irclog2html for #brlcad on 20070124

00:02.17``Eriksimh doesn't seem to support the old 704 :/
00:06.51``Erikhuh, ancient sumer used base 6 math, and that's why our arbitrary day time crap is in multiples of 6 (60, 24)
00:10.23``Erikhttp://www.marginalpeople.com/base%20six.html
00:11.17``ErikDAMN simpsons for making me learn!!!
00:17.20*** join/#brlcad buttdude (n=cfd26a68@bz.bzflag.bz)
00:19.33louipcdid they only have six fingers or something?
00:22.12``Erikheh, I think they did it all on one hand
00:22.14``Erikwith no concept of 0
00:22.23``Erikso closed fist is 1, ...
00:22.35``Erik<-- points at the url
00:23.07``Erikwas watching simpsons, skinner made a comment about finally being able to buy math books without base six stuff... so I had to g00gle
00:23.09``Erik:D
01:01.14brlcadheh, i think i remember that episode
01:01.28brlcadis that the cromulent one?
01:08.16``Erikno, the one where the dogs gonna die from a twisted stomach
01:23.33brlcadahh yeah
02:02.20Maloeran``Erik, I really thought they used a base of 12
02:05.43*** join/#brlcad Rangar (n=dave@203.118.156.252)
02:06.46MaloeranWoah, what are you doing here Hal? :)
02:07.01``Erikhehehe
02:07.34RangarMal; I'm here to whip you guys in to shape ;)
02:08.01``ErikI d'no if this is the right place for that kinda weirdassed kink
02:08.06``Erikalso; none of us are sheep.
02:08.48Rangarhahaha
05:39.34*** join/#brlcad cadguy (n=cadguy@c-69-250-146-28.hsd1.md.comcast.net)
06:36.45*** join/#brlcad clock_ (i=clock@84-72-89-136.dclient.hispeed.ch)
08:52.57*** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch)
10:50.52*** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch)
13:59.51*** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch)
14:02.57*** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch)
14:52.07*** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch)
15:48.27CIA-5BRL-CAD: 03erikgreenwald * 10brlcad/configure.ac: Add another X11 test to verify that the X11 library is actually linking in.
15:57.54brlcadthat's not evil!
16:13.05brlcadhm.. that's odd -- the test fails, but _XNewModifiermap is in /usr/X11R6/lib/libX11.a
16:19.19``Erikon some machines, xmkmf goes funky so it doesn't build the X_LIBS variable right
16:19.31``Erikrhel64 is a prime example
16:20.08brlcadi'm not complaining, as I'd been trying to find an x11 test that correctly fails on os x
16:20.47brlcadthe base system provides headers and stub libs that effectively all work .. except at run-time, any X11 call causes a crash
16:21.07brlcadI hunted/tested dozens
16:21.16``Erikurm, "a crash"?
16:21.34brlcadyeah, app would crash on the x11 call
16:21.54``Erikerm, does it matter if X11.app is running?
16:22.18brlcadthat's what I mean, a system that doesn't have x11.app (it's not installed by default on some/most/new systems)
16:22.32brlcadbut it does have the headers and libs so everything links
16:22.38``Erikah
16:22.54brlcadwhat's odd is that your link test for _XNewModifiermap fails
16:23.11``Erikcould always test if uname is darwin and hunt for X11.app *shrug*
16:23.15brlcadbut it actually is in the library, at least I see it in nm
16:24.35brlcadi was entirely trying to avoid a system-specific test
16:24.35``Erikhm, does the X_LIBS line look decent?
16:24.35brlcadyeah, looks right: X_LIBS='-L/usr/X11R6/lib -lX11 '
16:24.35``Erikhrm, *shrug*
16:24.41``Erikmebbe something to poke on after lunch... ya working from home today?
16:26.00brlcadnoir:~/brlcad morrison$ nm -g -o /usr/X11R6/lib/* 2>/dev/null | grep _XNewModifiermap
16:26.04brlcad/usr/X11R6/lib/libX11.a:ModMap.o: 000001ae T _XNewModifiermap
16:26.29brlcadfrom home all week
16:27.29brlcadagain, I'm not complaining.. I need/want the test to fail :-)  .. it's just entirely non-obvious why it's failing (which might mean it'll fail somewhere it shouldn't)
16:28.20``Erikhum
16:28.39``Erik(is nm an alias for otool -t?)
16:28.49``Erikor, -T, rather
16:29.27brlcader, not really .. it's the traditional nm
16:30.05``Erikyou got both my cphone # and lees, right? so you can verify dest once we figure it out? if you want to grab lunch...
16:30.18brlcadah, I just ate
16:30.22``Erikaight
16:30.27``Erik*shrug* later, dude :D
16:30.30brlcadlater
17:21.44CIA-5BRL-CAD: 03brlcad * 10brlcad/configure.ac:
17:21.44CIA-5BRL-CAD: restructure the X11 checks so that if --without-x11 is provided, all of the
17:21.44CIA-5BRL-CAD: associated x11 checks are also disabled as well so we don't waste time testing
17:21.44CIA-5BRL-CAD: for things that will not be used. similarly, don't auto-add /usr/X11R6 to the
17:21.44CIA-5BRL-CAD: search paths or add the extra restart paths unless we're in automatic or yes
17:21.47CIA-5BRL-CAD: mode
17:49.53CIA-5BRL-CAD: 03brlcad * 10brlcad/configure.ac: similarly disable all the java checks if we're --without-java
18:00.31CIA-5BRL-CAD: 03brlcad * 10brlcad/configure.ac:
18:00.31CIA-5BRL-CAD: clarify the license detail that this particular file, as part of the build
18:00.31CIA-5BRL-CAD: system, is BSD licensed. the lgpl was used before since this file is often
18:00.32CIA-5BRL-CAD: encountered first, but instead use the correct header and include a note about
18:00.32CIA-5BRL-CAD: how the license applied, referring to COPYING for more details.
18:30.37``Erikahhhh
18:41.31CIA-5BRL-CAD: 03brlcad * 10brlcad/configure.ac: similarly disable opengl/wgl checks if they are disabled via arg
18:49.22CIA-5BRL-CAD: 03brlcad * 10brlcad/TODO: add an entire new section for documentation, listing various documents that are written or planned. include notes and status for some that are already completed but perhaps not yet in CVS
19:16.04CIA-5BRL-CAD: 03brlcad * 10brlcad/configure.ac: stray "
19:33.39CIA-5BRL-CAD: 03brlcad * 10brlcad/src/librt/importFg4Section.c: copyright was acquired (by non-gov't) and later assigned to the government prior to open sourcing. remove old clause
19:37.27CIA-5BRL-CAD: 03brlcad * 10brlcad/src/librt/g_grip.c: seems to be a header that didn't update correctly -- accept implicit assignment of copyright via former brl-cad agreement, while retaining authorship details even though GSI is sadly now defunct.
19:45.12``Erikhttp://www.mirrorimage.com/air/page04.html
19:49.54dtidrow_workbrlcad: any good at diagnosing X11 problems?
19:50.05dtidrow_workor anybody else, for that matter
20:13.09``Erikwhat kinda 'problems'?
20:19.15*** join/#brlcad clock_ (i=clock@84-72-94-44.dclient.hispeed.ch)
20:20.15MaloeranAhah, online software engineering degrees are so depressing.  Introduction to Unix, HTML, XML, SQL, Basic, .NET? What the...
20:24.27dtidrow_work``Erik: it's dying deep down in an X call - XrmStringToQuark
20:24.27dtidrow_worksegfaulting
20:24.59MaloeranI thought I knew Xlib, but I don't remember seeing that function
20:25.00dtidrow_workwhich is about 13 call levels removed from the top X call  :-\
20:25.23dtidrow_workMaloeran: I'm guessing its very deep down in the bowels of xlib
20:26.25dtidrow_workthis seems to be rather old code, so the libs may have changed enough to break previously valid assumptions
20:28.05``Erikhuh, is the string being passed to it legit?
20:28.58dtidrow_workcan't tell - no decent debug info available that deep down
20:29.49``Erikhum, can't set a breakpoint at it and look at the argument?
20:30.01Maloeran*nods* It shouldn't segfault unless the address passed to it is unreadable
20:30.13dtidrow_work``Erik: nope, afaict
20:30.52dtidrow_workI can change the stack frame to that calling function, but there's no debug info at all available
20:30.59dtidrow_workat least with the libs I have installed
20:31.26dtidrow_workmeaning I can't see the function arguments
20:31.41MaloeranWhat's this code you are running? Do you have the source code?
20:31.43``Erikthere's only one argument, it'll either be on the stack or a register (depending on os and arch)
20:31.59``Erikonce you have the address, any debugger SHOULD be able to look at it
20:32.08``Erikor you can see if it even makes sense
20:32.25``Erikif you happen to know where heap and stack tend to be on that os/arch
20:32.46dtidrow_workgdb on x86
20:33.01dtidrow_worklinux/x86, that is
20:33.04``Eriklinux and fbsd are pretty different there with function call convention, last I looked
20:33.06``Erikaight
20:33.39``Erikiirc, linux/x86/gcc likes to try to do a register fill with arguments before jumping o.O
20:33.59``Erikbut my info might be outdated or just plain wrong, not exactly something I muched with much
20:34.01MaloeranArguments are passed on the stack on x86/Linux
20:34.18``Erikoh, fully on the stack? mebbe I'm thinking syscalls (int 0x80)
20:34.38``Erikok, so examine the stack, find the ptr you want, ...
20:34.45dtidrow_workyeah, there aren't too many available registers with x86  :-\
20:35.20``Erikthere are 8 gp registers.. but, uh, none of them are actually general purpose :(
20:35.32dtidrow_workhere's the call stack:
20:35.35dtidrow_work#0  0x05c1c731 in XrmStringToQuark () from /usr/X11R6/lib/libX11.so.6
20:35.35dtidrow_work#1  0x005ad1cb in XtSetTypeConverter () from /usr/X11R6/lib/libXt.so.6
20:35.35dtidrow_work#2  0x003021a3 in _XmRepTypeInstallConverters () from /usr/X11R6/lib/libXm.so.3
20:35.36dtidrow_work#3  0x0027f31d in _XmRegisterConverters () from /usr/X11R6/lib/libXm.so.3
20:35.36dtidrow_work#4  0x0030e926 in _XmScreenRemoveFromCursorCache () from /usr/X11R6/lib/libXm.so.3
20:35.36dtidrow_work#5  0x005b37c5 in XtInitializeWidgetClass () from /usr/X11R6/lib/libXt.so.6
20:35.38dtidrow_work#6  0x005b37f7 in XtInitializeWidgetClass () from /usr/X11R6/lib/libXt.so.6
20:35.39MaloeranAh, it's all on the stack.  On Linux amd64, it's rdi, rsi, rdx, rcx, r8, r9 ; xmm[0..5], xmm12, xmm13
20:35.42dtidrow_work#7  0x005b37f7 in XtInitializeWidgetClass () from /usr/X11R6/lib/libXt.so.6
20:35.50dtidrow_work#8  0x005b3c91 in XtInitializeWidgetClass () from /usr/X11R6/lib/libXt.so.6
20:35.52dtidrow_work#9  0x005b3f86 in _XtCreateHookObj () from /usr/X11R6/lib/libXt.so.6
20:35.54dtidrow_work#10 0x005b447b in _XtAppCreateShell () from /usr/X11R6/lib/libXt.so.6
20:35.58dtidrow_work#11 0x005e0e93 in XtVaAppCreateShell () from /usr/X11R6/lib/libXt.so.6
20:35.58dtidrow_work#12 0x005e1342 in _XtVaOpenApplication () from /usr/X11R6/lib/libXt.so.6
20:36.00dtidrow_work#13 0x005e148e in XtVaAppInitialize () from /usr/X11R6/lib/libXt.so.6
20:36.02dtidrow_work#14 0x08050e7b in setupWindow (argc=@0xbfde44e0, argv=0xbfde4564, model=0x82d65f0) at main.cc:96
20:36.04dtidrow_work#15 0x08050f94 in main (argc=1, argv=0xbfde4564) at main.cc:121
20:36.19dtidrow_workindeed - x86_64 is much better with regards to registers
20:36.22``Erikyeah, that's a backtrace, no the stack... it's generated by mixing 3  different things and throwing away a good bit of data...
20:36.45dtidrow_work``Erik: yeah, I liked MIPS
20:36.57dtidrow_workpity SGI pissed away their lead...
20:37.11``Erikheh, I was just reading about that earlier, x86-64 is no longer the name, intel has officially changed the name to be amd64, since it's a reimplementation of amd64's ISA...
20:37.34MaloeranI thought Intel opted for x64
20:37.50``Erikjra was asking me why his x86-64 desktop started calling itself amd64
20:37.55MaloeranMicrosoft refers to the architecture as x64, the open-source world refers to it as amd64
20:37.56dtidrow_workintel had it as EM64T, or something like that - just to avoid having to call it amd64  :-)
20:38.17dtidrow_worklast I heard, anyway
20:38.37``Erikhttp://en.wikipedia.org/wiki/EM64T
20:38.53dtidrow_workit's essentially intel's clone of amd64, though  :-)
20:39.10brlcaddtidrow_work: seems to be failing in the app setup, which takes several strings... passing any command line args? might try changing them and/or making sure something's not wrong with the input
20:39.28dtidrow_workAMD had better get its act in gear - Core is kicking Athlon a**
20:40.03``ErikXrmStringToQuark() shouldn't care what the string is as long as it's well formed... it LOOKS like a basic symbol table dealie
20:40.05dtidrow_workbrlcad: tried that - dies elsewhere due to a differently formatted data file
20:40.30brlcadthe fact that it's crashing deep isn't as important as it crashing at XtVaAppInitialize() from the app's perspective
20:41.10dtidrow_workMaloeran: and I'm pretty much an AMD fan as well, esp. the Opterons with their distributed memory architecture
20:41.15brlcadthat's one of the very first x calls it's probably making.. is the xserver actual running? :)
20:41.40dtidrow_work;-)
20:41.44brlcad:)
20:42.04``Erikheh, 'cept half the time, silly questions get answered with "oh, uhhhh.... n/m"
20:42.17Maloerandtidrow_work, I wouldn't say these Core Duo are quite stomping Opterons. You can't even put two of the chips on the same motherboard
20:42.27brlcadwell, I've seen even the most experienced dev try to kick off X11 apps from a Mac Terminal, localhost, forgetting to set display and/or xhosts
20:42.35dtidrow_work``Erik: that one was one step removed from 'is the power on?'
20:42.39brlcadso not entirely unheard of
20:42.51``Erikhum
20:42.54``Erikdon? is the power on?
20:42.57MaloeranPlus, Opterons are ahead in the very important performance/cost ratio
20:42.57``Erik:D
20:43.39``Erikbut intels chips are the ones being picked up by h^Hdell and apple
20:44.04MaloeranI thought Dell had switched to AMD, or perhaps it was HP
20:44.19dtidrow_workthey did, then the Core came out...
20:45.08dtidrow_workthe Core's perform better than equivalently-priced Athlons (apparently)
20:45.16brlcaddtidrow_work: another possibility.. linking in the wrong library
20:45.31brlcadrather suspicious that it's the first Xt call
20:45.52brlcadldd show anything useful?
20:46.07dtidrow_workI was looking at that, but don't have any experience with straight X11 stuff
20:46.58brlcadhow about just the basics.. ldd the binary and see if there are any 64-bit's being mixed with 32-bits and vice-versa
20:47.16dtidrow_workwell, I compiled the binary, and I don't have a 64-bit capable system
20:47.24dtidrow_workyeah
20:47.30brlcaddepends on the OS and compiler
20:47.31dtidrow_workmy thoughts exactly
20:47.37brlcadsome will let you
20:47.44brlcadand just crash at run-time
20:47.48brlcad(fun)
20:47.53``Eriklinux/x86/gcc, I think the rtld at least has the decency to check there
20:47.57dtidrow_worknot gcc/linux, in my experience
20:48.10brlcadyeah, but I wasn't making assumptions ;)
20:48.16dtidrow_workheh
20:48.24brlcadso this is linux
20:48.24``Erikbrlcad: I asked him his platform earlier.. :D
20:48.30brlcadah, i missed that
20:50.33brlcaderm, you compiled this yourself but can't step through in gdb?
20:50.55MaloeranThe libraries must not have debugging symbols I guess
20:50.57dtidrow_worknot that deep in the X libs
20:51.05dtidrow_workMaloeran: exactly
20:51.11brlcadsure, but again that's fairly irrelevant
20:51.26dtidrow_workI've got the params to XtVaAppInitialize
20:51.34brlcadfrom the app's perspective, something is wrong with the XtVaAppInitialize() call inputs or something that preceeds it
20:51.40dtidrow_workddd'ing it now  :-)
20:51.48``Erikheh
20:51.51``Erikgui dork
20:52.04brlcadddd has pretty struct graphs
20:52.21dtidrow_worknice data displays, too
20:52.55dtidrow_workah, same thing
20:53.20``Erikbut it's been over 10 years :(
20:53.31brlcad``Erik: it did :)
20:53.31dtidrow_workfor ddd?  yikes
20:53.49brlcadi last used ddd about that long ago too
20:53.55brlcad(at least seriously)
20:54.10MaloeranAh, I was writting my first lines of code 10 years ago
20:54.34brlcadheh
20:55.09dtidrow_workk, 'fess up
20:55.13dtidrow_workhow old?
20:55.19``Erik24 yrs since my first line of cbm basic2.0
20:55.29MaloeranAhah, woah
20:55.31dtidrow_worklol
20:55.42``Erikor, wait, no
20:55.49``Erikthat was on a coleco adam
20:55.50dtidrow_work``Erik: so about 50, eh?  ;-)
20:56.19MaloeranSo now Erik can say he has been coding from before I was born
20:56.35``Erikheh, no :D awz a young'n when I got my first machine... and, uh, it's not like you could run down the street to the software store in '83
20:57.11dtidrow_workyeah, remember those days - had to type programs in from 'Compute' magazine  :-)
20:57.45``Erikuh huh, my dad bought a coleco adam and a book called "games apples play" that was just listing after listing of trivial basic programs
20:57.56``Eriksome of 'em worked without modification on the adam
20:59.00``Erikoh, I had buck rogers on cassette, too, and a few others
20:59.19MaloeranEheh, cool
21:01.29MaloeranI had a C compiler and source code from Autocad. I think learning to code would have been easier with some kind of book or reference
21:09.32dtidrow_workfrom XtVaAppInitialize man page: "XtAppInitialize and XtVaAppInitialize have been superceded by  XtOpenApplication  and  XtVaOpenApplication  respectively."
21:09.54dtidrow_workinteresting - maybe I just need to change the code to this
21:15.20``Erikamusing... amazon is selling this book from between $250 and $450 used... but the entire book is available as a pdf for free on the authors website
21:15.34``Erik(On Lisp - Paul  Graham)
21:16.03dtidrow_workheh
21:44.18dtidrow_workwell, I'll have to futz with this more at home....
22:06.00*** join/#brlcad docelic (n=docelic@212.15.184.89)
22:20.43*** join/#brlcad docelic (n=docelic@212.91.113.204)
23:05.40*** join/#brlcad docelic (n=docelic@212.91.113.204)
23:40.22*** join/#brlcad Rangar (n=dave@203.118.156.252)
23:40.23Rangarphew, oh boy
23:43.27brlcadindeed
23:45.48``Eriktoxic flatulence?

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.