| 00:55.23 | bjorkintosh | is it possible to run the mged session on one machine and the display on another, remotely? |
| 00:56.34 | bjorkintosh | eg: having mged on asimov and the display portion on acclarke? |
| 01:06.16 | bjorkintosh | fbserv |
| 01:06.18 | bjorkintosh | never mind! |
| 01:18.23 | *** join/#brlcad archivist (n=archivis@host81-149-119-172.in-addr.btopenworld.com) [NETSPLIT VICTIM] |
| 02:07.17 | *** join/#brlcad Ralith (n=ralith@216.162.199.202) |
| 02:28.35 | yukonbob | hello, cadheads |
| 02:43.06 | starseeker | braces himself and starts committing step stuff |
| 04:03.35 | starseeker | is forced to do a retake tomorrow - too much junk that must be cleaned out of the original NIST tarball before it is committed |
| 04:04.37 | bjorkintosh | hmm. |
| 04:04.46 | bjorkintosh | i have fbserv running on an other machine... |
| 04:04.59 | bjorkintosh | is it at all possible to seperate the command window from the graphics window? |
| 04:05.15 | bjorkintosh | so that the graphics window is hosted on the remote machine? |
| 04:06.11 | bjorkintosh | great. a crash. |
| 04:10.07 | brlcad | bjorkintosh: you should be able to display it remotely via simple remote X11 |
| 04:10.18 | brlcad | just set your DISPLAY |
| 04:10.36 | bjorkintosh | i have. but i'd like to have the command window on this machine and the graphics window on the other machine. |
| 04:10.40 | bjorkintosh | is it currently possible? |
| 04:11.04 | brlcad | technically, probably |
| 04:11.14 | brlcad | but not exactly a tested configuration |
| 04:11.22 | bjorkintosh | ah okay. |
| 04:11.25 | bjorkintosh | any hints then? |
| 04:12.01 | brlcad | try "attach remotehost:0" |
| 04:12.17 | brlcad | half-blind stab |
| 04:12.21 | bjorkintosh | in mged? |
| 04:12.23 | bjorkintosh | okay. |
| 04:12.25 | brlcad | yep |
| 04:12.30 | brlcad | in the command window |
| 04:12.51 | brlcad | might not recognize the syntax |
| 04:13.42 | brlcad | ah, nfg |
| 04:14.43 | bjorkintosh | nfg? |
| 04:15.12 | brlcad | ~nfg |
| 04:15.38 | brlcad | ibot: wake up |
| 04:15.39 | ibot | ACTION throws a barrel-full of ice water on up and shouts "GOOD MORNING!!!!" |
| 04:15.49 | brlcad | nfg -> no fucking good |
| 04:16.13 | brlcad | bjorkintosh: well, you could set DISPLAY, then run mged -c, attach X |
| 04:16.51 | brlcad | could even run gui and the classic console will still work |
| 04:17.13 | brlcad | aha |
| 04:17.18 | brlcad | "gui remotehost:0" |
| 04:17.45 | brlcad | it'll create both, but should do the trick for arbitrary hosts |
| 04:17.48 | bjorkintosh | mged -c works. |
| 04:18.18 | brlcad | gui is what you wanted |
| 04:18.23 | brlcad | gui -d remotehost:0 |
| 04:20.25 | bjorkintosh | ah i see. |
| 04:20.36 | bjorkintosh | it creats an identical screen on the remote machine. |
| 04:20.45 | bjorkintosh | so i now have two command windows and two graphics windows. |
| 04:20.52 | bjorkintosh | one on this, and the other on the remote machine. |
| 04:20.59 | bjorkintosh | so what's the fbserv bit for then? |
| 04:22.22 | bjorkintosh | hah! this kicks ass. |
| 04:23.50 | brlcad | yeah, and they all work together |
| 04:24.33 | brlcad | so if you really only wanted one command window on hostA and a graphics window on hostB, you'd issue gui for the other and then close the corresponding windows you don't want |
| 04:24.58 | bjorkintosh | aha! |
| 04:25.29 | brlcad | (note, mged will shutdown if you close all the graphics windows) |
| 04:25.55 | bjorkintosh | i now have the image displayed on 2 different screens. :D |
| 04:26.00 | brlcad | if you want no graphics windows, then you go with mged -c and manually run gui |
| 04:26.23 | bjorkintosh | the machine running brlcad has no X, but i have two other ones which do. |
| 04:26.32 | bjorkintosh | and they're both currently showing the image as desired. |
| 04:27.03 | brlcad | fbserv is the framebuffer server -- framebuffers are used for displaying raster images, not for 3D interactive displays (we call those "display managers" or dm's) |
| 04:28.49 | brlcad | the graphics window is technically a display manager plus an overlayed framebuffer server -- you can see how those two interact with each other via the raytrace control panel |
| 04:32.13 | brlcad | bjorkintosh: you said mged crashed earlier -- do you happen to have a bomb log file(s) in your directory? |
| 04:32.33 | brlcad | if you do, please upload to ftp.brlcad.org/incoming via anon ftp |
| 04:32.48 | brlcad | or post it somewhere easily accessible |
| 04:33.31 | bjorkintosh | okay. if it happens again, i'll send it your way. |
| 04:33.40 | bjorkintosh | i deleted it immediately in annoyance. |
| 04:35.31 | brlcad | ah, k |
| 04:41.34 | bjorkintosh | Zzzz. or i won't wake up in the morning. |
| 04:41.36 | bjorkintosh | thanks again. |
| 05:37.25 | *** join/#brlcad hippieindamakin8 (n=hippiein@59.92.186.200) |
| 06:54.29 | *** join/#brlcad Ralith (n=ralith@216.162.199.202) |
| 07:17.53 | *** join/#brlcad Don__ (n=Don@c-68-62-76-34.hsd1.mi.comcast.net) |
| 07:36.06 | *** join/#brlcad Ralith (n=ralith@216.162.199.202) |
| 08:18.18 | *** join/#brlcad _sushi_ (n=_sushi_@84-72-93-63.dclient.hispeed.ch) |
| 09:48.42 | *** join/#brlcad mafm (n=mafm@223.Red-83-49-86.dynamicIP.rima-tde.net) |
| 09:50.20 | *** join/#brlcad Don_ (n=Don@c-68-62-76-34.hsd1.mi.comcast.net) |
| 10:48.41 | CIA-40 | BRL-CAD: 03davidloman * r33979 10/rt^3/tags/gsph0/: Tagged for GeometeryService phase Zero milestone. |
| 11:21.19 | d-lo | huh. Raytracing in DX11: http://tech.slashdot.org/article.pl?sid=08/03/31/1423247 |
| 11:46.12 | d-lo | Code Religion question: 'Goto' statement in modern C and C++: Good, Neutral, Evil, and why? |
| 11:49.02 | d-lo | My personal take is: Evil as it leads to spaghetti and can easily be avoided by some re-structuring. What are other people's opinions? |
| 11:59.40 | CIA-40 | BRL-CAD: 03davidloman * r33980 10/rt^3/trunk/ (4 files in 3 dirs): Removal of Boost libraries. Didn't work as well as I would have liked. Die boost, die. |
| 12:17.52 | *** join/#brlcad ``Erik (n=erik@ftp.brlcad.org) |
| 12:36.57 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-10.sbndin.btas.verizon.net) |
| 12:37.17 | brlcad | d-lo: it is neither evil nor non-evil, just one of many tools available (and one often misunderstood/misused) -- whether it's the right tool for the job depends entirely on the situation |
| 12:40.30 | brlcad | it rarely is actually needed and beneficial, but there are some times when a localized goto is better than various alternate restructurings |
| 12:42.20 | brlcad | that said, they are very rarely the best tool |
| 12:43.51 | *** join/#brlcad elite01 (n=omg@unaffiliated/elite01) |
| 13:04.01 | d-lo | brlcad: okay cool. I ran across some in rt^3, so i figured I would ask. |
| 13:13.56 | brlcad | d-lo: where at? |
| 13:14.32 | brlcad | ah, I see |
| 13:14.44 | brlcad | all throughout ogre and some of the new code |
| 13:15.34 | d-lo | I am not so much opposed to the stuff in ogre, since its being used as a lib. Just the stuff we have control over. :/ |
| 13:16.42 | brlcad | I see why he's using it, related to the 'exceptions' that libbu can throw with BU_SETJUMP/BU_UNSETJUMP |
| 13:17.18 | d-lo | lack of exp talking here: wouldnt a try/catch/finally work just as well? |
| 13:17.24 | brlcad | it's actually a pretty reasonable use in that place |
| 13:17.25 | brlcad | nope |
| 13:17.55 | brlcad | those are c++ constructs that only work with c++ exceptions, hence the quoted 'exceptions' :) |
| 13:19.49 | brlcad | there is an equivalent logical structure using if/then that could be used -- but it's only arguably better because you're already jumping |
| 13:20.06 | brlcad | jumps are very low-level C |
| 13:20.28 | brlcad | you could implement C++ exceptions using them (and many compilers do/did), for example |
| 13:20.31 | d-lo | hence 'faster' ? |
| 13:20.43 | brlcad | they are .. exceptionally faster |
| 13:20.47 | brlcad | but that's not the reason |
| 13:21.36 | brlcad | they're just one of the ways in C you can achieve a multiple-scope break in logic flow |
| 13:22.00 | d-lo | so its more coding style than anything else? |
| 13:22.22 | brlcad | try/catch/finally are really glorified syntactic setjump/longjump/goto's for that matter |
| 13:22.36 | brlcad | not really coding style -- he's responding to libbu |
| 13:22.42 | brlcad | actually, librt |
| 13:22.59 | brlcad | which will throw a jump on failure for some routines |
| 13:23.20 | brlcad | so he catchs it and returns gracefully |
| 13:23.24 | d-lo | hrm, I will have to look at that deeper sometime soon i suppose.... |
| 13:24.50 | brlcad | quick example: |
| 13:24.52 | brlcad | <PROTECTED> |
| 13:24.52 | brlcad | <PROTECTED> |
| 13:24.52 | brlcad | <PROTECTED> |
| 13:24.52 | brlcad | <PROTECTED> |
| 13:24.54 | brlcad | END_MARK: |
| 13:24.57 | brlcad | <PROTECTED> |
| 13:25.24 | brlcad | first pass through, the jump point is set on that if line |
| 13:26.29 | brlcad | and it evaluates false, does the unitize and rt_shootray |
| 13:26.34 | brlcad | if rt_shootray() has a failure, it may jump -- which will return logic back to that if line and it'll evaluate true |
| 13:26.54 | brlcad | at which point it goes to endmark, unsets the jump, and continues on its way |
| 13:27.21 | brlcad | in this particular example, there was nothing important that needed to occur, but could have easily had more after rt_shootray() that would have been important to avoid |
| 13:27.26 | d-lo | cool. Makes sense. :) |
| 13:27.40 | brlcad | you can acheive the same in that particular example with just: |
| 13:27.53 | brlcad | if (!BU_SETJUMP) { |
| 13:27.59 | brlcad | <PROTECTED> |
| 13:27.59 | brlcad | <PROTECTED> |
| 13:28.02 | brlcad | } |
| 13:28.06 | brlcad | BU_UNSETJUMP; |
| 13:28.16 | brlcad | but that's specific to that snippet |
| 13:28.42 | brlcad | and arguably less explicit |
| 13:29.03 | brlcad | that's closer to what we do in our own code, though, with the jumps |
| 13:30.08 | d-lo | pretty neat. Has it been benchmarked to an equivilent try/catch? I would be curious as to the results! |
| 13:32.14 | brlcad | you can't use try/catch |
| 13:32.23 | brlcad | they're not c++ exceptions |
| 13:32.37 | d-lo | righto, hence 'equivlent' |
| 13:33.08 | brlcad | that is an equivalent mechanism, what do you mean? |
| 13:33.43 | d-lo | Jumping is a failure recovery mechanism (at least used like this), correct? |
| 13:33.47 | brlcad | if/then vs if/goto? not really much difference performance-wise |
| 13:33.57 | brlcad | sure is |
| 13:34.20 | brlcad | or a roll-back logic redirect |
| 13:34.46 | d-lo | and the above code snippet is trying to call rt_shootray(&ap) all the while ready to catch a failure of that call.... right? |
| 13:34.58 | brlcad | yep |
| 13:35.23 | brlcad | specifically an abnormal failure that is raised by a jump |
| 13:37.15 | d-lo | so, the C++ equivlient would be something like: |
| 13:37.31 | d-lo | try { |
| 13:37.39 | d-lo | VUNITIZE(ap.a_ray.r_dir); |
| 13:37.39 | d-lo | rt_shootray(&ap); |
| 13:37.39 | d-lo | } catch (OmgWtfHappendToRtShootRayException e) { |
| 13:37.39 | d-lo | //stuff |
| 13:37.39 | d-lo | } finally { |
| 13:37.41 | d-lo | //more stuff |
| 13:37.46 | d-lo | } |
| 13:37.48 | brlcad | that would only work if rt_shootray threw c++ exceptions |
| 13:37.53 | d-lo | bah, that paste didn't work. |
| 13:37.55 | brlcad | it doesn't throw c++ exceptions |
| 13:37.59 | brlcad | rt_shootray can't throw a c++ exception because it's not a c++ library |
| 13:38.02 | d-lo | right, i get that. |
| 13:38.47 | brlcad | so it's not an option .. :) |
| 13:38.47 | d-lo | "But if it did", how would the performance compare..... |
| 13:38.52 | brlcad | oh hell |
| 13:38.57 | brlcad | c++ exceptions are absurdly expensive |
| 13:39.03 | brlcad | there have been lots of comparisons |
| 13:39.30 | brlcad | it doesn't really get much cheaper than a jump -- it literally amounts to a one-line jmp assembly instruction usually |
| 13:39.39 | d-lo | kk |
| 13:39.46 | brlcad | c++ exceptions are one to two orders faster |
| 13:40.03 | brlcad | which is why studio has entire compilation modes with c++ exceptions on/off |
| 13:40.04 | d-lo | on a side note, I need to learn asm :/ |
| 13:40.45 | brlcad | runs |
| 13:51.27 | *** join/#brlcad PrezKennedy (i=Matthew@whitecalf.net) |
| 14:41.51 | *** join/#brlcad samrose (n=samrose@c-24-11-214-181.hsd1.mi.comcast.net) |
| 14:56.01 | *** join/#brlcad BigAToo (n=BigAToo@64.255.115.3) |
| 15:36.21 | *** join/#brlcad Dr_Phreakenstein (n=phreak@216.151.24.198) |
| 15:49.11 | *** join/#brlcad ``Erik__ (n=erik@c-76-111-12-116.hsd1.md.comcast.net) |
| 15:50.21 | d-lo | ``Erik__: Heya! |
| 16:44.44 | *** join/#brlcad mafm (n=mafm@223.Red-83-49-86.dynamicIP.rima-tde.net) |
| 16:50.01 | ``Erik__ | oi |
| 16:51.14 | d-lo | howdy stranger! |
| 16:52.07 | d-lo | whats new? |
| 16:54.27 | CIA-40 | BRL-CAD: 03davidloman * r33981 10/rt^3/trunk/ (69 files in 15 dirs): Removed typedefs from iBMECommon.h. Refactored all .h and .cxx files accordingly. |
| 16:59.11 | CIA-40 | BRL-CAD: 03starseeker * r33982 10/brlcad/trunk/ (499 files in 49 dirs): (log message trimmed) |
| 16:59.12 | CIA-40 | BRL-CAD: Begin the process of adding NIST's STEP Class Libraries code to the BRL-CAD |
| 16:59.12 | CIA-40 | BRL-CAD: tree. This first commit sets up the build logic for a src/other/step |
| 16:59.12 | CIA-40 | BRL-CAD: subconfigure and imports a step tree which is the original NIST STEP 3.2 |
| 16:59.12 | CIA-40 | BRL-CAD: release, minus some temporary files and an old file in src/cldai that causes |
| 16:59.14 | CIA-40 | BRL-CAD: conflicts on case insensitive filesystems. This is not expected to be a |
| 16:59.16 | CIA-40 | BRL-CAD: buildable configuration - it is included to preserve the original state of the |
| 17:19.04 | *** join/#brlcad hippieindamakin8 (n=hippiein@59.92.185.81) |
| 17:25.59 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-10.sbndin.btas.verizon.net) |
| 17:39.22 | CIA-40 | BRL-CAD: 03starseeker * r33983 10/brlcad/trunk/src/other/step/ (83 files in 25 dirs): (log message trimmed) |
| 17:39.22 | CIA-40 | BRL-CAD: Update STEP code to version from |
| 17:39.22 | CIA-40 | BRL-CAD: http://www.statik.tu-cottbus.de/fileadmin/project_dth/scl/ This configuration |
| 17:39.22 | CIA-40 | BRL-CAD: is closer to building on many platforms but will need special options to |
| 17:39.22 | CIA-40 | BRL-CAD: configure not supplied by default. As this is not the final form the libraries |
| 17:39.26 | CIA-40 | BRL-CAD: will take in the BRL-CAD tree no attempt is made to work with this build |
| 17:39.28 | CIA-40 | BRL-CAD: structure. One more commit will be needed for a file renaming but after that |
| 17:39.38 | d-lo | go go gadget STEP! |
| 17:41.46 | CIA-40 | BRL-CAD: 03starseeker * r33984 10/brlcad/trunk/src/other/step/src/cldai/sdaiDaObject.cc: Add back in the properly named sdaiDaObject.cc |
| 17:41.52 | starseeker | OK, now the real fun |
| 17:44.07 | starseeker | takes what he has building so far and attemps to make something that won't bust distcheck |
| 17:49.11 | starseeker | debates nuking and re-adding vs. gradually shifting to working... |
| 17:49.52 | CIA-40 | BRL-CAD: 03indianlarry * r33985 10/brlcad/trunk/src/libged/clone.c: updated clone command to re-number and work with the naming convention used in 'build_region' tclscript |
| 17:53.42 | CIA-40 | BRL-CAD: 03indianlarry * r33986 10/brlcad/trunk/src/librt/mirror.c: updated rt_mirror to fix bug 2679693, looks like the radius vectors were getting paired wrong in the mirror |
| 18:01.34 | d-lo | ``Erik: You around? |
| 18:07.29 | *** join/#brlcad ``Erik_ (i=erik@c-76-111-12-116.hsd1.md.comcast.net) |
| 18:08.20 | d-lo | heh, ``Erik : you having connection problems or what? ;) |
| 18:19.35 | ``Erik_ | always |
| 18:19.56 | d-lo | whats new? |
| 18:21.40 | CIA-40 | BRL-CAD: 03starseeker * r33987 10/brlcad/trunk/src/other/step/ (16 files): Start setting up standard README, COPYING, NEWS, etc. files |
| 18:23.02 | d-lo | ``Erik: When could you sit down and talk to me about specifics of that Lisp web framework? I have the itch to move forward with a web- base game ;) |
| 18:24.34 | CIA-40 | BRL-CAD: 03starseeker * r33988 10/brlcad/trunk/src/other/step/ (arch_template/ bin/): Clear out arch_template and bin |
| 18:24.53 | ``Erik_ | *shrug* I should be in tomorrow, it's an odd thing though |
| 18:25.55 | d-lo | *chuckle* the ucw google search stil makes me laugh :) |
| 18:29.13 | d-lo | might use php though. Which would you recommend for storage sql or flat file? |
| 18:45.43 | CIA-40 | BRL-CAD: 03starseeker * r33989 10/brlcad/trunk/src/other/step/ (80 files in 7 dirs): More inching forward in STEP |
| 18:47.21 | CIA-40 | BRL-CAD: 03starseeker * r33990 10/brlcad/trunk/src/other/step/misc/ (. Makefile.am Makefile.defs flawfinder libtool.m4): Add step/misc directory |
| 18:54.40 | CIA-40 | BRL-CAD: 03homovulgaris * r33991 10/brlcad/trunk/src/other/boost/ (996 files in 124 dirs): boost update particularly phoenix |
| 18:59.31 | *** join/#brlcad madant (n=madant@117.196.146.10) |
| 19:09.55 | CIA-40 | BRL-CAD: 03starseeker * r33992 10/brlcad/trunk/src/other/step/ (8 files in 8 dirs): Add some Makefile.am files to various directories |
| 19:31.01 | brlcad | howdy madant |
| 19:33.10 | brlcad | notes that distcheck is busted |
| 19:33.22 | starseeker | brlcad: I know, I know |
| 19:33.25 | starseeker | working on it ;-) |
| 19:33.57 | CIA-40 | BRL-CAD: 03starseeker * r33993 10/brlcad/trunk/src/other/step/src/ (123 files in 8 dirs): Merge in most of the changes made thus far. Probably won't build yet as new Makefile.am stuff is untested, but getting close |
| 19:37.04 | brlcad | first commit step should have been to just extra_dist it from src/other or conditionalize it, then work on cleaning up src/other/step/*, then deconditionalize it -- that way it stays working |
| 19:37.27 | starseeker | ok, I'll put that in for now |
| 19:37.55 | brlcad | do you know how to conditionalize it? |
| 19:38.14 | starseeker | sort of |
| 19:40.17 | CIA-40 | BRL-CAD: 03starseeker * r33994 10/brlcad/trunk/src/other/Makefile.am: EXTRA_DIST step for now |
| 19:42.03 | starseeker | that get it for now? |
| 19:44.35 | brlcad | don't know, busy reading what we're going to talk about in 15 min ;) |
| 19:45.19 | brlcad | fyi, though, I have some questions -- like what's up with the stuff deleted -- commit message doesn't say what/why |
| 19:45.41 | brlcad | would have thought clcore was kinda important, for example |
| 19:45.56 | brlcad | unless it's just for the editor |
| 19:47.04 | brlcad | and it looks like build is still busted, you have subconfigure enabled in configure |
| 19:47.10 | brlcad | returns to reading |
| 19:49.32 | CIA-40 | BRL-CAD: 03starseeker * r33995 10/brlcad/trunk/configure.ac: Disable step subconfigure until it's working |
| 19:50.34 | *** join/#brlcad samrose (n=samrose@c-24-11-214-181.hsd1.mi.comcast.net) |
| 20:01.34 | CIA-40 | BRL-CAD: 03starseeker * r33996 10/brlcad/trunk/src/other/step/Makefile.am: Stuff the other dirs behind EXTRA_DIST for now, and bring them out later if they should be elsewhere. |
| 20:03.58 | CIA-40 | BRL-CAD: 03brlcad * r33997 10/brlcad/trunk/doc/TODO.BREP: add a missing important plane specialization of surfaces analogous to lines vs curves |
| 20:37.55 | madant | hi brlcad :) nothing much.. i need to sort out some crappy spirit and phoenix includes.. :) going to sleep now :) |
| 21:30.34 | CIA-40 | BRL-CAD: 03brlcad * r33998 10/brlcad/trunk/doc/TODO.BREP: Added few more queries. |
| 21:52.33 | *** join/#brlcad PrezKennedyJR (i=Matthew@whitecalf.net) |
| 22:03.00 | CIA-40 | BRL-CAD: 03brlcad * r33999 10/brlcad/trunk/doc/TODO.BREP: Breakout of raytracing dependencies |
| 22:06.53 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-10.sbndin.btas.verizon.net) |