Jump to content

Tux

Ultra Members
  • Posts

    1,217
  • Joined

  • Last visited

  • Days Won

    261

Everything posted by Tux

  1. Sorry to make a new release already, it's mainly to finish the previous one finally : after testing you could get the blue screen bug in mission 2 in shinobin (neogeo or neocd version) just by allowing the speed hacks in neogeo options in raine. You can get the same result if you use a rom cheat, and one of the cheats adapted from the original shinobi from mame is a rom cheat ! So really something had to be done ! So the main part : - disabled the 2 rom checks in shinobin, neogeo and neocd version to be able to use a speed hack or a rom cheat without getting this crazy blue screen on mission 2 - added cheats converted from the original shinobi in mame, it's just a reassembled version, so you can find the corresponding offsets for this version Except that : - Portuguese and Italian are now correctly autodetected in windows when you run raine for the 1st time (without any config file). Sorry I forgot these 2 in windows, hey I don't have anyone to maintain the windows version, so I do it by default, but it's not what I prefer ! - The watch points work better in the console when using musashi (which means the 64 bits version), and watches on words, see the git log for the details, it's quite technical. - Display again the neocd id after the short name of the game in the gui at the bottom of the screen. Just for neocd, not neogeo, that's where it's most useful when testing something which is not official. It was displayed before, but quite a long time ago. It's quite short but after that nothing is planned for quite a while ! http://raine.1emulation.com/download/latest.html
  2. No it makes no sense to try to detect the language for the 1st launch. Don't worry I am almost sure it's that, the gettext dll finally has a language detection for windows, and so it detects your language, but not raine, the result is what you see. In next version raine will detect it too so everything will be fine. Even if it's spectacular, it's still minor.
  3. And here is the converted cheat file for raine for shinobin, adapted from shinobi.xml in mame for the original version (it's just reassembled so there are offsets to find the corresponding addresses in ram or in rom) : https://github.com/zelurker/raine/blob/master/scripts/raine/shinobin.txt I didn't convert everything, because some had different addresses but it was not worth searching them. Almost everything is converted anyway.
  4. Disabled the rom check, there were 2 checks, 1 when the rom boots, and another just before mission 2. This guy might have hidden something else, I didn't replay the whole game, I adapted the mame cheat from the original shinobi to get invincibility to test that. You can have a look there if you are interested : https://github.com/zelurker/raine/commit/3fa97059ebc3234c635e71b43674975980d8288f After testing if you don't disable this a rom cheat like invincibility or a speed hack makes the crc to fail and you get a blue background at stage 2 ! After checking the whole rom asm : no there is no 3rd rom check, or it's written totally differently !
  5. Well sorry I can't guess what you broke in this case ! Actaully from your screenshot the locale gets initialized to Portuguese but raine doesn't even detect it so it stays in english internally, you even have an half translated screen : Background animation should have a translation normally, I checked and there is one for pt_BR. Eventually make sure that the contents of your locale directory is only 4 directories : es fr it pt_BR. I guess you should be able to overwrite the language setting from the gui anyway. (eventually select another language 1st and then once it works you can select what you want). For info yes here the language detection function works fine in windows so it detects french and so my raine starts by default in french, I usually change that to english because I am used to the english keyboard shortcuts. In next version the windows language detection function will recognize pt_BR and italian too by default, but you'll see that only if starting without any config file, otherwise the choice in the config file has the priority. All this might be because of the dll update, maybe the detection got clever and has its own windows language detection now... ! But normally when raine can't detect the locale, it falls back to C, which is english normally (default locale), so I am really not sure how you can do that !
  6. Nope sorry, no change on this side, and the region is not even in this dialog, and I can't reproduce your problem so I'll need some more info, obviously something changed on your system which initialized the language, which is quite a feat because I just noticed that I had forgotten the windows language detection function and it doesn't even know about pt_BR... ! It knows only about english, french and spanish, that's all ! So I guess you must have something in your environment which does that, type win+R and type cmd in the dialog to run a terminal, then type set and return, it will display your environment variables sorted in alphabetical order, check if you have something for LANG or LANGUAGE and report here. (you can scroll the window).
  7. Yeah but packages sizes add up on the server, the other day it ran out of space ! So it's best to try to save as much space as possible, even if these dlls are annoying. Actually if windows was less of the crazy os it is, these dlls should be shared in the system directory and not in the raine directory, that's what happens in linux, so in linux I don't need to maintain all these annoying dlls, I just give the list of what's needed and that's all. dlls are supposed to be libraries of shared functions useful to most applications, that's what these dlls contain here, but they are not shared at all because it's windows ! For the changelog, it's mostly the git log, I just added a few details that's all, you can see the current git log here : https://github.com/zelurker/raine/commits/master/
  8. Sorry, dll update mandatory after all, libstdc++ which is linked to gcc-14.2, for some reason it managed to read what it needed in linux despite the thing not being in the directory, the error shows only in windows. So return to the latest downloads, refresh the page eventually, the dlls are now for 0.97.3, still 4.1 Mb. It's annoying, but after all the last time I had to update it was in september 2023, so it's more than 1 year ago, not that bad. If someone else except you tested this, got the error and didn't report it, shame on him !
  9. Another guy who doesn't seem to realize that whatever the number of hours he spent on this, it's still ridiculous compared to the original work to create the rom, that's why all these works should remain free. Enabling the speed hacks for this rom might trigger this problem then, not sure, it depends if the crcs are computed just at start, or each time they are tested. I might make a patch for this later then... ! Don't know where you got this piece of info from, it's not on the page for the project, but thanks, it's good to know !
  10. Yeah but he promised it's the final version, so it's the last time !
  11. After fixing the windows compilation problem in arch (!), here it is. So the main thing is the new shinobi neogeo driver, see there : Otherwise, fixes/improvements included : - the changes for mer-curious changing his physical display and/or soundcard on his laptop are inside, there is a thread about all this somewhere. - fixed a crash when opening the console after an error while loading a game - fixed a bug in the game options dialog, there is another post about this somewhere too, thanks to mer-curious... - fix the region selection which got broken while moving around entries from the main gui menu. - Restore mouse position when exiting the program, mainly useful for linux, for window managers which react to this - Now use SDL_OpenURL to open an url from the gui, it probably doesn't change anything for now, but it will be useful if raine is recompiled for something else than windows or linux. For once the linux binaries were ready a few hours before windows because of an update problem in my arch setup, nothing major finally, but very annoying ! http://raine.1emulation.com/download/latest.html
  12. Rom version 1.1 released, it should fix the mentioned glitch. I updated raine source to accept this new version, this time the roms are directly in the main archive, no separate roms.zip, so you can either place the main archive directly into your roms directory but it's a waste of space, or extract the Shinobi-TTE-Roms directory, zip its content to shinobin.zip and you are good. Notice that the name of the main archive is accepted too because of this (TTE23-ShinobiNeoGeo-v1.1.zip). Well, that's good and all, but there was an update problem in arch for the mingw compiler, and it can't currently build raine ! I can build it in mingw64 native, but the dlls are totally different there, so I'll probably wait a few days to see if we get an update to fix that... ! Meanwhile the linux build is ok for 0.97.3 then ! (lots of small fixes except this new rom, the most obvious one being the fix for the region switch). Ok, finally I was able to make some 0.97.3 64 bits binary for 0.97.3, the exe comes from mingw, so it requires 1 more dll, but it's huge, and makes the whole thing almost twice as big as usually ! Oh well, it will do for now, I didn't test it extensively of course, but it seems to work fine. I'll make a proper release for windows when it will be possible, until then here is the link to the 64 bits windows binary : http://raine.1emulation.com/archive/raine64-0.97.3.7z Also the history.dat was updated to the 2025 version, I forgot that in the linux binaries... It's in git though.
  13. I admit it took me a small time of adaptation the 1st tries were not good neither, but after that the stage 1 is very easy (hint : use button C on the final boss, it's the easiest way !). Stage 2 is much harder, but the graphical glitch is much farther than that apparently ! The author is working on an update to try to fix that. I am not sure the bug happens in the mvs version (might be related to the number of sprites which can be displayed with neocd, they appear as plain blue, which means it's not a priority problem, the sprites are just not here). Time will tell I guess !
  14. Well I can't do much because I never compiled anything for an xbox and I don't have any, but if this version is from 2011 and you still have someone who can compile this port then then you are in luck, raine was already in git in 2011. The cause of the problem is very likely in neocd/neocd.c in the source directory, so just use a git log on this file going up from the date of your port, and you should probably find something in not too much time. That's of course assuming you still have someone to compile it ! I'd say it's probably a silly bug, like some memory area which has not been cleared on startup or something like that. The guy who did this port never came here at least to say hi, I heard from this once or twice that's all... oh well... !
  15. Ah, no idea, this xneraine is not from here, but I guess it's a recompilation of what was neoraine, that is a very old experiment from quite a lot of years ago now, neocd emulation has improved quite a bit in raine since then. You can try it in the pc version to compare. Notice there is a bug in the region selection in the current raine version, it's because of the changes mer-curious wanted to do in the main menu, I tried not to miss anything, but I missed the handling of the region entry. It will be fixed in the next binary, which I'll release soon now.
  16. https://h0ffman.itch.io/shinobi-neogeo interesting project. There seems to be a problem in mission 2 (which is not stage 2) with the fg layer, but there is still a big part of it which is fully playable anyway with neocd. And added the neogeo version from the roms.zip on the site, strangely the music is different ! It will be named shinobin, just rename roms.zip from the site to shinobin.zip
  17. I took the time out of curiosity to search when they did that change : in mame 177 precisely, it was in 2016! It doesn't come for this rom alone of course, there are quite a lot of changes in the cps2 roms, and they can't be applied automatically because there are some differences in the way raine and mame load these roms. Which means that I'd have to manually test each change, plus it would make the new roms incompatible with the version for raine on internet archive ! Well if you manage to find at least 1 useful change in all these changes, I'll consider it, otherwise just forget it ! The ideal would probably be to be able to load roms from some external definitions files so they can adapt to all these crazy changes for those who want to follow, but it would require quite some changes... ! Plus it would probably create some other problems in some drivers which modify the rom and expect some precise bytes at some precise places (which is not the case of the cps2 driver thankfully !).
  18. D : I remember now, it's just that the key sequence you type is the sequence which is searched, so if you type twice D, it means you are looking for DD, which can be dangerous in case you have a game whose label contains DD, like top hunter, the neogeo game... ! If you are not sure, it's better to enable the "return mandatory" option. I didn't do that to be standard, just to find a way to be as fast as possible with the keyboard, it works well. And for the mame update : it's a never ending headache, I stopped following their nonsense years ago, all this should have been done differently normally (preserving the old rom sets and just adding the new ones with an increasing number when they are found), it would preserve compatibility with all versions, and usually the most recent version hasn't got very important changes. So well I accept patches to update the romset if you feel like it, if not, you'll have to be patient because I really hate doing this stuff !
  19. 1 - I was tempted to say it's not really a bug, it's super minor and you are the only one using this at least from what's reported on the forum, but the fix is trivial, just reset exit_options when opening the game options, so I did it. 2 - gui options / return mandatory C - yeah if it can't find the key you just pressed in the following games in the list, then is starts to look for it anywhere in the game title, that's for when you don't remember the whole title ! This thing is made to be used very fast, and I can confirm it does its job well ! D - not sure, well just press return in this case, I won't bother to look why it didn't select it as it should have done in this particular case.
  20. No problem, it's very easy to reproduce. It comes from the very special nature of the console, contrary to other dialogs in the interface, it always exists, but since the interface has a parent system to be able to draw the background, it can create problems when the parent of the console disappears. Here because of the error, it takes one of the temporary messageboxes as parent and so the 1st time you try to display it, it just crashes because the parent has disappeared ! Yeah it's normal, this part of the gui is a hack, because I wanted the console to always exist, which is not the normal way of working for this gui. So there is a workaround, not too complex, maintain a list of active dialogs to be able to check if the parent is still active when drawing a dialog. If it's not, just remove the parent that's all. So patch very short and efficient. Pushed to git.
  21. xp last update was in 2010 or 2011, so it doesn't make much sense to loose too much time on it. I would have kept compatibility but virtualbox removed direct3d support for xp in 6.1 because of a big security hole in their driver, I would have preferred that they would simply add a big red button to click to allow users to install this despite the security hole, but they just removed direct3d support, so it's impossible for me to run any version of raine in xp in virtualbox now, so too many people want xp to disappear. On my end I quite liked some custom xp editions like tinyxp : small, fast and efficient, but oh well, I guess it's too outdated now. And it's not hardware obsolescence, you still do what you want with your hardware, you can install something else on it, it's software obsolescence only here. And the next victim will be win7, but maybe I won't have to deal with that... !
  22. Well you can always test the last sdl1 version to check, sdl2 started in 0.92, so the last sdl1 must be in 0.91.15. Also if you saw the post in the other thread, we became xp incompatible with 0.97.2, so it's probably because of the new function to detect the soundcard. It doesn't really matter anyway, xp is ancient now, and you are bound to hit an incompatibility like that if you keep adding new functions. Also something you probably didn't think about, but if you set Autodetect as default, then you will never see which card is selected, it probably doesn't matter much anyway as long as it can be heard, but if you want to have the detect and store result as before, the only way is not to have this autodetect as default. (or modify the interface to show which card was detected when Autodetect is selected, but I prefer to keep things as simple as possible). And for the unexpected return of the black border : I had actually tested this in windows, but apparently this function doesn't return the same thing when you call it in windowed mode and fullscreen ! So it got me, because I didn't think about saving the config and run again. And it doesn't make much sense, it's supposed to be state-less normally... ! Anyway !
  23. 1) autodetect as default : we used the other way for years and you are the 1st to have a problem with that, I think we can keep the old default for now, I prefer to see precisely my soundcard name when I go to options. 2) your haunting black border : well not for me in linux. Maybe it's windows only... I am terribly tempted to simply stop to support windows, what a marvelous idea it would be ! I might try that in windows later, but for now in linux : 1) removed my config file, started raine, went to fullscreen loaded dino, took a screenshot of the title screen, then quit 2) run raine again in the same config, loaded dino, took the same screenshot, got exactly the same picture, end of the test for now. ... and same result in windows, same picture in both case (compared the files they are identical). got it, I hadn't the latest binary in windows, it's still the same old problem because I changed the way to get the screen coordinates, in windows we get the height of an invisible windows title bar added, not in linux. 1 line fix, a function name instead of another, but it's really annoying. At least the site is fixed, back with some disk space, you can get the ultimate test binary there : https://raine.1emulation.com/archive/tux/raine.7z
  24. It's a miracle we can still post to the forum with a full disk, maybe the database uses another disk ? Anyway here is a mega.nz link for now : https://mega.nz/file/OYVAXI4b#uMADM9P6tyzFjRN3kUj90s1spQKI-B188WSJmvIBYDE
  25. Oh, ok I surrender, I added an Autodetect choice for the soundcard in the gui, and I fixed the detection of the sample rate which was linked to the sound driver instead of the soundcard... ! You won't have a test binary for now because after the forum, it seems the ftp is broken for now, I can't send raine.7z ! (deleted the file to avoid any mistake). It's still a rare case in my opinion, but it's quite interesting to see how it can be done without having something too big. Since I had to fix the sample rate detection too it's a little longer than what I anticipated but it's still ok, and maybe someone else would use that ! Notice that the behavior is the same as before : if you don't choose "Autodetect" explicitely and you have no config file, then it will detect the soundcard only once, there's no reason to detect it all the time except on a crazy laptop changing its output device all the time ! Difference from before : now if you change the soundcard in sound options and the sound card driver has a preferred sample rate, you'll see this sample rate selected in the options in real time. Not all drivers have a preferred sample rate, if there is none the frequency is not changed. By the way the error on the server is because of a full disk !
×
×
  • Create New...