windows bug, but it was fixed in some sdl version, make sure you are using a recent sdl (latest SDL2.dll date is april 2023, version 2.26.4 in the about dialog). If you still have the bug with the latest sdl, then maybe the bug is fixed only for some graphic drivers and in this case I'll have to take my workaround back. For now I can't reproduce this at all, with wine in linux or directly in win10.

Not for me, I tried the same game, samsho2, I don't have any sound association for it but opened "manage associations" anyway, which was empty, absolutely no effect on the game (except of course it stops the currently playing track if there is one).

sfz3a is using english text so for this one it's sure sfa3 is totally equivalent is setting the region switch to asia. For sfz3j, it's Japanese text which is obviously not in the rom, so I'll take just this one back for now, but I'll have to check those I removed earlier... progearj seems ok when using progear with japan region at least, so it's probably something specific to sfa... ! 1944 with japan region has almost no text from japan, it even writes USA in its 1st screen but not the copyright screen. So I'll assume the only problem was with sfz3j, if anyone finds another problem, let me know !

I was mistaken, I thought there was a new decryption code finally allowing to totally decrypt the rom, but no !!! In 2023 we are still stuck with opcodes / operands, and by the way raine was already using the same algorithm as xcopy when saving a cps2 rom in debug mode, but what you get are the decrypted opcodes. Some areas of the rom are totally readable directly, like the 8 1st bytes : $ff8000 this is the initial value of the stack pointer at offset 0, but if you go only to offset $60, same file, it should be the vector for irq0, and this vector is obviously unreadable, it's $14e for sfa3/sfz3j but it must be read from the encrypted area. What this means is the roms can't be compared directly, there's still no reliable way to do this ! Ton of crap... ! When you think that the rom encryption keys were in some volatile ram using a battery and so when the battery died you got a useless machine, it would probably have been declared illegal in this time, that's obvious planned obsolescence... ! And so no surprise that you are still unable to get a fully decrypted rom from that. Well, maybe I should just take back all the removed sets, or at least sfz3j then, I'll see... !

Yeah I was surprised too ! At least it was something easy. I tried xcopy but it's no good here, I need to decrypt the whole rom, not just 1 area. In debug mode raine saves automatically the whole rom when you quit, but with cps2 it's encrypted, that's what I need decrypted. I plan to optionally load the keys and if they are available decrypt the thing and run the game decrypted, but only if the keys are available. Anyway for now the weather is quite warm, Roland Garros has started even if it should be a lot less interesting this year than previously, and so it's not an ideal time to code, but I'll try to do something soon.

Yeah I am almost sure that's the reason too, and from yesterday. But it means I need some cps2 decryption code to be able to compare the roms to see if their only difference is really this region byte, and if there is more than that then keep the region clones, it's very possible I'll be obliged to re-insert a few of them and not just one. The code in mame doesn't look too hard to port, but I have other things in my mind, so it might take a little time. Anyway good news for the 1st report : it's actually some very old bug, it dates from the time when I started to add neogeo games on top of neoraine using the code for neocd ! There is a part I totally forgot for the initialization of the game because the width of neocd/neogeo games can be either 304 or 320, until now all neogeo games had a width of 304, which is the default, and which is obviously wrong for some of them, like here for samsho2 ! It's just a stupid bug, the big surprise is that it took us so long to notice it ! Here is a screenshot with it fixed

Well for the 1st, yes it's odd but no idea for now. For the 2nd it was already in 0.95.5c, so I am not the only one who didn't notice. For now no idea neither. edit : not 100% sure but it's probably related to cps2 encryption, sorry but I don't intend to take back sfz3j without further investigation here so it will take some time. I suggest either reverting to 0.95.5c for that or use sfz3jr1 which is still here (and if you can find a difference between sfz3jr1 & sfz3j, you can post it here !). For the 1st it might be related to the crazy way mame (yeah because fbneo just uses mame code) renders the sprites. Sorry but I don't intend to ever render the sprites using the same way, too time consuming for the cpu. Here again not 100% sure, would require more investigation, but not sure I'll want to investigate this further for now.

And I just posted the binary for 0.95.5d, the very last update for this 0.95.5 version. Just what we discussed in the past week, some improvements for the ips support (mainly load_continue support, but also the way the .dat files text is displayed), and the option to mute the sfa3 announcer in sound options.

And you can find the .dat files I built for the ips patches from romhacking there : http://raine.1emulation.com/download/extras.html Notice that the long text of some of the dat files (especially strider) will be displayed correctly only for versions more recent than 0.95.5c !

Yeah ok, I'll try to post that there soon, or in the extras section of the download page. For qsound sorry, that's the only area where we share a lot of code with mame. A long time ago the sound drivers were developed in common for mame and raine, now I just kept a thin compatibility layer to be able to use their code with as few changes as possible (but since they keep on changing their interface all the time we have an old version of their sound drivers now, but they are good enough for me for now). So the gurus who decide how to make a sound chip to sound are not here, even in the mame source there is no email, just a name, and the maintainers changed over the years. But no mention of any low pass filter here. So you can forget about this low pass filter for now.

Yes, and it's saved in the config file. Disagree on the capital letters, short game names are lower case, not upper case, and it's sfz3j not sfz3, but I can add /sfz3j to be more precise. And I agree for the announcer. Luckily there are already some quite long options there so this one can be long without any trouble. And translatable of course. edit : actually sfz3j and sfz3a are region clones, that is their only difference is their region bytes, so they are removed. The text and what the announcer says is based on the region : zero 3 for japan, and alpha 3 for the rest.

Well these are just qsound sound commands associated to samples of the guy speaking, I had to change to sound commands dialog to accept commands on more than 2 characters because his sounds seem to be from $100 to $13a. Filtering these out would silence him. But it would not be a cheat it needs to be included in the code, and I guess the option would go to sound options then, because I am not going to create a cps2 options just for that... It's committed, last option of the sound options dialog. After doing the sound associations for cps2 it was quite easy actually !

I didn't know there was a possible crash when loading different games using ips patches in the same session, if it's been reported it was lost, anyway I just fixed it. Yeah from what I remember only 1 collection was serious, it must be the fba pack. But emufrance allowed to fix this load_continue thing, so it's fine too. Found a few oddities in romhacking there : https://www.romhacking.net/?page=hacks&genre=&platform=26&game=&category=&perpage=20&order=&dir=&title=&author=&hacksearch=Go actually you have to rebuild the .dat files manually there, they didn't even include them, the majority is clearly for consoles but there are a few original arcade ips, including : - bublbobl ultra : more a proof of concept than a real bublbobl extension like buble bobble redux, but it's not worthless. - donkey kong : this is the one which made me bring bps patching to raine, although I knew I wouldn't like the graphics, nes graphics are horrible usually, but I was still curious to see it. - strider : some rom fixes from a real fan. he went as far as finding there was a music which should play when out of time in the rom and which never played, and managed to fix the rom to make it play, plus improved a few of the graphics. Maybe I should post at least some of the dat files I rebuilt by hand, but it's not very hard to do. I just enhanced the .dat parser to allow spaces or tabs as separators and any number of them. They have a few street fighter patches which are totally equivalent to some of the cheats here, so they don't seem interesting at all. These patches are clearly to burn some real eeprom and not to be used inside an emulator, but why not use them anyway ?

Well you have my thanks, because I even if I couldn't make my ips work, I added some code to handle this load_continue anyway, but the ips files I had were too unstable for any testing. This one works perfectly, I didn't need to change anything, the 2 hidden characters appear and are selectable without any trouble. Perfect ! You'll just need to wait for the next binary then... If nothing new happens till then, next week-end probably ! (but I wonder how these files on github can work, it was updated 2 months ago so it's recent... maybe for some modified emulator ? Anyway it's from a guy from Japan so it's hard to ask questions !). And it's normal loading this ips in the current official version hadn't any effect, all its offsets are > $100000 which means all the changes are inside the load_continue area !

After a quick look : - the github repo doesn't even have this file, but it has a crash when trying to select the char on the bottom left corner - it seems it's because it tries to patch a rom loaded with a LOAD_CONTINUE, which really makes things a lot less clear about what are the allowed offsets. I'll try to do something about it... ! But you could have reported this earlier, once again ! After looking : it's some crazy stuff as often from this repository. For example, the .dat starts by : #define IPS_EXTROM_INCL EXP_VALUE_002 if this line has a meaning, it's totally lost, it shouldn't be there normally, it's ignored by raine. Then the .ips inside try to patch a rom loaded with a continue which is already a risky endeavor, but they try to patch after the offset $200000 which is totally impossible with neogeo (the rom can't physically have a size over $200000, it's already split if it goes over $100000 but then $200000 is the last limit). So these patches are lost, if they are supposed to do something here, it's less than obvious. Anyway after playing with these patches (actually your file isn't even in the github repository anymore, but the 2 .dat files there produce a crash in the character selection screen when trying to reach the bottom left part), they always crash the game no matter what I do. So there must be some hack somewhere, but it's not obvious, and really I think I am starting to get really too much of this stuff ! Anyway I don't plan to emulate fbneo, I just wanted an easy way to handle rom patches, it was at the beginning just some support for ips files after the rom has been loaded, no .dat files. Then I added some .dat files because after this it seemed the logical next step (and I already got much more trouble with the file selector than what I had anticipated !), but here it goes really too far. So my suggestion is just to switch to fbneo and to stop to post pictures from fbneo, thanks !

Just posted 0.95.5c version, there are very few changes inside, mainly the few text changes from mer-curious, the change to allow more ips files, 2 new cheat files from ffman1985 : sfa3 and sfa3jr2, and that's all. It's just that there is nothing big in the works for now, so there is no real reason to wait for the ips fix. 0.95.5b was linux only to fix a make install problem. And there should not be any 0.95.5d !

Yeah nice find, I didn't even notice yesterday. It's fixed, 1 message only now, and it doesn't crash on you when too many ips files ! No idea, but for sure they are very different and not comparable. Maybe one day I'll look into it again... ! For your edited commit, well it was already pushed so it had no effect at all here, but don't worry, I just changed the message on my side... !

The message should probably be too many ips instead of too many roms... ! Anyway... !

As usual it seems you wait for a release to test something like that... And yes it's quite astonishing that all these files work without any conflict ! The normal max number of ips files was 20, which seems already a lot considering the conflicts. With 30 it's still not enough. With 40 it works, and without any conflict. The fix is very easy to do, if you had tested that before the release you would have it already... ! It's an interesting combination for others I guess... !

Screenshots of the currently selected game already appear in the background if you have some screenshots saved from this game (ctrl-s by default to save a screenshot). There is also a script to get some screenshots from the internet and save them in the screenshots directory, but you need perl for that. And no, raine never used dos4gw even if it was also the time of dos4gw in this time, it used something equivalent though (djgpp/go32 at the time). If you are curious about the "screenshots grabber", it's in the Extra part of the downloads page, towards the bottom of the page. In linux you must run it from ~/.raine, and you already have perl installed on your system so you just have to run it that's all.

Well for now there is not much content and it's still quite high in the topics list, so it's good as it is for now, maybe later

Excellent, you didn't waste anything here, glad it works anyway !

I'll take a look later at your kof2003 problem, but this kind of path is begging for problems here : spaces + even simple quotes and [ inside a path, and it's so long it becomes ridiculous. Normally there are workarounds for that and this kind of crazy path is accepted, but I hope you at least tried a simpler path before reporting that.

It started by ffman1985 scripts. Found a way to make them compatible with the normal cps2 versions and not just the phoenix versions. Then he noticed the sound associations didn't work well in cps2, normal since they were never really tested. He says there is a site in the internet with re-arranged versions of the cps2 tracks which sounds very nice with the associations... so they are fixed now, it was much easier than neogeo, their major commands stay the same for all drivers. Then I used the cheats to convert some region bytes switches to some regions which can be changed in raine gui, it allowed to identify region clones, that is clones whose only difference is the region byte itself, and get rid of them since we can now change the region to whatever we want. Notice some games like choko don't have english texts for the non japenese versions, which is not a surprise. Then there was the problem of some long standing graphical glitches because of the dreaded priority masks. So I finally added some emulation for these priority masks, it's quite amazing that the driver worked quite well without them for all this time actually ! (and thanks go to mame for their code, even if I couldn't believe it for very long, they probably spent a long time to find this). Except that : - the list of games now using the console scripts from ffman1985 is longer, see his thread for details. Here it is : hsf2d, sf2cejc, sf2ce, sf2hfj, sf2hf, sf2j, sf2, sf2ad, sfa, sfz2ald, sfz2alj, sfz2al, ssf2ud, ssf2xjd, ssf2xj, xmcotar1d - progear had been broken for 1 year ! It's now fixed. - the ips files can now be bps files when referenced in a .dat file. If you don't know what bps files are, never mind ! - some minor fixes in the multiple files selector for "load ips dat files" - since they never fixed sdl2_image 2.6.x which does not return palettes anymore for 256 colors png, I finally merged the right version of their load_png function to be able to keep the color cycling in the raine dialog on systems with a recent sdl2_image (linux only !) - raine can optionally be compiled using clang, it's mainly useful for debuging - add back button in game controllers to switch between dialog headers and main parts in the gui - Plus some minor stuff... ! That was quite a big update this time : http://raine.1emulation.com/download/latest.html Also both appimage files, 32 bits and 64 bits have been updated to 0.95.5, and there was a small problem in the makefile which forced me to make a 0.95.5b version just for the linux packages.

Yeah big update on cps2, ffman1985 started it with his scripts then he found the sound associations didn't work that well in cps2 which was not a surprise since it has never been really tested, so we fixed that, then he complained that we still have some graphical problems in some cps2 games, which were because of the way these crazy priority masks work, after a big hesitation I finally added some emulation for that, we also found a lot of region bytes for cps2 games and could identify clones where the only difference is the region byte, since it can be changed from the gui now it allowed to remove quite a few useless clones. I'd like to check a few more things but a release is probably going to happen soon. For the strings, yeah I saw, I'll check that later (but it can't be applied immediately because I need to check if it breaks some translations or anything else 1st). For your guide mapping, you are free to remap it to whatever you like on your setup, it's just a default configuration, it can be changed as all the default controls, but I won't change it again for now, sorry, absolutely not interested. Same thing for guide to close all gui dialogs, if nobody provides a patch for that, then it won't happen anytime soon. And I absolutely don't care at what retroarch does ! I am not against it if somebody writes the code and provides a patch, but I don't want to try to do it for now. Yeah left key on the keyboard is not exactly logical in this way, well it's because basically this gui is supposed to be very simple and it shows here, left and right keys are going to next or previous item if there is one on the current line, but if there is none then the current item is selected. In the game selection dialog, the game is loaded. Oh well, never mind, I didn't even know that actually, but it doesn't really bother me. edit : for the joystick mapping, it should logically be left and right either shoulder or trigger now to change pages, the reason why it's left and right for now is because it was added before sdl2 and its game controller interface which allows to identify precisely the controls. I agree I should change that one day, but well the left and right do their work for now... !