forums.marpirc.net

Dang it - I'm trying to come up with a clean set of diffs basing on the 0.122 ones, but the addition of ext_inp_header needs a bit of care, and I'm too tired after a working day to use more brains - will give it a try tomorrow hopefully.

A question for mahlemiut -
inpview12-xxx.diff and wolfxxx.diff both modify inptport.c, and that's likely not the only case; why keep distinct sets of diffs that touch the same files ?

It would be easier to run a diff -Nru on the vanilla vs the wolf tree of course, so I'm curious about the purpose of the distinct patches...

Thanks, --alessandro

I've had no issue with always applying wolf before inpview, just cat wolf*.diff inpview*.diff > wolfmain.diff. inpview certainly is a mainstay part of wolfmame so it makes sense to combine them.

*EDIT*; on a side note: are were the dipports.diff and mamerand.diff patches merged with wolf.diff at somepoint? since they stoped showing up in the source but the old dipports.diff still applies ok with no errors.

Hmm, I managed to come up with a set of patches that record new games and play back new and old games, but I must have botched something because I had to edit inptport.c reversing the memcp() logic that checks whether it's a XINP file or not, inside setup_playback()

I'll give that another look when I'm less sleepy...

The order you apply the patches shouldn't matter, you'll just get a few offset errors on the second patch. They are separate so that they can be worked on separately. One does not require the other to work.

There is a bug in MAME 0.123 where INP files are detected as extended when they're not, and vice versa. The memcmp logic is correct, it's just the following if(extended_inp) statement that's wrong (should be if(!extended_inp)). This is addressed in the the wolf123 diff.

dipports is still used, it just hasn't needed changing since 0.113. mamerand was removed, when the rand() fixes were applied to baseline MAME.

mahlemiut wrote:The order you apply the patches shouldn't matter, you'll just get a few offset errors on the second patch. They are separate so that they can be worked on separately. One does not require the other to work.

There is a bug in MAME 0.123 where INP files are detected as extended when they're not, and vice versa. The memcmp logic is correct, it's just the following if(extended_inp) statement that's wrong (should be if(!extended_inp)). This is addressed in the the wolf123 diff.

dipports is still used, it just hasn't needed changing since 0.113. mamerand was removed, when the rand() fixes were applied to baseline MAME.

Yup, I noticed that order didn't matter - but having separate diffs for the same file will obviously cause offset warnings either way, and I wondered whether that was necessary or otherwise...
As per the extended_inp, that's exactly what I changed - I just expressed myself sleepily.

So - are you interested in my 0.123 diffs ? I see mahlemiut.marpirc.net doesn't carry 0.123 yet

It's there, check the MARP news page for the correct URL. I guess I forgot to update the HTML. That's what I get for coding at 2am.

EDIT: update done, I did the HTML, but simply forgot to upload it. Go figure. You can always get a list of the binary and source packages at http://wolfmame.marpirc.net/

heh heh - I can figure that one out

will check those when I route around the customer proxy filtering .zip downloads - after lunch !

Thanks, ciao

mahlemiut wrote:dipports is still used, it just hasn't needed changing since 0.113. mamerand was removed, when the rand() fixes were applied to baseline MAME.

this goes with out mentioning but i might as well mention it :) If you are still using dipports113.zip in wolfmame123.exe, you [Barry] REALLY need to add it to the wolf123.diff file. And i would also recommend adding inpview to wolf123.diff too or at least adding both files to wolf123.zip for a true complete wolf source package. Alessandro is making versions that proly aren't including the dipports because he would have never known to look. I personally don't mind btw since my build scripts now incorporate all three files, but new people will get confused and just download wolf123.zip to make a wolf program that wouldn't be "official" and potentially have discprenacies.

What discrepancies? Each patch has distinct features, and none should affect anything else. No drivers are changed or altered in any way. And remember, the ability to playback wolfmame INPs is a standard feature of the baseline MAME.

different source sometimes no matter where it is could cause a discrepancy in playback. you "say" you are still applying dipports in wolfmame and that's not in wolf123.zip == sort of against the mame distribution source "full extent of the jam" law, even though it's in the same directory it's not realy implied.

I don't get what you're on about... All the source patches are available, and what goes into the binary is listed.

"This is the official binary recommended for use for MARP. It includes the extended INP support, input viewer, and -listxml update listed below."

I don't see why you can't see that dipports is not listed in wolfmame.txt (in the distribution file) or anywhere, except this thread. That's just the kind of loophole people get angry about not designating code you are compiling into mame and have a stupid fuss about. I'm just trying to avoid a stupid mame source violation by someone not smart enough to browse a directory. Anyway dipports0108.zip listed on wolfmame.marpirc.net tells anyone sensible that dipports was last used in wolfmame108, and not used in the current version, but from what you said it's used in the all the versions including wolfmame123. The same is NOT true for mamerand112 which probably was last used around 115 when mame added it.

Either say you are not using dipports or include the dipports source in the main current source distribution. It's easier than arguing.

What is compiled into the binary is listed on http://mahlemiut.marpirc.net/ which linked to on at least some of the news posts I make on MARP (certainly the latest was). dipports hasn't been updated since 0.113 simply because it hasn't needed to be.

cool that works, forgot about that site. all was always well.

And indeed I always included all three patches since building my SDLmame binaries:

wolf1xx.diff
inpview12-1xx.diff
dipports113.diff

I think it was clear enough from the pointers present on the SDLmame pages to wolfmame.marpirc.net, though I can see how someone who didn't read accurately before patching might think that "wolfmame patches" translates to wolf1xx.diff only... but I _do_ RTFM