IRC log for #brlcad on 20110405

00:30.13 starseeker Woot!
00:30.20 starseeker now has 3 Neo1973 phones
01:11.30 *** join/#brlcad crazy_imp (~mj@a89-182-208-247.net-htp.de)
02:03.50 *** join/#brlcad sachinjain (~sachin@117.211.88.150)
02:27.04 *** join/#brlcad Ralith (~ralith@S010600221561996a.vc.shawcable.net)
02:46.18 *** join/#brlcad IriX64 (~kvirc@bas2-sudbury98-1096601295.dsl.bell.ca)
03:22.50 *** join/#brlcad louipc (~louipc@archlinux/trusteduser/louipc)
03:58.55 brlcad starseeker: haha
04:49.44 *** join/#brlcad adityag (~ADITYA@182.237.144.88)
04:50.07 *** part/#brlcad adityag (~ADITYA@182.237.144.88)
05:57.41 *** join/#brlcad sachinjain (~sachin@117.211.88.150)
06:04.45 *** join/#brlcad KimK (~Kim__@ip174-71-95-176.om.om.cox.net)
06:39.26 *** join/#brlcad Stattrav (~Stattrav@111.93.134.142)
06:39.26 *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav)
08:10.42 *** join/#brlcad d_rossberg (~rossberg@BZ.BZFLAG.BZ)
09:00.22 *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav)
10:03.59 *** join/#brlcad juanman (~quassel@unaffiliated/juanman)
10:29.15 dloman Mernin all
10:36.32 kunigami good morning
10:38.54 ``Erik yargh
10:49.23 dloman so hows the lisp thing going =D
10:58.01 ``Erik awesome, files are moving back and forth, etc
10:58.54 ``Erik the next trick is tesselating geometry to send out on the fly, which is the brlcad.lisp shtuff (which'll also do images, wireframes, parts of files, etc)
10:59.25 dloman are there svn libs for lisp?
10:59.57 ``Erik probably, but cffi is quick and easy to write, way nicer than jni
11:00.55 dloman kewl
11:02.15 ``Erik librt is very macro and struct heavy, so it's tricky to interface it :/ almost temped to write getter/setter type funcs for some of the librt stuff :D
11:02.36 dloman ...write it in what language?
11:02.56 ``Erik C, in raytrace.h
11:03.12 dloman i'd say go for it :)
11:03.28 ``Erik rt_set_ray(point_t origin, vect_t dir); etc
11:03.40 ``Erik er, with an application sturct somewhere in there
11:12.56 CIA-105 BRL-CAD: 03davidloman * r44199 10/geomcore/trunk/ (2 files in 2 dirs): Add in an optional replyMsg arg to the objToChunk converter
11:22.39 CIA-105 BRL-CAD: 03davidloman * r44200 10/geomcore/trunk/src/GS/FileDataSource.cxx: Oops, forgot to commit FileDataSource::getObjs() ! It's implemented.
11:23.31 CIA-105 BRL-CAD: 03davidloman * r44201 10/geomcore/trunk/src/GS/DataManager.cxx: Cleanup/fixup DataManager::handleGeometryReqMsg. Now uses the shiney, new, and improved FileDataSource.
11:29.27 CIA-105 BRL-CAD: 03davidloman * r44202 10/geomcore/trunk/src/GS/GeometryService.cxx: Make GeometryService objects route GEOMETRYMANIFEST NetMsg's to the Datamanager
11:41.25 CIA-105 BRL-CAD: 03davidloman * r44203 10/geomcore/trunk/ (3 files in 3 dirs): Stub in GetCmd (client cmd) for getting geometry from server.
12:16.25 dloman ``Erik: whats with the GeometryBotReqMsg ?
12:17.04 dloman look identicle to GeometryReqMsg
12:17.18 ``Erik so far, eventually it should tesselate the geometry into an inmem db and send that
12:17.41 ``Erik to feed ogl, isst, vsl, etc
12:18.23 dloman why is a *Msg doing any work other than serialization and deserialization? tesselation is the job of the GE... isn't it?
12:19.19 ``Erik I d'no how ya have things wired up, I was just mimicking what I saw :D I imagine the msg would have to ask ge for a facetized version, no?
12:20.12 dloman perhaps, but that's not the way the gs is architected.
12:21.11 ``Erik okie, it's hard to read through all that inheritance, d'no where exactly it should fit... I just figured I'd put SOMETHING there so you could fix it up later :D
12:21.14 dloman some function being executed by some thread would 1) get the geometry from the DataManager, 2) use the GE to tesselate it and 3) make a GeometryManifest/Chunk to send the results to the requestor.
12:21.22 dloman kk
12:21.40 dloman there isn't that much inheritance :) Nothing compared to stuff like QT and Boost. heh.
12:21.47 dloman those make me cry
12:21.58 ``Erik at least they're documented :>
12:22.16 dloman :P They're also stable and released
12:23.41 dloman ..weren't created by one person, weren't created by someone doing their first soft dev project, etc, etc.
12:23.46 dloman the list is long.
12:24.05 ``Erik (and the only machine I have with doxygen can't configure geomcore :/ )
12:24.21 dloman cmake fails?
12:24.36 ``Erik FindBRLCAD.cmake can't see the (system) tcl
12:25.14 dloman bummer
12:25.15 starseeker ``Erik: if that's using our custom FindTCL there should be a way to expand the search paths
12:26.08 ``Erik is that pushed into rt^3, too?
12:26.24 starseeker uh, dunno
12:26.44 starseeker I was waiting until the ge/gs/gui reshuffle to go through all that
12:29.56 *** join/#brlcad juanman (~quassel@unaffiliated/juanman)
13:10.36 *** join/#brlcad adityag (~ADITYA@182.237.144.88)
13:34.36 CIA-105 BRL-CAD: 03erikgreenwald * r44204 10/geomcore/trunk/src/interfaces/cl/brlcad.lisp: bind libgcv. do RT_APPLICATION_INIT stuff. use &key to fill some of the struct.
13:35.05 CIA-105 BRL-CAD: 03erikgreenwald * r44205 10/geomcore/trunk/src/interfaces/cl/ (gsnet.lisp gsserver.lisp): clean up conditional stuff
13:57.34 *** join/#brlcad adityag (~ADITYA@182.237.144.88)
14:18.29 CIA-105 BRL-CAD: 03davidloman * r44206 10/geomcore/trunk/src/libNet/netMsg/GenericMultiByteMsg.cxx: Removed some stray debugging lines.
14:19.44 brlcad so my joking rant was apparently taken much more to heart than was intended
14:20.16 brlcad if anyone in here had a part in conveying how 'upset' I was, I really wasn't
14:21.34 brlcad the initial crap coding comment was totally meant to be tongue-in-cheek playful, the rest was simple matter-of-fact
14:21.42 CIA-105 BRL-CAD: 03davidloman * r44207 10/geomcore/trunk/ (include/GetCmd.h src/GS/cmds/GetCmd.cxx): Implement GetCmd for the test client.
14:22.13 brlcad particularly with release preparations under way
14:23.11 dloman Well, if you want honesty, i was working remotely yesterday and only had IRC to go by, but you came across as genuinely upset. Just how I read it.
14:23.32 dloman based on your above comments I assume others read it incorrectly also eh?
14:23.38 brlcad apparently
14:23.56 brlcad dislikes conflict, generally very happy-go-lucky
14:24.47 brlcad I was annoyed by the timing, and that probably made some of my remarks more cynical than usual
14:24.55 CIA-105 BRL-CAD: 03davidloman * r44208 10/geomcore/trunk/src/libNet/netMsg/GeometryChunkMsg.cxx: Typo fix.
14:25.48 dloman eh, it happens. As adults we all (should) have thick enough skin to weather someone's irritation, be it misinterpreted or not.
14:25.51 brlcad either way, the internet sucks at conveying emotion and intent -- I should (and do) know better and should have known better yesterday
14:26.35 brlcad also don't know if this played any part
14:26.44 brlcad but for the record
14:26.46 brlcad reverts should have NO NEGATIVE CONNOTATION
14:27.17 brlcad particularly for open source software projects, something I was taught long ago that is different from closed development
14:27.37 brlcad When there is dispute or need, it's supposed to encourage communication without any negative connotation.
14:27.46 brlcad You have just as much right to revert something I've done if a need arises.
14:27.53 brlcad If you un-revert something I've reverted, then we both stop and discuss since that implies there are conflicting needs.
14:28.05 brlcad just an FYI for everyone :)
14:28.59 dloman did i miss a revert war?
14:29.05 brlcad nope
14:29.25 brlcad but I did revert something yesterday and don't know if that played any part in people "interpreting" my upsetness
14:29.49 dloman kk. I've been trying to pay atention to the commit emails and thought I'd missed yet another important bit of info
14:30.29 dloman "SO i heard brlcad was so mad that he punched three old ladies in the face."
14:30.40 brlcad it was four dammit!
14:30.43 dloman nice
14:32.08 CIA-105 BRL-CAD: 03davidloman * r44209 10/geomcore/trunk/src/GS/GeometryService.cxx: Make sure the DataManager point is initialized to something other than 0x0 prior to attempting to use it.
14:34.19 CIA-105 BRL-CAD: 03davidloman * r44210 10/geomcore/trunk/src/GS/DataManager.cxx: Add some error logging so that DataManager won't handle a GeometryReqMsg silently anymore.
14:43.18 CIA-105 BRL-CAD: 03davidloman * r44211 10/geomcore/trunk/src/ (3 files in 2 dirs): Make test client handle geometry data sent by server.
14:45.48 dloman woot, back to where i was before the qt rip out :)
14:46.24 ``Erik starts looking for something else to break O:-)
14:46.35 dloman lol
14:46.45 dloman no need, i'm already too good at breaking things.
14:47.01 ``Erik so ya found the segfault?
14:47.08 dloman oh, and take that damned halo off your head. you're not fooling anyone :P
14:47.10 dloman yeah
14:47.31 dloman keith was right on with the pointer pointing at garbage
14:59.52 CIA-105 BRL-CAD: 03Markhobley 07http://brlcad.org * r2818 10/wiki/MGED_Commands: /* E */
15:11.55 CIA-105 BRL-CAD: 03davidloman * r44212 10/geomcore/trunk/src/libNet/NetMsgFactory.cxx: Make NetMsgFactory log an error when an unknown MsgType is encountered. (Rather than just returning NULL)
15:16.00 CIA-105 BRL-CAD: 03davidloman * r44213 10/geomcore/trunk/src/GS/DataManager.cxx: Fix DataManager's error reporting to the client. Was sending back error codes as MsgTypes. Now sends MsgTypes as MsgTypes!
15:20.04 *** join/#brlcad adityag (~ADITYA@182.237.144.88)
15:20.36 ``Erik now there's a cab: http://www.youtube.com/watch?v=OiKFCzMxJik
15:21.37 dloman nice
15:23.50 dloman question: int bu_file_readable(const char *path)
15:24.05 dloman the int that is returned.... is it 0 == readable ?
15:39.30 ``Erik nonzero is readable, src/libbu/stat.c ... if(bu_file_readable(file)) { do stuff to it; } else { bu_log("u y no has read?"); }
15:41.13 dloman kk, that's what I was thinkging, just got confused.
15:41.14 dloman danke!
15:53.49 CIA-105 BRL-CAD: 03davidloman * r44214 10/geomcore/trunk/include/FileDataSource.h: Add in a util function that determines if the supplied path exists and if it does, whether or not its a file or dir.
15:54.18 CIA-105 BRL-CAD: 03davidloman * r44215 10/geomcore/trunk/src/GS/FileDataSource.cxx: Woops, forgot a file on the last commit.
16:03.00 CIA-105 BRL-CAD: 03davidloman * r44216 10/geomcore/trunk/src/GS/FileDataSource.cxx: Make FileDataSource filter out all paths except paths ending in files.
16:12.01 *** join/#brlcad juanman (~quassel@unaffiliated/juanman)
16:12.37 CIA-105 BRL-CAD: 03davidloman * r44217 10/geomcore/trunk/src/GS/cmds/GetCmd.cxx: Update the GetCmd::getHelp() method to display correct help
16:22.33 *** join/#brlcad adityag1 (~ADITYA@182.237.144.88)
16:30.54 brlcad dloman: there is no guarantee that <0 means files doesn't exist
16:31.27 brlcad header documents 0 (i.e., false) if doesn't exist, and >0 for exists .. <0 is undefined
16:31.43 dloman didn't see that in the header :/
16:31.52 *** join/#brlcad hyarion_ (c05ben@peppar.cs.umu.se)
16:32.05 dloman just saw one sentence, so i read the code and asked to confirm. =D
16:32.21 brlcad they're meant to be simple if (bu_file_exists(file)) then do something else don't
16:32.23 *** join/#brlcad bhinesley_ (~bhinesley@adsl-99-104-170-214.dsl.bkfd14.sbcglobal.net)
16:32.51 brlcad complex return values usually indicate API weaknesses
16:32.58 brlcad so it's just a simple boolean
16:33.02 dloman kk
16:34.07 *** join/#brlcad roberthl_ (~robert@v1.rhl.me.uk)
16:35.49 *** join/#brlcad roberthl (~robert@v1.rhl.me.uk)
16:35.49 *** join/#brlcad roberthl (~robert@mediawiki/RobertL)
16:37.15 *** join/#brlcad alex_joni (~alex_joni@emc/board-of-directors/alexjoni)
16:37.58 *** join/#brlcad indianlarry (~indianlar@BZ.BZFLAG.BZ)
16:39.19 brlcad dloman: fwiw, things like FileDataSource::existsFileOrDir() really belong up in libbu (at least the implementation of it does)
16:39.32 brlcad so we're portably consistent and we get reuse
16:39.40 dloman right on.
16:40.03 dloman i'm pushing getting functionally complete and debugged to stable. Then I plan on doing a ****ton of clean up =D
16:40.15 brlcad if the API is lacking, then it's just as much work to make things fit cleanly into libbu as they do somewhere else
16:41.47 *** join/#brlcad roberthl (~robert@v1.rhl.me.uk)
16:41.47 *** join/#brlcad roberthl (~robert@mediawiki/RobertL)
16:43.14 brlcad yeah, with the current code, I think we just hadn't had a need yet to distinguish dirs beyond bu_file_exists() and bu_count_path()
16:43.49 brlcad if that need is there, then it's trivial to extend another call like bu_dir_exists() or bu_is_directory()
16:43.57 dloman well if the esistsFileOrDir(char*) is usefull then i can push it down there.
16:44.14 brlcad there's a whole class of posix file node matches like that
16:44.38 brlcad I wouldn't push it as it -- right now you only care if it's a file or a dir, but there are other types
16:44.41 brlcad could be a link
16:44.43 brlcad a pipe
16:44.44 brlcad a socket
16:45.04 brlcad existsFileOrDirOrSymbolicLinkOrPipeOr ...
16:45.19 brlcad each having useful purpose
16:46.47 brlcad that's why the current just followed the posix class ("man bash", jump down to CONDITIONAL EXPRESSIONS to see a list)
16:47.11 dloman kk
16:48.52 brlcad you could get a nearly identical result to your function with something like bu_file_type() that returns an enum type for regular file, dir, link, named pipe, etc
16:49.39 dloman bu_file_type(), got it.
16:49.51 brlcad that way you're still only implementing support for the types you need now but have the API in place that will extend to future node types
16:49.54 *** join/#brlcad ``Erik_ (Here@c-69-140-109-104.hsd1.md.comcast.net)
16:55.56 CIA-105 BRL-CAD: 03erikgreenwald * r44218 10/brlcad/trunk/src/conv/g-egg.c: eliminate as many globals as possible. pack the rest in the global struct.
16:57.11 brlcad pretty awesome: http://www.youtube.com/watch?v=xFrMl_5T0Rc&feature=player_embedded
16:59.04 brlcad open source cars
17:01.39 starseeker is that anything like these guys? http://www.40fires.org/
17:02.09 starseeker or the commerical side: http://www.riversimple.com/
17:05.16 starseeker huh - http://www.onawi.org/
17:05.27 starseeker that could be kinda cool
17:05.40 dloman am i reading g_transfer.c:server_geom() correctly? wdb_export_external() doesn't write to disk there?
17:05.58 dloman brlcad: Nice video, that guys a good speaker :)
17:13.23 brlcad dloman: wdb_export_external() writes to the "database instance", whatever that is
17:13.34 brlcad in g_transfer's case, it's an in-memory-only database
17:13.49 dloman yeah, figgered that out just now, heh.
17:13.56 brlcad so it doesn't write to disk, but only because it's in-mem
17:14.29 dloman if i instantiate a different type of db, then wdb_export_external to that, then that will get me writing to disk...... correct?
17:14.56 brlcad depends what kind of dbip you make, but if it's a file-based dbip, then yes -- it'll write out to disk
17:19.29 dloman db_open() with a RT_WDB_TYPE_DB_DISK flag?
17:27.47 brlcad sounds about right
17:27.56 dloman right on
17:28.14 CIA-105 BRL-CAD: 03starseeker * r44219 10/geomcore/trunk/src/ (6 files in 2 dirs): Get some renaming done, get a couple files building.
17:37.57 CIA-105 BRL-CAD: 03erikgreenwald * r44220 10/brlcad/trunk/ (configure.ac src/Makefile.am src/java/): no implementation, abandoned 8 years ago, unused, gone.
17:39.00 dloman http://www.local-motors.com/checkup.php?c=6998
17:39.01 dloman awesome.
17:41.53 ``Erik http://ecoroamer.com/drupal/CAD-files dwg for the thing
17:42.13 dloman ....do i smell a new addition to the db library?
17:42.44 ``Erik dwg is all engineering line drawings, no solid geometry, right?
17:42.55 starseeker right
17:43.10 starseeker good material for modeling, but not a solid model itself
17:44.24 starseeker license statement is... less than clear
17:44.39 starseeker someone should suggest one of the creative commons licenses to him
17:44.53 CIA-105 BRL-CAD: 03erikgreenwald * r44221 10/brlcad/trunk/src/conv/g-egg.c: quell warning that showed up on fbsd but not mac
17:46.34 CIA-105 BRL-CAD: 03starseeker * r44222 10/geomcore/trunk/src/libgvm/ (gvm_svn.h mem.c): Ah, right - we'll be using memory pools for the internals.
17:48.00 ``Erik looks like it's just a stretched f650 with a camper O.o
17:49.00 starseeker probably
17:49.12 dloman yeah *JUST* a streched f650
17:50.10 ``Erik they started with a 2007 f650 with a c7 diesel engine, stretched it, slapped a custom built camper on, run the camper on solar cells and heat captured off the exhaust O.o
17:52.34 CIA-105 BRL-CAD: 03starseeker * r44223 10/geomcore/trunk/src/libgvm/ (breakout.c gvm.h): move header entrys around, remove breakout.c (deal with that later)
17:52.59 ``Erik any news on red testing for release?
17:53.29 CIA-105 BRL-CAD: 03brlcad * r44224 10/brlcad/trunk/src/libged/wdb_obj.c: if ged_title() is wrong, then this one is probably wrong too since it's the daddy. make title not clobber units.
18:00.13 starseeker we've also got a report rtwizard isn't working... let me check...
18:00.46 starseeker crud
18:03.14 *** join/#brlcad Stattrav (~Stattrav@117.192.142.251)
18:03.14 *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav)
18:05.06 CIA-105 BRL-CAD: 03brlcad * r44225 10/brlcad/trunk/NEWS:
18:05.06 CIA-105 BRL-CAD: bob found/fixed a bug with the title command. presumably, this would convert
18:05.06 CIA-105 BRL-CAD: the working units to mm if you tried to set the title via the units command.
18:09.50 CIA-105 BRL-CAD: 03brlcad * r44226 10/brlcad/trunk/TODO: libged encapsulation issue reverted for release, schedule backlog task to restore with libdm/libfb decoupling addressed. similar decouple needed for screengrab command (though that command may just end up migrating?).
18:11.42 *** join/#brlcad adityag (~ADITYA@182.237.144.88)
18:19.40 CIA-105 BRL-CAD: 03davidloman * r44227 10/geomcore/trunk/include/FileDataSource.h: Add a //TODO about swapping out existsFileOrDir with bu_file_type()
18:31.40 CIA-105 BRL-CAD: 03davidloman * r44228 10/geomcore/trunk/src/GS/DataManager.cxx: Add server side logging for geometry requests
18:38.56 CIA-105 BRL-CAD: 03davidloman * r44229 10/geomcore/trunk/src/GS/FileDataSource.cxx: Make note about the Logger failing after the call to BRLCAD::MinimalDatabase cstr.
18:42.00 CIA-105 BRL-CAD: 03davidloman * r44230 10/geomcore/trunk/src/GS/GSClient.cxx: Display the name of the GeometryChunk as it comes in from the server
18:44.26 CIA-105 BRL-CAD: 03brlcad * r44231 10/brlcad/trunk/TODO: file node entity type support to libbu
19:21.22 CIA-105 BRL-CAD: 03brlcad * r44232 10/brlcad/trunk/TODO: red matrix edit works!
19:26.59 CIA-105 BRL-CAD: 03brlcad * r44233 10/brlcad/trunk/TODO: red is looking better, doesn't seem to list rgb/color attributes multiple times any more, but now seems to be wiping out the shader and region_id at least when set to plastic and -1 respectively.
19:42.57 *** part/#brlcad adityag (~ADITYA@182.237.144.88)
20:38.54 *** join/#brlcad Wildstar (42bd9f25@gateway/web/freenode/ip.66.189.159.37)
20:39.05 Wildstar Hello
20:39.35 Wildstar Is there a PDP-11 source code for BRL-CAD?
20:41.13 Wildstar BBL
20:44.13 CIA-105 BRL-CAD: 03starseeker * r44234 10/geomcore/trunk/src/libgvm/ (CMakeLists.txt mem.c objects.c): Add preliminary stab at repository_object conversion routine.
20:45.44 dli PDP-11 is 16bit, amazing request for brl-cad :(
20:50.48 CIA-105 BRL-CAD: 03erikgreenwald * r44235 10/geomcore/trunk/src/interfaces/cl/ (gsclient.lisp gsnet.lisp gsserver.lisp): macros!

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