00:00:01: <Doxin> that seems like less work
00:00:25: <bvowk> heh
00:00:39: <impomatic> The assembler is tricky because of macros and multi-line equ
00:00:46: <bvowk> iirc the exmars code to use the pmars parser wasn't bad?
00:01:00: <Doxin> impomatic: I know
00:01:02: <bvowk> You just needed to feed it the code and then convert?
00:01:22: <Doxin> bvowk: well I was trying to make bindings for exmarsd
00:01:28: <Doxin> but that thing is a right mess
00:03:13: <Doxin> it seems like everything has just copy-pasted code from pmars and I'd like to not do that as well.
00:03:26: <bvowk> I'd personally recommend you try putting in the effort to pull that pmars parser into yours, as it's the standard..
00:03:46: <bvowk> And making an assembler that works and breaks in the same way is hellish.
00:04:12: <Doxin> the problem isn't so much wrangling that parser to work
00:04:14: <bvowk> Look at what PyCorewar had to do, and it's not quite perfect.
00:04:25: <Doxin> the problem is moreso that that'll leave my code an utter mess
00:04:33: <impomatic> Unfortunately there's macros like this (and worse) to content with in published warriors:
00:04:38: <impomatic> for 20+!(r=1)*(x=qfac)*(n=power)
00:04:39: <impomatic> for n%2
00:04:41: <impomatic> for !r=(r*x)ORESIZE
00:04:43: <impomatic> rof
00:04:44: <impomatic> rof
00:04:46: <impomatic> for !x=(x*x)ORESIZE+!n=n/2
00:04:47: <impomatic> rof
00:04:49: <impomatic> rof
00:05:35: <Doxin> eh
00:05:45: <bvowk> I wonder if we should update the standard to allow for fewer degrees of freedom?
00:05:51: <Doxin> I'll see where I give up I guess.
00:06:07: <Doxin> bvowk: as long as all the pmars oddities are accounted for in the spec it should be fine.
00:06:08: <bvowk> And then convert the scary published warriors so we can keep them going ahead.
00:09:04: <bvowk> Maybe someone should just move the pmars parser into a library we publish separately :)
00:14:54: <Doxin> :P
00:15:24: <Doxin> if I ever finish this thing I'll make sure to slap it on github with a reasonable license
01:51:33: -!- Doxin has quit (Remote host closed the connection).
11:51:08: -!- Doxin has joined #corewars.
12:00:19: -!- Payhn has quit (Quit: ZNC - http://znc.in).
12:02:26: -!- Payhn has joined #corewars.
15:05:56: <Doxin> impomatic: there's something awry in the spec as shown on http://www.koth.org/info/icws94.html
15:06:13: <Doxin> line 0196 is wrong. field should probably be expr there too?
15:15:43: <Doxin> impomatic: also line 0218 seems wrong
15:17:36: <Doxin> maybe it's time for a '16 draft.
15:46:22: <impomatic> Thanks, I'll take a look tonight and fix it if necessary
18:05:39: <bvowk> hrm.
18:52:14: <Doxin> well I just managed to produce 8MB of backtrace. this is going to be "fun"
19:03:47: <bvowk> That's it?
19:19:51: <Doxin> bvowk: it's massive
19:20:07: <Doxin> bvowk: also it is inside a mixin which makes it doubly painful
19:22:46: <bvowk> heh
19:23:03: <bvowk> You've got some untangling to do.
19:24:10: <Doxin> I sure do :P
19:28:49: <Doxin> yeah this is impossible to untangle :I
20:17:19: <impomatic> Doxin: I see what you mean about 0196 and 0218.
20:19:28: <impomatic> There's quite a bit missing from that grammar as well.
20:19:30: <impomatic> Doxin:
20:19:47: <impomatic> { } * addressing modes.
20:21:39: <impomatic> ! < > == and probably some other stuff in expressions.
20:24:21: <impomatic> SEQ SNE NOP LDP STP opcodes
20:30:37: <impomatic> ORG EQU END are pseudo-ops and should have a separate entry
20:39:16: <impomatic> Inversed just claimed first place on the 94 and 94nop hills with Overseer :-)
20:46:25: <Doxin> impomatic: I guess it's pretty much moot if I can't get this darn parser generator to not segfault
21:08:00: -!- glogbot has quit (Ping timeout: 260 seconds).
21:08:08: -!- glogbot has joined #corewars.
21:12:17: <bvowk> damn.
21:12:28: <bvowk> Inversed is on a tear.