Tux

ok 0.91.10 released with these fixes for windows, I'll try a linux binary tomorrow

Ah I found the forum was very quiet lately, so it was because of a technical error ?!!! Sorry to learn that... Oh well no wonder there are still things to fix about all these shiny new cheats, too much to test, I'll have a look, but I don't know when, it shouldn't take 10 days though ! ok bug confirmed, I hadn't thought about multiple scripts using a choice from a list like that running at the same time, I'll have to find something, it will take a little time... edit again : it's fixed, I'll have to make a new binary for that, and maybe a linux binary too, the biggest change since 0.91.9 except this one is for linux to be able to run its asm code with the recent change which makes the stack not executable by default. Oh well, not an ideal setup to make a binary right now, it will wait until tomorrow. Thanks for the bug report at least !

Yeah I plan to do that too when I get access again to an old 1Ghz pc I have, but for now it's quite far, it will be later. I plan to test with univbe too to check if there are some modes where triple buffer can be enabled, for now the only mode which can work for sure is mode-x 256x240, I am almost sure it's possible in vesa, but maybe univbe is required and I can't test this in an emulator...

I don't know launchbox, what I can say : if you pass the rom file on the command line, it will be seen as a zip file for neocd, which you don't want here. The correct way is to pass the game name on the command line, not the zip file. for aes/mvs, notice you can switch most neogeo games between the 2 if you use something like the unibios, there is a shortcut when you boot the game to decide if you want to boot it in aes or mvs, but anyway... !

This was a hack for version 0.28 but which was not fully understood at the time and which was totally forgotten for all this time, it allows to switch to a given video mode and then make it to display only 1/3 of its lines on screen, excellent for extra wide games like darius. These modes were impossible to select since the vesa modes display only the supported video modes now, and they are not directly supported, it's like a virtual video mode, you need to switch first to a mode with 3x the number of lines you want for it to work... ! So... ! I found this by looking in the code how the triple buffer worked in dos, and found this instead. So I added them back, but I had some trouble to test of course, I wanted to wait until I have access to an old pc to test this on a real screen, but I don't have it available right now. Well I just found that dosemu in linux emulates this perfectly, here is a screenshot : And ingame : Oh yeah cut a part of the title bar on this one, testing a new screenshot tool... Anyway it works great, and it would work on any pc able to switch to 1024x768, so all normal pcs then ! I just added the 3 most standard modes at the end of the list of the standard supported modes, they appear with a "dark " prefix in the list, and are used automatically when loading a game if "auto mode change" is selected, and it's the default. (these standard modes being 640x160, 800x200, and 1024x256, used here by darius). Quite amazing to see such a come back !

There is a post elsewhere in the forum about that, apparently the issue was fixed by a driver update, look for the post for details, it shouldn't be too hard to find there was a picture posted...

And finally some "dark modes" as they are called, I added some because I was curious to see if worked in the virtual machine : it doesn't ! I added the 3 most "standard" ones, 640x160 (which uses 640x480), 800x200 (for 800x600), and 1024x256 (for 1024x768), the 3 appear with a prefix "dark " in the list of vesa modes if you already have the corresponding standard vesa mode. For me the screen is changed indeed I see only the upper half of it, so it's not usable, but it was worth testing. I'll leave it for others to test eventually, trying to load darius in there with "auto mode change" selected automatically takes 1024x256. I suppose that when it works the modes are darkened because of the video hacks used to display them. For me it's stretched but not dark... too bad !

About the crash when out of memory : it's because of a swap problem with djgpp/cwsdpmi, for some reason it should swap when going over the physical memory, but instead of that it crashes, no idea why, it's specific to djgpp. I investigated about that using the virtual machine, crashes just mean to restart it here, so it's quite quick to test. If you are interested, there is a workaround : download the cwsdpmi distribution from there : http://sandmann.dotster.com/cwsdpmi/ then run cwsparam in the same directory as cwsdpmi, and when it asks for the swap file, type "" to disable it (really the double quotes, not just return). Then when it quits, copy the cwsdpmi.exe to the raine directory, and now you should get a nice error message instead of a crash. I remember seeing djgpp swapping a very long time ago, so it worked at a time, no idea why here it doesn't work, but anyway you really don't want any swap with raine, very very bad idea !

For me at least, the data segment suddenly stopped being executable, which crashed all the asm in raine. That's with the latest binutils update, 2.35, it's very likely related even if I didn't find any info about it. You won't see any problem as long as you don't recompile the code with a linker from this 2.35 binutils release, so the released binaries will be ok. For the new ones, I just committed a patch to git, it was easier to make than what I had feared, just call mprotect on the data segment of 1 of these functions, and since all the asm functions are grouped in the same segment by the linker, there is just 1 mprotect call to make. It's very linux specific, but afaik no other os did that, except osx, but I don't care about osx anymore, and this patching wouldn't work in osx (unless they have /proc/pid/mappings, which I doubt !). It took me a big part of the day to find this, I just wanted to test something else this morning and stumbled on this... it was a bad day starting ! Anyway, back to normal now ! Oh, by the way this new linker also emits some warnings about relocations, but they are just harmless warnings, so they will stay for now. I patched only 1 function with the problem because that makes it shorter and more readable, it's the contrary for all the other functions, so they will stay as they are for now ! Warning looks like that : /usr/bin/ld: linux-gnu-sdl/objectd/68000/s68000.o: warning: relocation in read-only section `.text' /usr/bin/ld: warning: creating DT_TEXTREL in a PIE

I'd say return to the seal one which gives you less problems then. The dos version had a mixer built-in, you could display it while a game is running, but it's mostly when some specific sound is too weak/too loud, not to tune the whole volume. Max mixer volume does what it says, it just pushes the output to the max, but it's usually already the case. The problems might be because it's a weird brand, I don't know. I can't help you more on that I am afraid ! I liked working a little bit on this too, it brings some memories, and it's fun to have quite a lot of the recent stuff to be at least loadable in dos !

Hum, not here, but the sound is regularly interrupted by silences here in the virtual machine when using allegro, so it's really not ideal. Maybe try a different sample rate, but otherwise out of ideas ! On my side I checked the triple buffer, in fact maybe you can do it with this good old display doctor to get some very special vesa modes, but with my setup in a virtual machine, no vesa mode can work with triple buffer. It actually requires to allocate 3 times the mode and switch the active page between the 3 buffers. There are some very specific modes which work with mode-x : 240x256 (and I confirm this one works, although it's not very convenient), 320x256. Then there is a serie of modes which divide their vertical lines, for those very wide games : 640x100 !, 400x150, 640x150, 640x200, 640x120, 640x160, 640x240, 800x150, 800x200, 800x300... For all these modes, it setups a vesa mode with 3 times the vertical lines before dividing the displayed lines by 3, but they would probably never appear in the mode selection list since it would require that the final mode is supported (640x100 !!! never seen something like that !). So these triple buffer modes were very rare actually ! It would be possible to painfully update the list displayed to add these weird modes, but it's not worth the trouble, too few of these games, and they always look very strange with these impossibly wide screens... So to sum up : more problems with the allegro sound version, and triple video modes work, but are a pain to setup and so very rarely used in reality ! Oh well, these were hard times !

Yeah that's what I was going to propose too, it was to be expected, this seal is too old to be reliable now... The download page is not used to having 2 dos versions so they show both as "Dos binary", the new one is 1.8 Mb (almost the same size, just a little shorter), and is named rainead-0.91.9.7z (a like allegro). Good luck and thanks for testing ! Annoying for the triple buffer, not a big surprise too, this triple buffer worked because of some old hacks from Antiriad, we tried to maintain them with new versions of allegro, but something clearly broke somewhere, and it's going to be very hard to find where exactly since I didn't make these and I don't really know how it's supposed to work ! I might eventually test later with some binary which has a working triple buffer, I guess 0.34 worked with triple buffer then ? But it's very old... ! Oh well... ! Maybe later, at least it's good to know !

Should be ok, 32 bits versions are optimized for pentium3, but not the dos version which is still optimized for pentium in case of old hardware like this one so there should be no trouble. Maybe the sound ? I'll wait for news from you then !

Oops sorry I hadn't updated the old versions page, and it's 0.91.8. There were some very minor tweaks in 0.91.9 for the dos version, it now restores the dos screen when leaving, and fixes the web site name displayed in dos, but they are not worth building a new binary alone, but maybe you would be interested by the fixes anyway... ! Ok, I just built a 0.91.9 version for dos and uploaded it, you can find it in the latest version page : http://raine.1emulation.com/download/latest.html The only question I have is for the sound, in dos there is the choice between seal which is dead since 1999 if I am not mistaken, or allegro which is a little more recent. Seal seems to work a little better in dos emulators like dosemu, so it's the one used in this build. But I suspect the allegro version is better when you have the real hardware. So test this and report if you have problems with the sound. Notice that with 32 Mb of ram you won't be able to load all the games, some are really big for dos standards !

Easy way to do it : just delete the bld files you don't like, this will disable the blending for the game ! By the way there might be some update available for these bld files, and if not it might be a good idea to try to find who produced them to post about the glitch you found... Post here if you find where to post for that, I am curious to know !

The 7 credits at once has always been here, I won't fix it for now (not one of my prefered games by far, so I won't loose too much time on it). For any other problem : weather way too hot for now, maybe later, I don't know when ! tested it in the heat : only the music goes mute for me, but slowly, it's a fade out and it comes back at the end of the level. The other sound effects seem ok. And it's not 7 credits but 4, but it's still too much anyway. I'll probably leave it as it is !

Just some quick hot fixes, mainly the region selection which was broken for neocd, a bad sound in opwolf which has been here since the old dos days of 0.28 and before, wrestlefest is back with some updated cheats, incredible that I didn't get any post about that finally, but I don't miss them (it was badly updated last time), truxton2 / snowbros 2 were hanging if using the 32 bit version (with mz80), and the status & clones settings from the game selection list are saved. That's all for today ! http://raine.1emulation.com/download/latest.html

Oh sigh, there is a bug in the region switch for neocd then... hum, bad timing weather too hot currently and tired. It's probably something silly and easy to fix anyway, it worked before of course... Ok, fixed and uploaded 0.91.9 for that, see the topic I am about to create for that !

Finally updated the opwolf driver, a real update this time, the 1st one in more than 10 years ! - Automatic rom loading - samples live decoding instead of the long static sample list + the hack to insert the sample codes in the rom - the driver is now shorter and uses the mame rom - actually the only difference is the last 68000 rom which was probably badly dumped in the raine version, never found where it creates a problem if it creates one, but it's probably safer to update this romset anyway ! It won't interest a lot of people since this bug survived for so long, but really this code needed some care... !

Glad it's specific to an intel video chip ! By the way you can crash easily this version by changing from opengl to normal blits and then back to opengl, and playing a few seconds each time. Well it's still the deprecated api which shows here, I can't do much about it, it's safer to exit if you want to leave opengl and reload with a different video setting. Yeah I remember you already asked the version number in the tiltle bar... ah yes I added it when a game is loaded, but forgot when there's no game... You are really a version number lover, I guess it's because you keep a few different versions at the same time... oh well... it's done !

Yeah it's an endless race to try to keep compatibility at the rom level with mame, all that because when they find a newer set, instead of adding it as a new set, they replace the main one by the newest they just found, breaking compatibility with all the sets all the time. Too tiresome. I do it mainly for cps & neogeo, for the others it's from time to time, but very rarely. terraf is a horrible game anyway ! On my side I hate breaking compatibility with existing sets (I mean the ones which worked in previous versions on raine), even if there are utilities to fix that.

Finally it's a quiet day here and I was curious, so I looked into it. The opwolf in raine uses the z80 rom from opwolfb to emulate the c-chip, so it's a kind of improved bootleg but not exactly the real opwolf, the c-chip is not emulated. The sound problem was not as hard to find as I thought first. There were 7 bytes for the sound commands from 0 to 6, Antiriad tried byte 6 as sound activator, it was byte 4 (as found in mame). This means he never finished stage 1, or didn't care. Also the aiming is a little too high in opwolf (not in opwolfb). So I fixed the sound and the aiming, which are very short fixes. There are hacks everywhere in this driver... anyway ! It's fixed but not released, it's ultra low priority since it's been like that for 20 years, but it will be fixed in next release then.

minimum curiosity, try emulator inputs in "controls", f11 is a good idea yeah, you could have tried. There might be a problem specific to opwolf, I never really played it. I might check that later, but it's not my kind of game, so it will probably take time ! edit : yeah I could use the cheats to finish quickly the 1st stage, I can confirm the problem, I guess it's the kind of problem without importance for Antiriad since it doesn't impact the playability... This bug has probably been there since 98, and it's the 1st time I hear of it, so I guess there is no emergence to fix it, I might take a look later, but not now...

Well up to you, I just re-uploaded the 2 raine archives (because the part from sdl which redirects to stdout.txt was in the static part of the lib, so updating just the dll does no good !), and by the way I found a few more sdl patches, mainly security fixes, so re-uploaded dlls32-0.91.7.7z and dlls64-0.90.7z with an updated SDL.dll inside. Most people don't need to update here, but it's probably better to get these versions for the new downloads. No version number change of course, it's just the patches + the configure options which change for sdl. If your binary creates a stdout.txt file on launch, it's the old version, you can also check from the compilation date in the about dialog...

Yeah I know it's super annoying but it's the way sdl1.2 with its default configuration works in windows ! Which means that if I build it automatically from linux like any other package, then you get this ! I knew it was like that, but I wanted to be sure you couldn't work with the stdout.txt in this case... I'll update the dlls package today then with a sdl.dll which doesn't do that !