The above video goes away if you are a member and logged in, so log in now!
Heh, yeah, . Still discovering new things, even today, actually, lol.
Snes9x kept randomly crashing, every other compile, when I was trying to debug the input code. I debugged it to a line in newlib's dlmalloc-ps2.c, for free(), which is weird, since ps2sdk has its own free(), malloc() and such which have been tested and are mostly bugfree.
A lot of the libc symbols in ps2sdk were defined with "__attribute__((weak))" which essentially means they get redefined if another symbol is found to be "stronger" like a globally defined symbol in newlib. I had to modify newlib to make its symbols weak, then remove the weak attribute from ps2sdk's symbols.
I guess I was using the newlib symbols before, but wasn't triggering whatever bug was causing that crashing. Back to the input debugging, .
I'd like to compare Rag to steam here.
They keep on delaying their products to fix bugs, again and again, and it IS necessary, (that's why we've never seen a "black ops PS3"-like incident with steam games) but HOLY HELL do they PWN all they surround when they finally release their product.
Do all the bugfixing you need for this, Rag, it's still awesome you're so devoted to this.
Or half life EP3
Originally Posted by personuser
Uh...yeah, guys...I dunno why I even posted that last night...ya :P it was late and I wasn't thinking :P
Originally Posted by henshinzhu
a DOH!!! moment there homer?
128 installs , last on 24-May-2013
PS2 50002 PAL Silver
+ Network Adaptor + 500 Gig
Hitachi IDE + FMCB 1.8b
PS2 79002 PAL
Slim Unmodded FMCB 1.8b
black CFW 6.60 B10 PhotoFast 5400 with 2 x 8 Gig micro SDHC
Yes, perhaps :P
Originally Posted by spud42
I have the input problems figured out, I think. I'd messed up the mapping IDs a bit, so input wasn't reporting the right ID numbers that had been mapped. I'd left a couple of lines that only reported new button presses, when all button presses should be reported, since Snes9x does the input handling... I'm not quite sure why those lines were there. Also, it looks like mapping the directional buttons and directional axes at the same time isn't possible, so I need to set the default mapping to "None" for the axes. I still need to test if loading a mapping works and if I can close and reopen a controller between the GUI and Snes9x without any problems.
Is this a problem in the configuration system?
Originally Posted by ragnarok2040
In Rockbot I am using both the analog (axes) and direcional (hat) works at same time without problem.
It's a problem somewhere in the mapping code in Snes9x, I think. Mapping the axes might override their button counterparts, e.g. Axis Up/Down overrides Up and Down mappings. It's also possible that reporting the position of the analog stick is enough to cancel out button input even if it's centered.