| 00:23.12 | *** join/#brlcad ibot_ (i=ibot@pdpc/supporter/active/TimRiker/bot/apt) |
| 00:23.12 | *** topic/#brlcad is http://brlcad.org/ || BRL-CAD is an open source solid modeling software suite || Developers needed! Read the HACKING file for details on getting involved |
| 00:41.32 | *** join/#brlcad DTRemenak (n=Daniel_R@c-24-23-59-104.hsd1.mn.comcast.net) |
| 01:14.29 | *** join/#brlcad DTRemenak|RDP (n=DTRemena@c-24-23-59-104.hsd1.ca.comcast.net) |
| 01:45.14 | *** join/#brlcad IriX64 (n=IriX64@toronto-HSE-ppp4308865.sympatico.ca) |
| 01:46.15 | IriX64 | mmmm would you be interested in the source to that linuxbrlcad.zip i sent to ftp.brlcad.org a while back? |
| 01:47.46 | IriX64 | ill take that as a yes, you're free to delete it if it doesn't pass muster. |
| 01:49.01 | brlcad | the source? |
| 01:49.06 | brlcad | why, did you change the sources? |
| 01:49.15 | IriX64 | eh? |
| 01:49.35 | brlcad | why would I want the source to linuxbrlcad.zip? |
| 01:49.51 | IriX64 | to go along with the bibaries. |
| 01:50.03 | brlcad | but I mean did you actually modify the sources? |
| 01:50.27 | brlcad | presumably you obtained the sourcecode from cvs or from a source distribution |
| 01:50.47 | brlcad | so if you didn't modify the sources.... why would I need/want them? |
| 01:51.08 | IriX64 | do the binaries not have stuff thats not there in the distribution? |
| 01:51.22 | IriX64 | or did you even look at the binaries |
| 01:51.24 | IriX64 | ? |
| 01:51.28 | brlcad | i looked at the binaries |
| 01:51.36 | IriX64 | stock? |
| 01:51.38 | brlcad | you're avoiding my question though :) |
| 01:51.45 | brlcad | did you modify the sources? |
| 01:51.57 | IriX64 | i hate questions life would be simpler without them. |
| 01:52.03 | IriX64 | yes i did. |
| 01:52.32 | brlcad | care to share what and why? anything significant? |
| 01:52.45 | brlcad | if it's useful, i'd integrate it into cvs |
| 01:52.55 | IriX64 | geometry browser has been enhanced |
| 01:53.08 | brlcad | I'm not going to go goose hunting for mysterious changes though :) |
| 01:53.17 | brlcad | enhanced? |
| 01:53.44 | IriX64 | just compile the code. |
| 01:54.35 | brlcad | you enjoy being difficult and obscure, or is it just a hobby? :) |
| 01:54.47 | IriX64 | both ;) |
| 01:56.37 | IriX64 | smoke break. |
| 02:30.46 | IriX64 | ah well client troubles, maybe another time. |
| 02:32.12 | IriX64 | rhubarb anyone? :) |
| 02:32.29 | IriX64 | this client isn't equipped for that though. |
| 02:40.29 | *** join/#brlcad IriX64_ (n=IriX64@toronto-HSE-ppp4308865.sympatico.ca) |
| 02:41.31 | IriX64_ | mention rhubarb one time ... my system goes to war on me. |
| 02:43.21 | IriX64_ | man wheres the status screen on irssi? |
| 02:44.23 | IriX64_ | ill be back. |
| 02:46.34 | *** join/#brlcad IriX64 (n=Who@toronto-HSE-ppp4308865.sympatico.ca) |
| 02:47.03 | IriX64 | clients should be all the same :) |
| 02:50.00 | IriX64 | sugar or salt? |
| 02:51.03 | brlcad | it's like rhubarb pie, but with a sugary crunchy topping in a baking dish |
| 02:53.25 | IriX64 | thought you meant a stick of rhubarb lightly sprinkled. |
| 02:57.35 | IriX64 | IriX64@hagarsfi-f038a0 ~/brlcad-10.0.0 |
| 02:57.35 | IriX64 | $ ./configure --enable-almost-everything --with-x --enable-math --enable-optimi |
| 02:57.35 | IriX64 | zations --disable-shared --build=i586-unix-freebsd |
| 02:58.09 | IriX64 | did i miss any options? |
| 03:02.55 | IriX64 | cleverly programmed into each build cycle is a looooong smoke break :) |
| 03:02.58 | brlcad | --with-x is the default |
| 03:03.36 | IriX64 | thank you wasn't sure thought i looked in configure --help and saw it wasn't or something. |
| 03:03.46 | IriX64 | or read it wrong. |
| 03:28.39 | CIA-9 | BRL-CAD: 03brlcad * 10brlcad/autogen.sh: (log message trimmed) |
| 03:28.39 | CIA-9 | BRL-CAD: yet another fairly massive rewrite, this time mostly just restructuring. keep |
| 03:28.39 | CIA-9 | BRL-CAD: track of the basename for better output messages and refactor the initialization |
| 03:28.39 | CIA-9 | BRL-CAD: steps into a function. also refactor the libtool failure check into a function |
| 03:28.39 | CIA-9 | BRL-CAD: while we're at it as well. all steps towards being able to cleanly handle |
| 03:28.40 | CIA-9 | BRL-CAD: manual build steps of projects using recursive configure. oh, and this change |
| 03:28.42 | CIA-9 | BRL-CAD: removes the directory change to the autogen.sh script, so it must once again be |
| 04:16.08 | CIA-9 | BRL-CAD: 03brlcad * 10brlcad/autogen.sh: allow external overrides on HELP, QUIET, VERBOSE, and VERSION_ONLY. document the settable project defaults |
| 04:29.20 | CIA-9 | BRL-CAD: 03brlcad * 10brlcad/autogen.sh: utilize the RUN_RECURSIVE define so that we don't recurse more than once during autogening |
| 05:34.16 | CIA-9 | BRL-CAD: 03johnranderson * 10brlcad/src/tclscripts/mged/mike-tux.ppm: The file in the cvs repository was corrupted. Sean, be careful when using scripts to update all files :-) |
| 05:34.47 | brlcad | oops |
| 05:37.25 | brlcad | ahh, some lines actually end with a color that equaled a space/tab char |
| 06:39.30 | *** join/#brlcad PKMOBILE (n=Apathy@c-69-250-236-100.hsd1.md.comcast.net) |
| 07:19.25 | CIA-9 | BRL-CAD: 03brlcad * 10brlcad/autogen.sh: add a manual_autogen function that runs the manual configuration steps. this makes it possible to add support for manual recursive configure build support. |
| 07:51.13 | *** join/#brlcad clock_ (i=clock@84-72-94-219.dclient.hispeed.ch) |
| 08:56.18 | ``Erik | hah |
| 08:58.11 | ``Erik | (dare I 'orange' mike-tux.ppm to get ride of those two revisions?) |
| 13:40.38 | brlcad | nah |
| 13:59.47 | *** join/#brlcad dan_falck (n=danfalck@pool-71-111-76-8.ptldor.dsl-w.verizon.net) |
| 15:01.09 | CIA-9 | BRL-CAD: 03brlcad * 10brlcad/NEWS: speling |
| 15:17.39 | *** join/#brlcad bjorkBSD (n=bjork@ip70-178-169-173.ks.ks.cox.net) |
| 15:34.04 | *** join/#brlcad ChanServ (ChanServ@services.) |
| 15:34.04 | *** mode/#brlcad [+o ChanServ] by irc.freenode.net |
| 16:16.19 | CIA-9 | BRL-CAD: 03brlcad * 10brlcad/autogen.sh: allow environment variable overrides on AUTORECONF, AUTOCONF, AUTOMAKE, LIBTOOLIZE, ACLOCAL, AUTOHEADER. fix --help so it works again (need to rewrite so it's not just a configure pass-through) |
| 17:50.30 | *** join/#brlcad matt_ezeki_230 (n=matt@host117-24.pool8710.interbusiness.it) |
| 17:50.39 | matt_ezeki_230 | hello guys... |
| 17:53.26 | brlcad | hello |
| 17:59.08 | matt_ezeki_230 | brlcad: hi, do you remember an our 3-4 days ago chat? We talked about to show a solid inside an other solid as broken line and you suggested me to use direct region flag... |
| 17:59.35 | brlcad | as a potential direction |
| 18:00.05 | brlcad | along with maybe merging the mged wireframe with a rendered version |
| 18:02.09 | matt_ezeki_230 | yes, show potential direction as broken line after a raytrace |
| 18:02.53 | matt_ezeki_230 | I tried to find any example of use of direct region flag |
| 18:03.06 | matt_ezeki_230 | but without success |
| 18:03.39 | matt_ezeki_230 | and neither a definition of direct region flag in MGED_Quick_Ref, VolumeII... |
| 18:03.45 | matt_ezeki_230 | VolumeIII.. etc... |
| 18:04.10 | matt_ezeki_230 | would you be so kind as to give me a simple example of use? |
| 18:05.09 | brlcad | it's in the manpage |
| 18:05.13 | brlcad | for rtedge |
| 18:06.14 | brlcad | somewhat obscure detail to be putting into introductory training docs ;) |
| 18:33.03 | matt_ezeki_230 | ok... I think i'm near at hand the solution! I gave the command rtedge -W -s850 -c"set dr=1" but it returns the following error: |
| 18:33.48 | matt_ezeki_230 | rt_do_cmd("set): command not found |
| 18:36.03 | matt_ezeki_230 | BRL-CAD Release 7.8.2 ..... |
| 18:38.01 | matt_ezeki_230 | /usr/brlcad/bin/rtedge -M -W -s850 -c"set |
| 18:38.15 | matt_ezeki_230 | opendb dr=1"; |
| 18:38.53 | matt_ezeki_230 | rt: rt_dirbuild(dr=1") failure |
| 18:39.02 | matt_ezeki_230 | any suggestion?? |
| 19:05.46 | brlcad | run it on the command line, not inside mged |
| 19:06.01 | brlcad | otherwise you get into tclsh issues and how to go about escaping options, etc |
| 19:15.22 | matt_ezeki_230 | ok...but something was wrong... |
| 19:15.27 | matt_ezeki_230 | i gave |
| 19:15.35 | matt_ezeki_230 | in mged: |
| 19:15.48 | matt_ezeki_230 | 1) in a.s rcc 0 0 -2 0 0 4 1 |
| 19:16.00 | matt_ezeki_230 | 2) in b.s rcc 0 -2 0 0 4 0 1 |
| 19:16.10 | matt_ezeki_230 | 3) r reg.r u a.s u b.s |
| 19:16.19 | matt_ezeki_230 | in command line: |
| 19:17.03 | matt_ezeki_230 | /usr/brlcad/bin/rtedge -W -s850 -c"set dr=1" test.g reg.r >test.pix |
| 19:17.21 | brlcad | rtedge doesn't send to stdout |
| 19:17.27 | brlcad | rtedge -o test.pix |
| 19:18.03 | matt_ezeki_230 | really? I obtained a correct graphic with both cylinders... |
| 19:18.12 | brlcad | or rtedge -F/dev/Xl -o test.pix -W -s850 -c"set dr=1" test.g reg.r in order to both see it in a window and render to file |
| 19:20.25 | matt_ezeki_230 | ok the graphic is raytraced but there are any broken lines in cylinders intersection.... |
| 19:20.45 | matt_ezeki_230 | neither using -c"set dr=1"... |
| 19:24.45 | brlcad | the "broken lines" are likely related to the default curvature tolerance |
| 19:24.56 | brlcad | you'll have to tweak that as they are the same region |
| 19:25.13 | brlcad | so you're only using curvature information to draw the edge |
| 19:33.02 | CIA-9 | BRL-CAD: 03brlcad * 10brlcad/AUTHORS: bowers middle is anthony, so sayeth the rtedge manpage |
| 19:38.56 | brlcad | ahh, I see what you mean by the "broken lines".. it's because those two cynlinders begin to merge there within the edge/curvature tolerance so it's no longer considered an edge |
| 19:39.25 | brlcad | that doesn't seem to be a configurable tolerance either, it's on or off (e.g. dn=0) |
| 19:41.09 | brlcad | i could make that configurable if you really would like that feature .. depends if it's significant as edges such as those are questionably still edges |
| 19:41.55 | brlcad | e.g. you'll never really get the X to come fully together as the point at which they meet is very much a smooth crossing |
| 20:11.52 | CIA-9 | BRL-CAD: 03brlcad * 10brlcad/autogen.sh: |
| 20:11.52 | CIA-9 | BRL-CAD: woot, this little "script" is in excess of 1000 lines... in celebration, finally |
| 20:11.52 | CIA-9 | BRL-CAD: break tradition with other autogen.sh convention by adding a more useful --help |
| 20:11.52 | CIA-9 | BRL-CAD: option instead of passing it on to configure. also add support to override any |
| 20:11.52 | CIA-9 | BRL-CAD: of the command (autoreconf, automake, etc) options via a set of _OPTIONS |
| 20:11.54 | CIA-9 | BRL-CAD: environment variables. |
| 21:00.48 | *** join/#brlcad digitalfredy (n=digitalf@200.71.62.161) |
| 21:13.41 | *** join/#brlcad lg_ (n=lg_@mstr195175-16578.dial-in.ttnet.net.tr) |
| 21:13.54 | lg_ | hi... |
| 21:14.25 | brlcad | howdy |
| 21:15.03 | lg_ | finally got into the irc-channel... just a real brlcad-noob, however ;-) |
| 21:15.32 | brlcad | welcome, lots of noobs abound ;) |
| 21:16.04 | lg_ | anyone among those who ever tried brlcad's mged on a mac? |
| 21:16.24 | brlcad | plenty do, it's one of the primary dev platforms |
| 21:17.16 | lg_ | hm, than i must have done something really stupid - i could not find out how to navigate, e.g. zoom in and out, using keyboard shortcuts... |
| 21:17.34 | lg_ | thought it was an exotic platform for brlcad |
| 21:19.00 | lg_ | and without this kind of usability, it is even hard to try out ;-) |
| 21:19.47 | brlcad | the mged modeler in brl-cad wasn't designed to be "discoverable", that's likely a large portion of your difficulties with it |
| 21:20.22 | lg_ | ;-) but i suppose there is a way to navigate with keyboard / shortcuts, right? |
| 21:22.14 | brlcad | yes and no |
| 21:22.38 | brlcad | the command line is clearly key oriented and *everything* you can do with the gui can be done without the gui interface |
| 21:23.04 | brlcad | did you see the shiftgrips and mged quick reference on the website? |
| 21:23.24 | brlcad | the introduction to mged tutorial on there is also pretty good for getting walked through the basics |
| 21:23.40 | brlcad | takes just a couple hours to get through the whole series if you're diligent |
| 21:25.22 | lg_ | ok. i have mged with a model opened, and the geometry browser says there are three objects, one spere and a cylinder among those. i am in the graphics window, and the view menu claims that 'I' and 'O' should zoom in and out. Do I have to press the Apple-Key, the Ctrl-Key or anything together, or what does it mean? I simply don't see an effect. I also do not know how to move to an initial view showing all objects, if I do not know their posi |
| 21:25.44 | lg_ | ;-) I know the pdf-documentation, it is pretty well done |
| 21:30.53 | lg_ | i think my entry level problems are even too basic to be solved with the documentation - I do not know how to control the gui even |
| 21:32.03 | brlcad | mged's graphical is not the greatest ;) the command line is a fair bit more powerful |
| 21:32.13 | brlcad | there's work on-going on a new graphical interface (two projects actually) |
| 21:34.19 | lg_ | i heard about it. but don't misunderstand me, i like editing by the command line. it is just necessary to have some viewer app, e.g. to explore geometry imported from other applications. right now, i see a black mged graphics window with one bright pixel in the middle, and entering the shortcuts simply shows no effect. that is my current problem ;-) still, is there some information about the new guis available now? |
| 21:34.51 | lg_ | and, by the way, did anyone bind to the brlcad-libs from c++? |
| 21:35.49 | brlcad | ahh, you found a bug |
| 21:36.22 | brlcad | I and O were zoom in/out but the command that they bound to no longer exists |
| 21:36.22 | lg_ | no, that cannot be true, it's a bug? after all my new grey hair now??? |
| 21:36.37 | matt_ezeki_230 | brlcad: maybe I posed you a too much trivial case...or too much particular issue. If you agree I could send to you 3 files: 2 database and an "instruction files" so you can see what I need to do... |
| 21:37.18 | brlcad | matt_ezeki_230: sure |
| 21:37.23 | lg_ | i was convinced to be the most stupid brlcad-beginner in the universe (or at least the ip4-address-space) |
| 21:37.30 | brlcad | heh |
| 21:39.05 | lg_ | so what is the new way to navigate a scene / model now, if the commands where removed? or have they been removed accidentally? |
| 21:41.42 | brlcad | there's generally five different ways to get something done |
| 21:42.26 | brlcad | e.g. zoom command on the command line, first and third mouse buttons |
| 21:42.53 | lg_ | hey, one would even be sufficient... imagine me, sitting in front of mged, the geometry got imported, is shown in the geometry browser, and I cannot manage to see it |
| 21:43.09 | lg_ | ok, lets try that |
| 21:43.39 | brlcad | if you just want to see it, double click it in the geometry browser ;) |
| 21:43.47 | brlcad | right-click in browser gives a menu |
| 21:44.19 | brlcad | e or draw command will display geometry, tops command shows the top-level hierarchy |
| 21:44.36 | lg_ | it's there! |
| 21:45.00 | lg_ | the double-clicked showed it. |
| 21:45.04 | brlcad | these commands are all in the mged quick reference and are covered by the tutorials ;) |
| 21:45.16 | brlcad | (cept the browser isn't covered) |
| 21:45.39 | lg_ | i am on a mac (laptop), using a one-button mouse. so i am used to use CTRL-left to emulate the right mouse button, what seams not to work in mged! |
| 21:45.41 | brlcad | it's effectively "experimental" as there are a couple known limitations (in particular for one-button mouse systems) |
| 21:46.26 | brlcad | cmd-click will display the menu on os x |
| 21:46.48 | brlcad | move over the menu item and then hit return instead of clicking |
| 21:47.08 | brlcad | and yes, that's very clunky.. |
| 21:47.23 | brlcad | I'd suggest learning the command line first unless you have a keyboard aversion |
| 21:47.41 | brlcad | that interface will carry over to both of the new modelers too |
| 21:48.13 | lg_ | ahem...yes, keeping the cmd-key and the mouse pressed, moving at the same time to the menu item and pressing enter, i think two people are needed to operate that on my mac ;-) |
| 21:49.36 | lg_ | the new modelers could insist on keeping the eyes closed while performing these artistic operations ;-) |
| 21:50.10 | brlcad | they'll certainly be more discoverable sans documentation at least :) |
| 21:52.06 | lg_ | ok, now, after the first success, i got some motivation. don't take me wrong, I went through the doc, but i really had some problems because some basic functionality seamed not to work. and i started with an imported model, as that is where i am right now, i am considering to use brlcad on existing model data. so, even if i should have done so, i think it is not surprising that i first went through the modeling exercises. |
| 21:52.37 | brlcad | by the way.. if you happen to be hitting keys and the model starts spinning on you.. hit 0 ;) |
| 21:52.47 | brlcad | that's caused by xyzXYZ bindings |
| 21:53.09 | lg_ | beautyful |
| 21:53.43 | *** join/#brlcad DTRemenak|RDP (n=DTRemena@c-24-23-59-104.hsd1.ca.comcast.net) |
| 21:53.59 | matt_ezeki_230 | brlcad: i'm trying to send you the 3 files.. |
| 21:54.04 | brlcad | similarly (f)ront, (l)eft back right top bottom, and (3)5/25 az/el view |
| 21:55.00 | brlcad | matt_ezeki_230: thanks.. hadn't noticed.. |
| 21:55.06 | lg_ | by the way, the geometry i imported from formz by iges arrived like a charm, great. still, as far as i understood, brlcad is moving to step niw, right? |
| 21:55.43 | matt_ezeki_230 | thanks to you for your help |
| 21:55.54 | matt_ezeki_230 | brlcad: thanks to you for your help |
| 21:57.51 | lg_ | ok, i think i will play a while now ;-) |
| 21:59.00 | brlcad | lg_: as far as mouse gui bindings for os x, shift+ctrl+mouse is zoom, shift+option+mouse is translate, ctrl+option+mouse is rotate |
| 21:59.19 | brlcad | some more details on those in the shift-grips guide |
| 22:00.00 | lg_ | ah, one more question (the answer may be hidden in the docs, but i could not fing it so far), not for the gui. can i use wildcards to operate on many objects, that are not grouped? as mentioned, i have existing geometry imported. |
| 22:00.09 | brlcad | lg_: some work on a step importer was started though it is currently stalled effort as the lead dev that was working on it is off on something else right now |
| 22:00.35 | brlcad | we do have the step specs though, and an updated version of the nist express parser |
| 22:01.40 | lg_ | that's a pity, as the iges importer seams to support only rather old iges4. while, as menitioned, the geometry is fine, i could not use e.g. blocks (or references or instances or symbols or however it is called) |
| 22:01.46 | brlcad | lg_: yes you can -- the mged command line has two evaluation modes -- the default is like a command shell with globbing operators eg ls *.r ; the other is tcl evaluation which is a lot more involved to explain if you don't know the Tcl language |
| 22:02.10 | lg_ | yes, i know a bit about the story, i am on the free-architecture mailing list, and the step issue had been discusses |
| 22:02.12 | brlcad | yeah, it would be nice if someone worked on updating the iges converter too |
| 22:02.25 | brlcad | limited resources and time of course.. |
| 22:02.48 | lg_ | ok, no surprise if you are working on the gui and frontend stuff |
| 22:03.05 | brlcad | yeah, that's the most "hurting" imho |
| 22:03.51 | brlcad | we can hook in all existing importers and exporters and still be worlds better than most other projects even with the limited functionality and limitations on some of them |
| 22:04.13 | brlcad | but it's hard to avoid the old mged gui at the moment |
| 22:04.22 | brlcad | people expect/want something radically different |
| 22:04.33 | brlcad | most at least |
| 22:04.37 | lg_ | yes, but that is bound to specific applications |
| 22:04.49 | brlcad | very much so |
| 22:05.59 | brlcad | the follow up modeller addresses that problem with plugin profiles so you can have things like a gui with terminology changes specific to domains (e.g. CADD, CAE, CAM, CAA, etc) |
| 22:07.31 | lg_ | i dream of something like small, nice tools to create high-level objects for architecture, a manager app that creates geometry from these and sends them to brlcad, a viewer showing them with librt, tracing view rays to find objects when the user hits a mouse button to implement snap, tracing back from the object to the again small app that generated the high level object that initiated its creation... nice design for an architectural cad sy |
| 22:08.29 | brlcad | initial focus will likely be towards generic CAD integrating all of our existing functionality for geometry management, image manipulation, importers, exporters, renderers, etc |
| 22:09.04 | brlcad | then move on towards whatever the community demands most (which at this point seems to be CADD, CAM, and CAA in probably that order) |
| 22:09.16 | lg_ | ok, a generic 3d cad would be an outstanding project right now |
| 22:10.01 | brlcad | we have all the implementation pretty much done, the brl-cad libraries provide 80-90% of what is needed already implementation-wise |
| 22:10.10 | brlcad | what's missing is a gui ennvironment to wrap it all up in |
| 22:10.24 | lg_ | my interest is in architectural cad, and that requires a level above the geometry. that does not mean that generic cad cannot be used for architecture, it is another way to work. |
| 22:10.40 | lg_ | yes, the gui is a key issue to get these things accepted |
| 22:11.14 | brlcad | sure, architecture (CAA) is very closely related to CADD needs.. drafting diagrams, constraints, parametrics, rendered plans, etc |
| 22:11.46 | brlcad | lot more focus on 2D than you generally see in solid modeling leveraging extrusions more, sketches, etc |
| 22:12.24 | lg_ | yes, but for caad you need some kind of intelligence in the objects. i do not like the 2d approach too much, i want an useable 3d model to extract information from |
| 22:13.24 | lg_ | what i mean by intelligence is that a window belongs to a wall, that a corner of a wall is constructed by well-known rules depending on material etc, and that the actual geometry is the result of these |
| 22:14.37 | lg_ | small generator applications could handle that very well, and i would feed the brlcad not directly by user input, but by a library of objects that the user generates |
| 22:15.12 | bjorkBSD | window belongs to a wall ... |
| 22:15.16 | bjorkBSD | or ANYTHING which would hold it. |
| 22:15.21 | bjorkBSD | door ... |
| 22:15.23 | bjorkBSD | roof. |
| 22:15.44 | bjorkBSD | floor? (wtf?!) |
| 22:16.21 | bjorkBSD | but idiomatically (note linguistic metaphor) |
| 22:16.25 | bjorkBSD | it belongs in a wall. |
| 22:17.04 | lg_ | yes, but the object should "know" this. it can still be flexible, but for architecture, it makes a lot of sense that placing a window in another object implies that an opening is cut into that object, and, if the object is moved, the window changes its position, too (some kind of parent-child-hierarchie here). |
| 22:17.05 | bjorkBSD | ... and if you're a certain company in redmond, windows belong everywhere :D |
| 22:17.29 | lg_ | than it belongs under the earth... |
| 22:17.30 | bjorkBSD | i thought that was the whole point of CSG |
| 22:17.37 | brlcad | those are constraints and hierarchical geometry views in a generic sense, also being thought about.. how I could say, for example, that I want to create a "typical" brick-mortar 2-level house |
| 22:18.06 | brlcad | where i'd only specify dimensions, numbers of windows, doors, etc .. extreme example given the size/complexity but gives the idea |
| 22:19.32 | brlcad | similar smaller scale might be that I want a nut and bolt, and I'd just specify the length, threadsize, etc and the model would retain that knowledge and the underlying connections (e.g. all the primitives or splines or whatever used to make the threads unioned to a cylinder, etc) |
| 22:19.33 | lg_ | yes, it should still be allowed to influence and manipulate, but, for example, the bricks are only available in some certain sizes, that defined usual openings, all kinds of dimensions, construction details. it is a whole layer over the actual csg stuff happening |
| 22:20.42 | bjorkBSD | lg have you ever used catia by any chance? |
| 22:21.23 | lg_ | no, i think it is close to what i want |
| 22:22.56 | bjorkBSD | fork up the dough :D |
| 22:23.08 | bjorkBSD | or **** it and tell us what it's like. |
| 22:23.16 | bjorkBSD | <PROTECTED> |
| 22:24.23 | lg_ | bjorkbsd, are you from a archicture-related field? |
| 22:24.57 | lg_ | i ask because i am curious about people using this kind of apps in architecture. most caad is much more primitive imho |
| 22:26.29 | brlcad | invest dozens of thousands into catia and company or invest that money/time/interest by contributing to an open source CAD project ;) |
| 22:27.29 | lg_ | i would prefer the second, while it is hard for someone who is not really good in writing code (i am actually an architect) |
| 22:28.02 | bjorkBSD | not at all, lg_ |
| 22:28.10 | bjorkBSD | i'm from a "wow this is COOOL" field. |
| 22:28.37 | brlcad | if that's not brl-cad, so be it, but it's a shame that so many are willing to dump so much money and time into commercial packages when there's clearly enough interest and talent to get something together |
| 22:29.00 | bjorkBSD | i'm especially interested in the interface. |
| 22:29.13 | lg_ | still i try to contribute by collecting information, promoting the idea etc. just set up a website to allow people to share their ideas on it (free-architecture.org). |
| 22:29.16 | brlcad | all I know is that we're a couple decades of development ahead of the other open source "competition" and we still have a couple decades of dev-time catch-up to go |
| 22:29.21 | bjorkBSD | but curse the paucity of coding skills! |
| 22:29.51 | lg_ | bjorkbsd, what is wowsocool? |
| 22:30.03 | brlcad | lg_: it's all appreciated ;) actually your architectural expertise alone is enough to warrant respect and be useful |
| 22:30.28 | brlcad | no single project can cross domains without having consistent participation from the major CAD domains |
| 22:31.37 | bjorkBSD | lg_ modeling mechanical components and watching mechanisms come to life until i can afford to create them. |
| 22:32.00 | brlcad | right now the biggest void is a lack of developer involvement, though documentation, tutorials, and advertising (website, release notes, news, conferences) take up a huge amount of time too |
| 22:32.22 | lg_ | i think if the brlcad people continue work on a really robust csg engine, it is the best they can do, as there are few implementations i know about. still, i would love to see some coordinated effort to allow to chain all these tools to form something more complete and useable. and take e.g. the load of developing viewers and guis from those working on the engin |
| 22:32.56 | lg_ | bjorkBSD, ok, mechanics, I guess your are closer to the typical brlcad user profile than me than ;-) |
| 22:34.08 | matt_ezeki_230 | brlcad: now i go away (to sleep because here in Italy it is 12:34 A.M. :-) ). If would you be so kind as to try the code in file README....you have my e-mail (file README again) but i'll be back here tomorrow. Bye and thanks a lot for your help! |
| 22:34.21 | brlcad | lg_: brl-cad's CSG engine and raytrace engines is actually one of the oldest and best around -- better than just about every commercial iplementation as well from a robustness and performance perspective |
| 22:34.41 | bjorkBSD | comparable to catia? :O |
| 22:34.46 | lg_ | hey, matt_ezeki, good night, here in istanbul it's already 1.34 will be hard tomorrow morning |
| 22:34.59 | brlcad | lg_: you're right though.. a coordinated effort is desparately needed and leveraging existing functionality |
| 22:35.11 | brlcad | bjorkBSD: actually yes |
| 22:35.15 | bjorkBSD | <PROTECTED> |
| 22:35.22 | bjorkBSD | then we have a gold mine on our hands :-? |
| 22:35.39 | matt_ezeki_230 | lg_: he he ! yes, will be hard :-) |
| 22:35.41 | brlcad | brl-cad libraries were designed with large-models and performance in mind, primary focus has always been on analysis purposes |
| 22:36.24 | brlcad | where it very much lacks is in the modeling interface side |
| 22:36.29 | bjorkBSD | how did you become involved with brl-cad, brlcad ? |
| 22:36.54 | brlcad | and support for brep/nurbs primitives.. that's one weak area in the engine |
| 22:38.05 | lg_ | i have been so fed up with the fact that all the csg-engined I used in architecture led to geometry that was not suitable for any other editing after some booleans. so i appreciate a csg engine ;-) |
| 22:38.08 | brlcad | bjorkBSD: anyone can become involved.. I became involved many years ago working under Mike Muuss (BRL-CAD's original architect) |
| 22:38.51 | brlcad | lg_: are you familiar with where brl-cad comes from and what it is/was designed for? |
| 22:38.53 | lg_ | brlcad, seams that you have been in the development for a while? |
| 22:39.24 | brlcad | about 7 or so years now, lead developer |
| 22:39.47 | lg_ | brlcad, yes, i think that is why its powerful, it is to create models for analysis, while many modelers and cad i used where for visualization or limited complexity |
| 22:40.20 | brlcad | i was behind the pressure to make it an open source project, took several years and lots of effort .. to say the least |
| 22:40.55 | lg_ | can imagine it was hard to make this move with an established project like brlcad |
| 22:42.22 | lg_ | i follow the development of radiance, which was also open-sourced some years ago, after maybe 20 years of development, also controlled by governmental (and other) institutions. still, it is great that these things happened finally, so we can send a thank you know, because we can play around with all these nice toys ;-) |
| 22:43.07 | brlcad | yeah, radiance is pretty cool .. some of the brl-cad devs used to work with them way back when |
| 22:43.57 | bjorkBSD | hmmm. |
| 22:44.39 | brlcad | difference there being radiance was somewhat lingering/abandoned so it was open source, whereas brl-cad continues to be actively used and developed but was open sourced to "give back" and spark expanded collaboration/development/etc |
| 22:46.34 | brlcad | i've toyed with the idea of writing a radiance-compatible interface to brl-cad for a while myself.. or even implementing a new tracer based on brl-cad's raytrace library |
| 22:46.54 | brlcad | infinitely doable, very cool stuff |
| 22:47.16 | lg_ | well, i do not know the internals of radiance development over the time, but there is a lot of development around it, many projects forked, and a lot of commercial applications (most of the serious ones) are based on it. the problem is that is seams to be hard for new devs to join the core dev team |
| 22:47.18 | brlcad | hard to stay focused on the modeler :) |
| 22:47.36 | bjorkBSD | brlcad, i can read C to some extent, but when you move this stuff to c++ ... |
| 22:47.40 | bjorkBSD | that'll be the end for me :( |
| 22:47.59 | *** join/#brlcad DTRemenak (n=DTRemena@c-24-23-59-104.hsd1.ca.comcast.net) |
| 22:48.26 | brlcad | bjorkBSD: the core will likely always remain C .. most of the initial 300+ plugins are all C |
| 22:48.32 | lg_ | yes, but i think an even cooler interface would be to the more general simulation environment esp-r, than i could use the solid model to do all kinds of thermal simulation, too, and still use the interface from esp-r to radiance ;-) |
| 22:49.03 | brlcad | bjorkBSD: it's only the higher level logic that will be C++ and even that will be utilizing the C layers :) |
| 22:49.40 | brlcad | lg_: that's the problem with CAD development in general.. it spans so many genre's it's hard to stay focused on just a few concepts and do them well :) |
| 22:49.57 | bjorkBSD | <PROTECTED> |
| 22:50.13 | lg_ | that is nice, it means i will see some examples how to call the brlcad libs from c++ ( i know a bit c++) |
| 22:50.42 | brlcad | that's why the focus is on a "platform" for plugin-based development so it becomes more manageable to have isolated groups focus on their genre of expertise/interest |
| 22:51.26 | lg_ | but, for the radiance interface, you would have to mesh all the geometry, as radiance supports no csg at all. and than you can also use the great (new) wavefront-obj support in radiance, so g2obj is all we need right now |
| 22:52.09 | brlcad | something akin to eclipse for CAD mixed with a game, physics, and render engines :) |
| 22:52.43 | lg_ | i am really curious about what you are doing there. will it be open-source, too, and can we lurk to the dev code soon? ;-9 |
| 22:52.46 | brlcad | lg_: we do the geometry meshing now for a variety of purposes.. half the exporters are mesh-only |
| 22:53.24 | brlcad | the ADRT interface that was added to brl-cad a year or so ago is a high-performance triangle-only engine as well |
| 22:53.31 | lg_ | cad-eclipse, i know a lot of people who dream about this idea (even if it may be sick to dream of cad) |
| 22:53.50 | brlcad | twing created an interactive triangle-only raytracer and path tracer based around it |
| 22:54.23 | brlcad | that's what made things like http://ftp.brlcad.org/images/stryker_slat.png possible |
| 22:55.17 | brlcad | sticking CAD into Eclipse is just wrong .. there are fundamental lessons that can be learned from their project though |
| 22:56.11 | lg_ | brlcad, for radiance, obj is great, the only point where a special exporter would make sense are primitives like spheres, cylinders etc which are supported better then meshing them. but in real models, these primitives are rare. cad in eclipse is nonsense, but the idea of a development framework with plug-ins is great |
| 22:57.08 | lg_ | still loading the image, i am working on 56k-modem-speed right now :-( |
| 22:57.52 | brlcad | correct me if i'm wrong but radiance is both an interface for rendering as well as a rendering system itself. it shouldn't be that much work to implement the radiance interface (so data files, procedural textures, geometry, etc work unmodified) |
| 22:58.26 | brlcad | or even hook in the radiance renderer into some CAD system as a plugin ala how blender has various rendering methods exposed |
| 22:59.03 | brlcad | lg_: eek.. 56k?! didn't know people still use those.. :) |
| 22:59.52 | lg_ | should be possible, the core of radiance is available as a library. i am not sure how easy it is to use the different ways radiance optimizes rendering, as the octree-sorting etc. |
| 23:00.23 | lg_ | there is even a clean and new c++-rewrite, that allows to access all functionality plus some extras from c++ |
| 23:00.43 | brlcad | i'd bet adrt outperforms radiance for full global illumination rendering |
| 23:01.27 | lg_ | (yes, i still don't have a dsl line, cause the phone is registred on the owner of my flat, so i have to wait until she is doing all the paper work. so long i enjoy the good-old-times internet feeling) |
| 23:03.01 | lg_ | (the radiance-rewrite is called radzilla, the author is Carsten Bauer, a physician and always open to good ideas, more info on http://www.cb-d.de/radzilla.html, if you want to spend your spare time ;-) |
| 23:03.02 | brlcad | adrt was implemented based off of recent interactive raytracing research, some of the techniques presented at siggraph over the past 3 years or so only |
| 23:03.28 | lg_ | ok, got the image, you render this interactively? |
| 23:04.23 | brlcad | no no.. that was a full light transport simulation |
| 23:04.37 | lg_ | there has been some nice interesting real-time raytracing work done at the university of saarbrŸcken. they have both a software (and open-source for non commercial use, afaik) and a hardware version |
| 23:04.57 | brlcad | took several days, 8 trillion rays, many million per second |
| 23:05.16 | lg_ | ok, and the problem was the number of triangles? |
| 23:06.01 | brlcad | saarbricken? not sure I've heard of them |
| 23:06.28 | lg_ | saarbrucken (u-umlaut, in germany) |
| 23:07.07 | brlcad | we (used in the singular sense of the dev that collaborated with them on ideas) worked with saarland .. names escaping me at the moment |
| 23:07.46 | lg_ | yes, saarland is the state, saarbrucken the "capital" |
| 23:07.58 | lg_ | http://www.openrt.de/ |
| 23:08.03 | brlcad | yeah, that's them |
| 23:08.18 | lg_ | they have some nice projects, i think |
| 23:08.19 | brlcad | slusalek |
| 23:08.29 | brlcad | we interacted with them pretty extensively |
| 23:09.14 | lg_ | did not know that. i never met, found out by a friend and followed a bit what they developed. now they seam to try making some money |
| 23:09.28 | brlcad | adrt actually outperforms for some geometries (non-hardware implementation) |
| 23:09.41 | brlcad | yeah, they are focused on selling the library |
| 23:10.00 | brlcad | adrt is effectively an open source free version developed by "someone else" ;) |
| 23:10.51 | lg_ | ;-) i have rendered quite large scenes in radiance by using instances, which allows gigantic geometry by sharing memory. but it is certainly not the renderer for those in need of speed. |
| 23:10.59 | lg_ | will have a look at adrt |
| 23:11.02 | brlcad | basically, an implementation of a lot of ingo wald's research |
| 23:11.38 | brlcad | adrt's deficiencies are in it's polish, user interface is barely existant .. if you can't code, it's not really useable without a lot of hand-holding |
| 23:12.33 | lg_ | i have an awful model, we are rendering on a 4node cluster for two months now to get 20 pctures 1600x1200. maybe adrt might be something if we ever try to get animations. |
| 23:12.38 | brlcad | brl-cad's librt also has an implementation of shirley's photon mapping, but that is a tricky beast for solid models .. |
| 23:13.27 | lg_ | ok, that is a field unknown to me, an optic effect that never (?) appears in architectural scales |
| 23:13.30 | brlcad | adrt has a layer called isst (interactive shot selection tool) that is used for realtime raytracing of very large detailed models (full vehicles) |
| 23:13.47 | brlcad | that obtains about 20-30fps on a 4 node cluster |
| 23:13.55 | lg_ | (while i am not sure if this is really true when i think of a dome covered by specular material) |
| 23:14.52 | lg_ | the model i am talking about is a model of the whole interior of the hagia sophia, with a resolution down to 1x1 cm |
| 23:15.25 | brlcad | a volumetric geometry? |
| 23:15.36 | brlcad | or it is also facetized? |
| 23:15.58 | lg_ | we cannot process it in any cad, we hold a directory tree of geometry files and assemble the model only during the rendering. i once converted it to a vrml and got a 3GB file... |
| 23:16.22 | lg_ | it is not volumetric, it is for radiance, only triangles and polygons |
| 23:16.48 | *** join/#brlcad DTRemenak (n=DTRemena@c-24-23-59-104.hsd1.ca.comcast.net) |
| 23:16.54 | brlcad | 3GB isn't that large for us, especially if that's the vrml |
| 23:17.33 | brlcad | that should easily fit in core |
| 23:17.39 | lg_ | the most of the geometry is held in obj and dxf files. i could not browse it at least, for interactive navigation |
| 23:17.59 | lg_ | i should try it? |
| 23:18.11 | brlcad | oh i believe you.. most CAD systems fall apart on large models :) |
| 23:18.53 | brlcad | at least when trying to bring it all up, that's where we generally do much much better assuming the model will at least fit in memory |
| 23:19.06 | lg_ | what kind of input interface could i use? if i have to access it from an own application, what language bindings do exist, what geometry formats are already supported? got curious |
| 23:19.46 | brlcad | depends if you'd want to view it with adrt or librt .. the latter has much more tool support, adrt is going to be much faster for triangles |
| 23:20.12 | lg_ | most cad systems i know have some kind of magic limit for geomety size / memory ;-) |
| 23:20.19 | brlcad | there isn't an obj importer (only an exporter) though.. so some processing would be needed |
| 23:20.46 | lg_ | the idea was to have adtr for setting up a way to animate the whole scene |
| 23:20.48 | brlcad | would be trivial to add one, just nobody has needed it yet |
| 23:20.59 | lg_ | (which is impossible in radiance now) |
| 23:21.07 | bjorkBSD | brlcad, were you the one who mentioned a playstation/xbox keypad interface? |
| 23:21.13 | lg_ | what is supported? |
| 23:21.30 | brlcad | once it's in brl-cad's .g format, getting it to adrt is pretty trivial -- otherwise a direct dump to adrt format is possible as well (it's pretty simple raw format) |
| 23:21.45 | lg_ | don't tell me you are running brlcad on a playstation? ;-) |
| 23:21.55 | lg_ | ok, i can get it into brl-cad format |
| 23:22.02 | brlcad | bjorkBSD: i'm not sure.. doesn't sound familiary |
| 23:22.27 | brlcad | lg_: do you have brl-cad installed now? |
| 23:22.33 | bjorkBSD | eh. a video games interface, basically. |
| 23:22.47 | bjorkBSD | where all you have is the control pad for interacting with it. |
| 23:23.04 | brlcad | lg_: if you do, ls -la /usr/brlcad/bin/*-g are the importers |
| 23:23.22 | lg_ | bjorkBSD, i am wondering if a psp3 will be my next linux installation;-) brlcad, yes of course, i am playing with it, i tested the iges-interface so far |
| 23:23.31 | *** join/#brlcad DTRemenak|RDP (n=DTRemena@c-24-23-59-104.hsd1.ca.comcast.net) |
| 23:23.34 | brlcad | g-* are the exporters as well as asc2g and g2asc for the ascii intermediate format |
| 23:24.50 | brlcad | most common importers probably being iges, vrml, dxf, off, ply, and stl |
| 23:25.10 | brlcad | there is a binary export plugin for Pro/Engineer if you happen to have that |
| 23:25.31 | lg_ | yes, i would have liked an obj importer, but iges is nice as i can export natively from the modelers |
| 23:25.41 | lg_ | i have most of the model in formz |
| 23:25.43 | brlcad | and a Unigraphics exporter as well, though I haven't compiled that in ages |
| 23:26.09 | lg_ | eh, does the vrml-importer support the old vrml1? |
| 23:26.33 | brlcad | it used to |
| 23:26.35 | brlcad | :) |
| 23:27.10 | lg_ | hm, there is no vrml-g or anything similar |
| 23:27.11 | brlcad | it was updated to vrml 2.0 several years ago without retaining backwards support (don't ask me why..) |
| 23:27.31 | brlcad | ahh, that's right.. i mean the exporter |
| 23:27.57 | bjorkBSD | logo anyone? |
| 23:28.02 | brlcad | vrml is one of the other bastard few importers that hadn't been written yet |
| 23:28.13 | lg_ | than maybe the reason was to get rid of all the vrml-1 related support questions. |
| 23:28.13 | bjorkBSD | repeat 4 [fd 100 left 90] :D |
| 23:28.16 | lg_ | logo? |
| 23:28.19 | brlcad | trivial format, just one more of those things that someone could spend a week or two on |
| 23:28.43 | brlcad | don't generally get CAD solid models in vrml format :) |
| 23:28.47 | bjorkBSD | ANCIENT lisp dialect and turtle geometry platform |
| 23:29.04 | lg_ | yes, i remember, bjork... |
| 23:29.06 | lg_ | ;-) |
| 23:29.12 | bjorkBSD | fun stuff aye? |
| 23:29.53 | lg_ | ok folks, i think my boss will kick me out of the office tomorrow when he finds me sleeping there |
| 23:30.00 | brlcad | heh |
| 23:30.09 | lg_ | it is half past three (in the morning) here! |
| 23:30.11 | brlcad | lg_: good talking to you, hope to see you around more often ;) |
| 23:30.17 | brlcad | where abouts is here? |
| 23:30.22 | bjorkBSD | turkey! |
| 23:30.38 | lg_ | certainly, you have been great help. i am in istanbul, turkey |
| 23:30.40 | brlcad | ~tr |
| 23:30.42 | ibot | turkish is #turklug, or #debian.tr, or trke yardm iin #debian.tr |
| 23:30.48 | brlcad | so it is |
| 23:30.58 | bjorkBSD | ~ibot |
| 23:31.07 | bjorkBSD | oh fine don't talk to me [-( |
| 23:31.11 | bjorkBSD | ~ibot_ |
| 23:31.12 | ibot | well, ibot_ is not me |
| 23:31.18 | lg_ | yes, in fact i am german, but i am living here |
| 23:31.26 | brlcad | lg_: aha |
| 23:31.32 | lg_ | so good night / morning! |
| 23:31.35 | brlcad | ciao! |
| 23:31.43 | brlcad | ~x en de good night! |
| 23:31.47 | bjorkBSD | is ibot a bot? |
| 23:31.56 | lg_ | ;-) |
| 23:31.57 | bjorkBSD | <PROTECTED> |
| 23:32.02 | lg_ | multilingual bot... |
| 23:32.21 | bjorkBSD | ~brl-cad |
| 23:32.22 | ibot | it has been said that brl-cad is a powerful constructive solid geometry solid modeling system that includes an interactive geometry editor, ray tracing support for rendering and geometric analysis, network distributed framebuffer support, image and signal-processing tools. |
| 23:32.27 | lg_ | we say iyi aksamlar here, by the way |
| 23:32.30 | bjorkBSD | <PROTECTED> |
| 23:32.58 | brlcad | lg_: is that turkish? |
| 23:33.02 | lg_ | yes |
| 23:33.32 | lg_ | at least it should be, i am still learning ;-) and it is difficult in irc because all the special characters disappear |
| 23:33.38 | brlcad | bueno, pues.. iyi aksamlar ;) |
| 23:33.41 | *** join/#brlcad DTRemenak|RDP (n=DTRemena@c-24-23-59-104.hsd1.mn.comcast.net) |
| 23:34.03 | lg_ | great, cu, and thanks for help and nice talk |
| 23:34.08 | lg_ | lars |
| 23:34.16 | *** part/#brlcad lg_ (n=lg_@mstr195175-16578.dial-in.ttnet.net.tr) |
| 23:34.47 | bjorkBSD | brlcad, do you ever use a lightpen or tablet with brlcad? |
| 23:34.59 | bjorkBSD | seems they're very uncommon. |
| 23:36.30 | brlcad | actually just purchased a tablet to test with brl-cad, should be here next week |
| 23:37.16 | bjorkBSD | how much did it cost? |
| 23:39.07 | brlcad | not much 100-150 or something |
| 23:40.20 | bjorkBSD | definitely not as cheap as a keyboard and mouse :D |
| 23:41.26 | bjorkBSD | the game industry ... i'd love to take a good look at spore sometime. |
| 23:58.26 | ``Erik | neat |
| 23:58.29 | ``Erik | wacom? |
