IRC log for #brlcad on 20100108

00:07.35 ``Erik "Anaesthetists - They do it until you fall asleep"
00:22.34 *** join/#brlcad R0b0t1 (n=Enigma@unaffiliated/r0b0t1)
01:02.42 *** join/#brlcad Ralith (n=ralith@69.90.48.97)
02:28.58 *** join/#brlcad dtidrow (n=dtidrow@c-71-238-51-148.hsd1.mi.comcast.net)
02:49.52 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2162 10/wiki/MGED_CMD_rpp-cap:
02:51.11 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2163 10/wiki/MGED_CMD_decompose:
02:53.01 *** join/#brlcad dtidrow (n=dtidrow@c-71-238-51-148.hsd1.mi.comcast.net)
02:55.36 Nohla ``Erik or somebody, what means stem?
02:57.51 ``Erik in what context?
02:58.50 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2164 10/wiki/MGED_CMD_bot_decimate:
03:00.18 Nohla sorry, in the 6th tutorial, f.e., it's say:
03:00.31 Nohla To make the stem a region, type at the Command Window prompt:
03:00.32 Nohla
03:00.32 Nohla <command>r stem1.r u ball1.s u ball2.s u ball3.s[Enter]</command>
03:01.20 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2165 10/wiki/Category:MGED_bot_operators: New page: Commands that operate on [[BoT|bags of triangles]] [[Category:MGED object generators]] [[Category:MGED]]
03:02.48 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2166 10/wiki/Category:BoT_operators: New page: Commands that operate on [[BoT|bags of triangles]] [[Category:MGED object generators]] [[Category:MGED]]
03:03.14 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2167 10/wiki/Category:MGED_bot_operators:
03:03.50 ``Erik ah, a thin connecting piece, umm, like with a wine glass, the piece between the bulb and the base
03:03.54 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2168 10/wiki/MGED_CMD_bot_decimate:
03:04.36 Nohla okok
03:05.14 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2169 10/wiki/MGED_CMD_bot_condense:
03:05.18 ``Erik which one is tutorial 6?
03:05.18 Nohla ``Erik the writing tides a bit :)
03:05.31 Nohla create a globet
03:05.33 Nohla :)
03:05.47 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2170 10/wiki/MGED_CMD_bot_face_sort:
03:05.48 ``Erik ah, hah, yes, my example was apropos!
03:05.57 Nohla you guessed :)
03:06.19 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2171 10/wiki/MGED_CMD_bot_vertex_fuse:
03:06.51 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2172 10/wiki/Bot: Redirecting to [[BoT]]
03:07.25 ``Erik http://monstermugs.co.uk/images/aaaglDOMINMA.jpg this glass has a black stem and a black base
03:08.43 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2173 10/wiki/MGED_CMD_bot_face_fuse:
03:08.54 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2174 10/wiki/Category:MGED_BoT_operators: New page: Commands that operate on [[BoT|bags of triangles]] [[Category:MGED object generators]] [[Category:MGED]]
03:09.31 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2175 10/wiki/Category:BoT_operators:
03:10.48 CIA-38 BRL-CAD: 03Ssd 07http://brlcad.org * r2176 10/wiki/Category:MGED_bot_operators: oops someone delete this
03:16.52 Nohla ``Erik thanks
03:17.09 Nohla I'm near finishing :)
03:18.43 ``Erik awesome :D
03:22.35 Nohla well, finishing 6th, but 3rd, 4th and 5th were jumped, jejejje
03:22.55 Nohla a friend helped me with this
03:23.29 Nohla chosed 6th just in case he take too much time to do it
03:29.46 *** join/#brlcad akafubu (n=akafubu@unaffiliated/akafubu)
03:30.27 ``Erik it is still another translated document... part of the first translation set for BRL-CAD :)
03:32.44 Nohla shure
03:33.08 Nohla hope to advance quickly this month
03:33.24 Nohla february is a coplicated month to me
03:34.36 Nohla ``Erik could you help me by explaining the meaning of "framebuffer"
03:34.57 ``Erik um, do you know much about how computer graphics work?
03:35.23 Nohla this context: If
03:35.23 Nohla you want to view the goblet without the wireframe, go to the Framebuffer
03:35.23 Nohla option of the Raytrace Control Panel and click on Overlay.
03:35.45 ``Erik that's referring to the window on the screen that displays the raytraced image
03:35.52 Nohla well, read something today :P may that help
03:36.25 ``Erik doesn't think any tutorials refer to other aspects of it, it's just the output window for the raytracer
03:36.49 ``Erik does that make sense?
03:37.35 Nohla is it like shape?
03:37.50 Nohla of objets
03:38.30 ``Erik no, the shapes exist in the geometry, but we do raytracing to make a nice picture of it, the results of the raytracing pops up in a new window (or glued into the geometry editing window)
03:39.33 ``Erik http://brlcad.org/gallery/s/screenshots/gnu_tux.png.html
03:39.57 ``Erik this shows the framebuffer on the left, command window on the bottom and geometry widnow on the right
03:40.38 Nohla but, it's like the silhouette
03:40.50 ``Erik it's the results of raytracing
03:41.09 ``Erik he ran the "rt" command to raytrace, it popped up a new window and filled it with the raytrace results
03:41.48 ``Erik http://brlcad.org/gallery/s/screenshots/ronja_screenshot.png.html <-- has the framebuffer on the top right
03:41.59 ``Erik embedded in the geometry window
03:44.26 ``Erik http://brlcad.org/gallery/s/screenshots/t62_mged.jpg.html actually is closer to the part your'e looking at, the treads (red lines) are visible because the framebuffer (the silver tank) is in Underlay mode, if you click Overlay, the red lines disappear
03:45.43 Nohla well
03:46.09 Nohla I can understand but can't tell it in spanish yet :)
03:46.11 Nohla jejej
03:46.22 ``Erik it's a strange functionality, perhaps we should mark that as 'for later' and brlcad can help figure out how to say it when he's online? :)
03:46.23 Nohla too tired I think
03:47.09 Nohla brlcad must be dead
03:47.38 Nohla He's never outline :P
03:48.12 ``Erik "offline", you mean? :)
03:48.46 Nohla yes XD
03:48.57 Nohla you see? toooooo tired
03:49.07 ``Erik I saw him in the office today, he may've decided to go home and finally sleep, or he's out carousing on town
03:49.11 Nohla ``Erik thanks again
03:49.26 ``Erik no problem, thank YOU for doing the translation :)
03:50.13 Nohla night
03:50.16 ``Erik goodnight
03:52.43 Nohla ``Erik framebuffer could be the drawing of the visible surface?
03:54.08 ``Erik it's... the buffer where the raytraced frame results are put... um, for raster images
03:54.34 Nohla well, that was my best bet for tonight, better try tomorrow
03:55.02 ``Erik um, in that example, they get the raytrace dialog box and push 'render', and it raytraces what's in the geometry window
03:55.19 ``Erik and it makes two "layers" in that window, one with the red wireframe, the other with the raytraced image
03:55.36 ``Erik underlay/overlay is controlling the relation of those two layers
03:56.01 ``Erik go sleep and think about it tomorrow :D
04:06.10 brlcad cool, so eto still has evaluation errors, but ebm is fixed
04:07.40 brlcad Nohla: fantastic progress! :) :)
04:09.40 brlcad was en route, now home
04:09.51 brlcad back to the codegrind!
04:10.10 ``Erik huh, don't recall seeing your car when I left, didja run out for food or something?
04:10.17 brlcad yeah
04:17.03 Nohla brlcad I was just going into bed
04:17.45 brlcad Nohla: hasta man~ana! .. que te duermes bien!
04:18.13 Nohla tomorrow night we could try to let me understand how to define blablabuffer (I can't remember word :P)
04:19.03 Nohla and, if you dont mind to jump some tutos, I'll bring 6th
04:19.29 Nohla but, until that moment, good night and day of tomorrow :)
04:20.00 Nohla sorry if you can't imagine wath I'm traying yo say :)
04:20.04 Nohla zzzzzz...
04:23.07 ``Erik framebuffer
04:23.40 Nohla ...I was talking about everything I was trying to say :)
04:24.31 Nohla real zzzz...
04:24.34 Nohla kisses
09:24.39 CIA-38 BRL-CAD: 03d_rossberg * r37170 10/brlcad/trunk/src/other/openNURBS/ (opennurbs_point.cpp opennurbs_quaternion.cpp):
09:24.39 CIA-38 BRL-CAD: including common.h has to be preferred to including brlcad_config.h directly
09:24.39 CIA-38 BRL-CAD: (otherwise the Windows build won't work)
11:18.27 *** join/#brlcad mafm2 (n=mafm@119.Red-81-34-12.dynamicIP.rima-tde.net)
12:17.18 *** join/#brlcad d_rossberg (n=rossberg@BZ.BZFLAG.BZ)
12:19.42 d_rossberg i got a brlcad build error (implicit declaration of function): http://pastebin.bzflag.bz/m1afc8bab
12:20.58 d_rossberg it looks like regex.h is included from tcl/generic instead of /usr/include
13:04.49 starseeker growls first week of the year and I'm out sick...
13:11.08 ``Erik yean, d_rossberg, I've been seeing that on several machines, too... there's some fix I did to tclInt.h a long time ago for that, I guess it was during an update... I'll have to dig up what I did
13:41.26 d_rossberg ``Erik: it could depend on whether libregex is build or not (i.e. the one from the system is used)
13:43.49 d_rossberg in the first case the libregex include path precedes tcl/generic, in the later case tcl/generic precedes the system include path /usr/include
14:21.08 ``Erik tcl itself has a regex.h that confuses the preprocessor, I'll look into it once I'm settled in here (just got to the office)
14:47.37 ``Erik hrm, it's there O.o odd
14:57.27 d_rossberg the regex.h in tcl/generic uses additional macros like __REG_NOFRONT
15:04.37 *** join/#brlcad docelic (n=docelic@78-2-71-58.adsl.net.t-com.hr)
15:31.57 brlcad starseeker: oof, sorry to hear that
15:40.25 brlcad d_rossberg: yeah, hrm!
15:40.43 brlcad if libregex were built, it would get the right header
15:41.11 brlcad but since it's using a system one, the -Isrc/other/tcl/generic is overriding ...
15:42.32 brlcad the solutions that are coming to mind are:
15:42.34 brlcad 1) make -I/usr/include a REGEX_CPPFLAG
15:43.30 brlcad 2) remove the __REG_NOFRONT/__REGNOCHAR so they're defined
15:44.05 brlcad 3) remove <regex.h> from regionfix.h and just declare the extern regex functions in use
15:44.39 ``Erik the one I was seeing was implicit declaration of regcomp/regexec/reg(somethingelse), but it's now showing up on any machines at the moment O.O
15:44.53 brlcad 3 is probably the simplest, but is of course a total "punt"
15:45.19 brlcad it should only happen with default configure options where it uses a system regex
15:45.24 brlcad --enable-all won't see it
15:45.40 brlcad rather, a system regex and a non-system tcl
15:46.24 ``Erik which is my standard mac build, which showed it, but now it doesn't, odd
15:50.01 brlcad my default mac build uses system tcl/tk
15:50.15 brlcad (didn't use to, but I fixed that last spring/summer)
15:52.08 ``Erik hm, I usually build with ./configure --enable-optimized --prefix=/usr/brlcad/HEAD --enable-tcl-build --enable-tk-build --disable-png-build CFLAGS=-I/opt/local/include LDFLAGS=-L/opt/local/lib
15:52.32 brlcad hm, dunno then
15:52.36 brlcad debug that shit up
15:53.01 d_rossberg however, there remains something unclear: #include <~> and #include "~" should differ by their search paths
15:53.27 d_rossberg #include <~> should search in the system directory first
15:55.51 brlcad d_rossberg: -I override system search dirs
15:55.53 brlcad at least for gcc
15:56.37 brlcad "" simply checks current dir first before checking -I and system dirs
15:56.49 brlcad an implicit -I.
15:57.40 CIA-38 BRL-CAD: 03brlcad * r37171 10/brlcad/trunk/src/other/tcl/generic/regex.h: change the file we have control over. according to the header, it's legit to remove the __REG_NOFRONT define so we get declarations of regcomp() and regexec().
15:58.29 brlcad that should at least fix the warnings even if it does get that file instead of a system one (and fortunately those function signatures aren't likely to change)
15:58.40 brlcad issue will just be tclInt.h getting included
15:58.58 brlcad but if it's getting that regex.h, it's already got -Isrc/other/tcl/generic
16:00.24 brlcad http://gcc.gnu.org/onlinedocs/gcc/Directory-Options.html <-- See second sentance under -Idir regarding override behavior.
16:01.51 brlcad there's probably some magic -isystem foo we could use, instead of -I for all src/other codes to get the desired search order behavior, but it'd be gcc-specific
16:04.27 CIA-38 BRL-CAD: 03brlcad * r37172 10/brlcad/trunk/TODO: inspiration
16:06.30 d_rossberg i'm using gcc's man-page here
16:07.26 d_rossberg the "-I-" and "-iquote" options are looking interesting
16:07.39 brlcad my man page has the same quote :) .. "Directories named by -I are searched before the standard system include directories."
16:08.18 d_rossberg maybe "-iquote" should be the standard rather than "-I"
16:08.55 ``Erik huh, pbcopy/pbpaste (mac only)
16:09.07 brlcad they're still searched before system dirs
16:09.38 d_rossberg right, but only for #include "~"
16:09.38 brlcad plus it's still go tthe issue of being gcc-specific flags
16:10.51 d_rossberg what other compilers do you have?
16:11.14 d_rossberg on windows it is a totally different story
16:11.16 brlcad icc and sunw
16:11.31 d_rossberg i see ...
16:11.37 brlcad aside just from underlying philosophy
16:11.42 ``Erik mipspro if we want to blow the dust out of an old clunker O:-)
16:11.49 ``Erik msvc
16:13.24 d_rossberg i think it is part of the C standard that include<> searches the system headers first and include"" the user provided headers
16:14.17 d_rossberg therefore you have to declare the system and user provided headers correctly in gcc
16:14.35 d_rossberg (and any other compiler)
16:14.52 brlcad would be more incined to put system dirs on the -I path (e.g., -I/usr/include) before adding the configure magic to detect that -iquote works and to use it instead of -I
16:15.16 brlcad in the absense of any compiler flags, that is gcc's behavior .. :)
16:15.33 brlcad spec has nothing on application options
16:17.18 brlcad did that commit fix the build?
16:17.41 ``Erik src/brlcad/src/librt/regionfix.c:151: warning: implicit declaration of function 'regfree'
16:17.45 ``Erik but the other two are fixed
16:17.53 brlcad hum
16:18.08 brlcad i checked regfree .. looked like it was already deeclared
16:18.14 ``Erik looked like, yes
16:18.41 d_rossberg overwritten by the TCL function ;)
16:18.48 ``Erik but not in the -E output
16:19.32 brlcad right, it should be a TclReFree()
16:19.55 brlcad sure it's object-clean?
16:20.13 brlcad hm, nvr mind :)
16:20.14 ``Erik just rm -rf'd his build dir and did an autogen.sh
16:20.20 ``Erik wow, 404 on pastebin
16:20.49 ``Erik http://paste.lisp.org/display/93148
16:21.01 ``Erik that's what the tcl regex.h resolves to out of -E on my mac
16:21.28 d_rossberg there is still one error left: implicit declaration of regfree
16:22.30 brlcad mm.. all the regex code in tcl includes regguts.h which includes regcustom.h which includes all those proper "tcl-style" declarations/defines .. not clear why regex.h has them at all
16:25.34 ``Erik including tclInt.h in regeg.h USED to fix it, odd that it stopped
16:25.54 CIA-38 BRL-CAD: 03brlcad * r37173 10/brlcad/trunk/src/other/tcl/generic/regex.h:
16:25.54 CIA-38 BRL-CAD: remove the entire tcl-protection block that was spliced in from tclcustom.h so
16:25.54 CIA-38 BRL-CAD: that we get regfree() declared. all of tcl's regex sources include regguts.h
16:25.54 CIA-38 BRL-CAD: which includes regcustom, so they should get their integrated behavior for their
16:25.55 CIA-38 BRL-CAD: built-in regex.
16:26.09 brlcad including tclInt.h fixed a different problem
16:26.23 brlcad these are warnings .. we were just ignoring the warnings before
16:26.27 brlcad not they're errors
16:26.40 brlcad that should do the trick
16:26.44 ``Erik hm, was years ago, I've forgotten O.o (and the history didn't transfer very nicely
16:27.00 ``Erik worked here
16:27.06 brlcad have to check tcl-runtime to make sure regular expressions still work, but tcl compiles clean .. hopefully regionfix.c compiles clean now too
16:27.46 d_rossberg regionfix.c compiles now :)
16:37.35 *** join/#brlcad jnewt4 (n=jnewt@ppp-70-252-130-22.dsl.ksc2mo.swbell.net)
17:50.20 *** join/#brlcad akafubu (n=akafubu@unaffiliated/akafubu)
18:28.38 *** join/#brlcad Ralith (n=ralith@d142-058-090-011.wireless.sfu.ca)
18:38.43 *** join/#brlcad Ralith_ (n=ralith@d142-058-090-011.wireless.sfu.ca)
19:20.22 CIA-38 BRL-CAD: 03brlcad * r37174 10/brlcad/trunk/src/rt/do.c: comment style formatting consistency cleanup
19:46.30 ``Erik huh, the rb trees ARE used
20:37.50 brlcad if_tk.c:292: error: too many arguments to function 'Tk_PhotoPutBlock'
20:40.49 CIA-38 BRL-CAD: 03brlcad * r37175 10/brlcad/trunk/src/libfb/if_tk.c: Tk_PhotoPutBlock doesn't take an interp, at least with tk 8.4
21:10.36 ``Erik O.O
21:10.39 ``Erik http://pastebin.bzflag.bz/d3691b911
21:10.49 ``Erik er, hell... src/librt/primitives/arb8/arb8.c:2092: warning: passing argument 2 of 'rt_arb_3face_intersect' from incompatible pointer type
21:11.18 ``Erik const point_t x[6]; throws that, const point_t *x; in another file, ...
21:11.25 ``Erik neat, huh?
21:45.47 *** join/#brlcad akafubu (n=akafubu@unaffiliated/akafubu)
22:09.30 *** join/#brlcad Ralith (n=ralith@69.90.48.97)
22:23.36 *** join/#brlcad akafubu (n=akafubu@unaffiliated/akafubu)
22:34.56 CIA-38 BRL-CAD: 03brlcad * r37176 10/brlcad/trunk/src/librt/prep.c:
22:34.56 CIA-38 BRL-CAD: if we're asked to prep an rtip with nothing in it, we're done prepping that
22:34.56 CIA-38 BRL-CAD: rtip. set needprep to false so code elsewhere doesn't keep recalling prep
22:34.56 CIA-38 BRL-CAD: spewing 'no (primitives|regions) left to prep' messages. this is related to sf
22:34.56 CIA-38 BRL-CAD: bug report 2927515 (a particular arb5 fails to raytrace) from John Dalton
22:34.59 CIA-38 BRL-CAD: (john-dalton), quelling the spewing of 'endless' error messages.
23:14.41 *** join/#brlcad Nohla (n=jesica@168.226.179.239)
23:32.48 CIA-38 BRL-CAD: 03brlcad * r37177 10/brlcad/trunk/src/rt/ (do.c hurt.c):
23:32.48 CIA-38 BRL-CAD: do frame was only checking if there were primitives prepped, but not whether
23:32.48 CIA-38 BRL-CAD: there were any valid regions remaining to render (which there should always be
23:32.48 CIA-38 BRL-CAD: as even all regionless objects (even primitives) are promoted to regions). make
23:32.49 CIA-38 BRL-CAD: it check and halt if there aren't any so that we don't dispatch all rays only to
23:32.51 CIA-38 BRL-CAD: have prep and do_run do nothing useful.
23:34.00 *** join/#brlcad akafubu (n=akafubu@unaffiliated/akafubu)
23:37.29 CIA-38 BRL-CAD: 03brlcad * r37178 10/brlcad/trunk/NEWS:
23:37.29 CIA-38 BRL-CAD: rt now reports prep failures a little more clearly, in part by aborting earlier,
23:37.29 CIA-38 BRL-CAD: so as to not spew a prep failure per ray when rt is invoked from within mged.
23:37.29 CIA-38 BRL-CAD: this is in response to sf bug report 2927515 (a particular arb5 fails to
23:37.29 CIA-38 BRL-CAD: raytrace) from John Dalton (john-dalton), quelling the spewing of 'endless'
23:37.32 CIA-38 BRL-CAD: error messages.
23:47.54 CIA-38 BRL-CAD: 03brlcad * r37179 10/brlcad/trunk/src/librt/primitives/arb8/arb8.c: quell constness conversion warning. make rt_arb_3face_intersect take a const pointer to plane_t, assume it's a [6] array.
23:48.04 brlcad that should fix the warning ``Erik
23:57.32 CIA-38 BRL-CAD: 03brlcad * r37180 10/brlcad/trunk/src/libfb/if_tk.c: so 8.5 did change Tk_PhotoPutBlock to have an interp, so we need to check the version of tk we're working with given it's an api incompatibiliy.

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.