freenode/#mezzano - IRC Chatlog
Search
21:04:38
froggey
I've reworked IRQ handling and reenabled the HDA driver. audio playback should work again and not break virtualbox integration
21:20:14
froggey
the disassembler's main function (disassemble-subfunction) operates on compiled functions, objects tagged with +object-tag-function+
21:20:52
froggey
it's any function that isn't a closure or a funcallable instance (generic functions, interpreted functions, etc)
21:23:28
froggey
you need to use (sys.int::function-code-size function) and (sys.int::function-code-byte function index) to read a compiled function's machine code
21:27:07
froggey
it should actually remove any deleted files in the given directory, like emptying the recycle bin. I haven't thought about it beyond that
21:28:29
housel
I'm working on adapting quicklisp, and that's one of the operations it wants to be able to do
21:30:14
froggey
I think delete-file with a directory argument should cover that. the hyperspec isn't very clear here
21:32:31
housel
quicklisp uses delete-file as its default implementation, but quite a few platforms have another function for directories
21:37:14
froggey
ebrasca: you can bind sys.c::*trace-asm* to T or :FULL and the compiler will print what it's passing to the assembler. I used that and objdump/other disassemblers while developing most of the system
21:37:51
froggey
the native disassembler is a very recent addition, though it's very nice to have and I use it a lot now
2:31:36
fittestbits
sorry about the emoji - host: device>top-level-dir>... (without the space between the ":" and "device")
3:06:56
fittestbits
I guess those brain cells died - don't remember that. Didn't even remember using > instead of / - its been a long time ...
3:14:13
p_l
in this case the syntax is more from Genera, but I suspect it might have been there since MIT CADR system 76
3:18:15
p_l
Symbolics first made LM-2, which was AFAIK upgraded CADR that was fully compatible but more reliable, then the software evolved and got ported to 3600, and finally I think around "System version 6" or so it got named "Genera"
3:19:54
fittestbits
Oh, Symbolics, I see. Didn't follow their progress after the 3600. AI and lisp machines got over run by PCs from below and unix workstations from above.
3:25:14
fittestbits
Seemed to me that at that time, PCs were under powered to run lisp with enough performance to compete with programs in C, plus, the people doing AI in lisp didn't have a killer app.
3:32:35
p_l
fittestbits: an application that was credited with repaying (with interest) every cent DARPA ever spent on "AI"