| 00:14.12 | brlcad | gah, drat |
| 00:14.28 | brlcad | starseeker: tcl upgrade is still missing something, maybe in the init.tcl script |
| 00:15.08 | brlcad | regress test is failing on the fastgen test with a Tcl_Init error Can't find a usable init.tcl in the following directories... |
| 00:15.31 | brlcad | hm, actually all the mged tests fail with the same error |
| 00:15.36 | brlcad | (on mac 10.4) |
| 00:16.02 | Twingy | o.O |
| 00:18.00 | *** join/#brlcad smurfette (n=Pandora@c-69-242-189-29.hsd1.mo.comcast.net) |
| 00:18.11 | brlcad | starseeker: the failure is on an uninstalled build -- the auto_path is missing the source search directories which I believe is a mod to our bundled sources |
| 00:38.51 | starseeker | ok, I'll see if I can find it |
| 00:41.11 | starseeker | realizes he's probably going to have to do a diff on the old tree vs. the default tarball |
| 00:41.14 | starseeker | argh |
| 00:50.24 | brlcad | svn log the dir, there aren't that many changes |
| 00:50.59 | starseeker | is it an atomic change or was it rolled in as part of an update? |
| 00:56.49 | starseeker | oh, the unix/Makefile.in? |
| 01:08.04 | CIA-40 | BRL-CAD: 03starseeker * r33844 10/brlcad/trunk/src/other/tcl/unix/Makefile.in: Add back in unix Makefile.in changes |
| 02:33.53 | starseeker | tries to figure out where auto_path is getting /usr/lib |
| 02:49.48 | starseeker | oh, it's sucking in my system TCLLIBPATH |
| 02:59.30 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-71.sbndin.btas.verizon.net) |
| 03:04.12 | starseeker | blinks |
| 03:04.24 | starseeker | the failing line is: $this configure -menu $itk_component(menubar) |
| 03:04.41 | starseeker | $this contains ::.archer0 |
| 04:44.49 | starseeker | ah, ding nabbit - I forgot my list of National Archive stuff at work |
| 04:45.16 | starseeker | will have to go in on monday |
| 05:04.18 | starseeker | ah, nuts |
| 05:04.33 | starseeker | me remembers regression testing currently needs an in-tree build and starts over |
| 05:17.22 | starseeker | ah HAH |
| 05:17.38 | starseeker | didn't look at the auto_path closely enough |
| 05:18.56 | starseeker | http://pastebin.bzflag.bz/m5ffe7078 |
| 05:19.08 | starseeker | somehow, /usr/lib/* stuff got sucked in |
| 05:19.41 | starseeker | anybody else tried archer with a recent svn checkout built on a machine with system tcl/tk/itcl/itk already installed? |
| 05:19.55 | starseeker | but using local tcl/tk/itcl/itk for brlcad? |
| 05:35.35 | starseeker | eyes libtclcad's tclcad_auto_path |
| 08:48.36 | *** join/#brlcad _sushi_ (n=_sushi_@84-72-93-63.dclient.hispeed.ch) |
| 10:00.44 | alex_joni | how would one go on with converting a mesh to a solid? |
| 11:48.26 | d-lo | alex_joni: You would have to enclose the mesh or, if its easier, extrude it out into a solid. |
| 11:48.48 | alex_joni | d-lo: I think the mesh is closed |
| 11:49.44 | d-lo | alex_joni: Well, if its already enclosed, there is a property you can set to make it a solid instead of just a shell. |
| 11:49.47 | alex_joni | (it's a 3D mesh in STL or any other format) |
| 11:50.17 | d-lo | Get the mesh into solid edit mode. |
| 11:50.59 | alex_joni | d-lo: downloading brl-cad atm :) |
| 11:51.18 | alex_joni | (7.12.6 win) |
| 11:52.19 | d-lo | Good deal! :D |
| 12:01.37 | d-lo | Meshes are called BoTs in Brlcad btw. |
| 12:10.43 | alex_joni | d-lo: got a link to a crashcourse in brlcad? |
| 12:10.52 | alex_joni | or.. can I do the stuff I need in Archer? |
| 12:11.49 | d-lo | Archer and BRLCAD are very similar. BRLCAD is much more functional whereas Archer is less-so, but much prettier on the User Inferface side of the house. |
| 12:12.02 | d-lo | Both, however, should do what you want to do with BoTs.... maybe |
| 12:12.03 | alex_joni | I did start mged in the past |
| 12:12.20 | alex_joni | how would I go on with importing a STL ? |
| 12:12.21 | d-lo | http://brlcad.org/wiki/Documentation |
| 12:13.12 | d-lo | page 21 of this pdf: |
| 12:13.16 | d-lo | http://brlcad.org/w/images/6/66/Converting_Geometry.pdf |
| 12:13.26 | d-lo | shows the stl-g conversion steps. |
| 12:13.31 | alex_joni | coo |
| 12:14.00 | d-lo | It is common practice to try several conversion paths if the initial try doesn't work. |
| 12:14.38 | d-lo | aka if stl->g path doesn't work, load the stl file into pro/E or Unigraphics and export a DXF... then try the dxf -> g |
| 12:15.12 | d-lo | I have used Blender on occasion to convert files to .dxf and use the dxf->g converter.... |
| 12:15.30 | d-lo | how large is the stl file you are talking about? |
| 12:15.54 | alex_joni | hang on.. just figured out I have only obj's |
| 12:15.58 | alex_joni | converting them to STL now |
| 12:16.00 | alex_joni | < 2MB |
| 12:16.32 | alex_joni | 100k for the STL |
| 12:16.45 | d-lo | ok, no ginormous files then. |
| 12:16.59 | alex_joni | (not at this point ;) |
| 12:17.04 | alex_joni | I have some 40M ones |
| 12:17.07 | alex_joni | but not for now |
| 12:17.18 | alex_joni | STL needs to be ASCII? |
| 12:17.34 | alex_joni | n/m -b does the trick |
| 12:17.41 | d-lo | there were some recent reports of dxf files acting rather strange once the file size approaches 1.5 gigs... |
| 12:17.45 | alex_joni | Using solid name: s.stl |
| 12:17.45 | alex_joni | 2100 facets |
| 12:17.46 | alex_joni | Making region (r.stl) |
| 12:17.58 | d-lo | sounds like its working :) |
| 12:18.01 | alex_joni | d-lo: ROFL :) |
| 12:18.05 | alex_joni | 1.5gig :) |
| 12:18.39 | alex_joni | ok, the .g opened in archer |
| 12:18.52 | alex_joni | I have a tree on the left (all->r.stl->s.stl) |
| 12:19.19 | d-lo | sounds like it worked |
| 12:19.53 | alex_joni | stupid question.. is there a way to see it? |
| 12:20.03 | alex_joni | or should I raytrace it? |
| 12:20.32 | *** join/#brlcad mafm (n=mafm@223.Red-83-49-86.dynamicIP.rima-tde.net) |
| 12:20.42 | mafm | hi |
| 12:21.40 | d-lo | raytracing should work |
| 12:21.43 | d-lo | Hi mafm! |
| 12:33.56 | *** join/#brlcad elite01 (n=omg@unaffiliated/elite01) |
| 12:38.34 | alex_joni | 14:15 < d-lo> there were some recent reports of dxf files acting rather strange |
| 12:38.34 | alex_joni | <PROTECTED> |
| 12:38.43 | alex_joni | oops.. wrong paste :/ |
| 12:42.23 | alex_joni | d-lo: can you check if this is ok? http://eneas.juve.ro/~juve/brlcad/s2.g |
| 12:44.45 | d-lo | Looks good so far... what did you want me to check> |
| 12:44.46 | d-lo | ? |
| 12:46.00 | alex_joni | if there is anything there |
| 12:46.09 | alex_joni | I can't seem to get it to display or raytrace anything |
| 12:46.24 | alex_joni | ahhh... found it ;) |
| 12:46.33 | alex_joni | need to right-click on s.stl and enable the view |
| 12:46.45 | d-lo | Okay, well i found a problem. |
| 12:46.54 | d-lo | the 'bolt holes' on the base are an issue. |
| 12:47.24 | d-lo | they are supposed to be a subtraction from the main bot, but instead, they are *part* of the main bot.... thus they are confusing the raytracer. |
| 12:47.42 | alex_joni | d-lo: I don't really care about the holes ;) |
| 12:47.53 | alex_joni | I need to weld the subparts, so the holes are not interesting .. |
| 12:47.54 | d-lo | They should be removed or replaced. |
| 12:48.02 | alex_joni | I see |
| 12:48.09 | d-lo | well, they holes are making the geometry invalid. |
| 12:48.40 | d-lo | just do a raytrace looking down into the hole from the top and then up from the bottom.... its a magic 'hole' :) |
| 12:49.58 | d-lo | Since they were supposed to be a subtraction, the holes are 'mini-bots' in themselves, so they should be seperate solids. Since they aren't, there are two solids existing in one and have overlapping faces. |
| 12:50.35 | d-lo | the raytracer determins solidity by the amount of faces hit along the path of a ray (1st = in, 2nd = out, 3rd = in, etc) |
| 12:51.03 | d-lo | so when there are two faces overlapping, you get the first two hits immediately |
| 12:51.28 | d-lo | and trick the raytracer into thinking its exited the solid, when, in reality, its inside it. |
| 12:51.50 | d-lo | Now, this may not matter to your needs, but it will make the raytraced pictures inaccurate. |
| 12:52.27 | alex_joni | I saw that |
| 12:53.25 | d-lo | Looks like the same phenominon is happening on the top with those 4 (or 8) cyls |
| 12:54.23 | d-lo | I have some homebrewed tools that can eliminate the boltholes and those cyls very quickly if you want me to. Just need an email addy to send the file back when I am done. |
| 12:54.49 | alex_joni | d-lo: I have about 10 of these files, but I don't think the bother is worth it :) |
| 12:55.04 | alex_joni | if I can convert it to a solid (even if it looks funny around those holes), I'm happy |
| 12:55.34 | alex_joni | now.. you said there is a flag I can change so I can convert it from g to some other solid format? |
| 12:55.45 | d-lo | alrighty. Just noticed that the smaething is happening in the middle with those tab thingys about halfway up |
| 12:56.04 | d-lo | Just curious, why do you need it 'solid' ? |
| 12:56.07 | alex_joni | I notice in attributes it says" |
| 12:56.18 | alex_joni | "This is a solid object (not just a surface)" |
| 12:56.59 | *** join/#brlcad hippieindamakin8 (n=hippiein@202.3.77.141) |
| 12:59.27 | d-lo | I believe thats the setting I was talking about.. my desktop is chewing on optimizing your BoT... when its done I will look at that attr. |
| 13:00.31 | d-lo | where did you see that message>? |
| 13:02.44 | alex_joni | in archer under advanced |
| 13:02.57 | alex_joni | I switched to mode Advanced |
| 13:03.03 | alex_joni | and I have an attributes tab on the right |
| 13:03.09 | starseeker | fyi, Archer is experimental |
| 13:03.20 | starseeker | mged is still the "workhorse" at the moment |
| 13:03.42 | alex_joni | starseeker: I know that.. a bit too much to chew on atm though |
| 13:03.51 | starseeker | ah :-) |
| 13:04.28 | d-lo | I pulled out those geometric flaws and reoptimized the bot. (got the vertex count down by over 65% ;) ) |
| 13:04.47 | d-lo | but I am seeing some issues with those triangled shape thingies in the middle. |
| 13:04.51 | d-lo | What is this btw? |
| 13:07.02 | alex_joni | d-lo: it's a support thingie from a factory |
| 13:07.19 | alex_joni | side wall support |
| 13:07.52 | alex_joni | you asked earlier why I need a solid, I need to import it (an ACIS sat file) into a robot programming and simulation package.. |
| 13:08.44 | d-lo | Cool... well, those bolt holes are going to mess with your 'Solidity' |
| 13:09.03 | d-lo | if you need 100% prime cut Geometry, you will need to get rid of those flaws. |
| 13:09.27 | alex_joni | I see that g-iges does something, and converts 1 solid to NMG's before exporting to IGES (according to the log) |
| 13:10.34 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-71.sbndin.btas.verizon.net) |
| 13:11.04 | d-lo | that sounds about right. |
| 13:11.47 | alex_joni | but trying to import it in another app it fails on the iges format |
| 13:15.19 | alex_joni | I don't suppose there's an g-acis hidden somewhere? |
| 13:15.21 | alex_joni | or g-step ? |
| 13:18.42 | alex_joni | d-lo: http://eneas.juve.ro/~juve/brlcad/s32.g (this one should be without the holes) |
| 13:23.23 | d-lo | alex_joni: No, no g-acis nor g-step. Do you have another application that could be used as middleman for the conversion? |
| 13:25.11 | alex_joni | yeah, that's what I'm trying with IGES |
| 13:25.37 | alex_joni | (using Alibre which I have here, but i get "Object reference not set to an instance of an object." - whatever that might mean |
| 13:26.16 | d-lo | Okay, took a look at s32.g and it looks 100% valid. |
| 13:26.46 | d-lo | since its already a BoT, you might try g-dxf, then dxf->step or dxf->acis |
| 13:28.32 | alex_joni | is dxf 3D? |
| 13:28.45 | alex_joni | (the dxf that g-dxf writes I mean) |
| 13:29.07 | *** join/#brlcad Elrohir (n=kvirc@p5B14DAEE.dip.t-dialin.net) |
| 13:29.52 | d-lo | Yes, I believe it is. |
| 13:30.00 | alex_joni | it seems so |
| 13:35.40 | d-lo | Good deal. Everything working then? |
| 13:36.38 | *** join/#brlcad mafm_ (n=mafm@223.Red-83-49-86.dynamicIP.rima-tde.net) |
| 13:39.53 | alex_joni | d-lo: nope ;) |
| 13:40.32 | d-lo | whats brokeded? |
| 13:40.32 | alex_joni | d-lo: ACAD doesn't want to open it .. |
| 13:40.42 | d-lo | in what format |
| 13:40.43 | d-lo | ? |
| 13:40.57 | alex_joni | dxf |
| 13:41.14 | d-lo | what other formats does ACAD open? |
| 13:41.31 | alex_joni | it says "improper table entry name on line 20" |
| 13:41.53 | d-lo | whats line 20 look like then? Any illegal characters? |
| 13:42.17 | alex_joni | dxf, dwg, acis, 3ds |
| 13:42.25 | alex_joni | 20: r.stl/s.stl |
| 13:42.41 | alex_joni | I tried changing that to a number, then it moves on 4 lines, and complains about a missing table something |
| 13:42.42 | CIA-40 | BRL-CAD: 03starseeker * r33845 10/brlcad/trunk/src/libtclcad/tclcadAutoPath.c: Clearing the Tcl auto_path before tclcad_auto_path does its work gets archer up and running with local tcl/tk. Need to check its impact on builds using system tcl/tk |
| 13:43.43 | d-lo | you could try dxf->3ds in Blender :/ |
| 13:45.51 | alex_joni | heh, now that would lead a nice conversion path :D |
| 13:46.13 | alex_joni | dxf (mesh) -> obj -> stl -> g -> dxf -> 3ds -> dxf -> acis |
| 13:47.00 | alex_joni | d-lo: thanks for the effort though :) .. I'll probably go with a non-free conversion that creates a solid in ACAD from the mesh |
| 13:47.12 | starseeker | arrrgh, now archer crashes when trying to open a file |
| 13:47.15 | starseeker | http://pastebin.bzflag.bz/m28988fe4 |
| 13:49.57 | d-lo | alex_joni: You can't take the original dxf straight to g using dxf-g? |
| 13:50.57 | alex_joni | d-lo: nope, it seems it does only see 2-3 polylines |
| 13:51.14 | d-lo | Hrm, where is this dxf coming from? |
| 13:51.29 | alex_joni | some design studio :) |
| 13:51.49 | alex_joni | it would have been great if they used a SOLID in Acad when they drew it |
| 13:52.03 | alex_joni | then it would just have worked to export it from Acad, and I would be done with it |
| 13:53.05 | alex_joni | I did notice this when using g-dxf though: |
| 13:53.07 | alex_joni | db_walk_subtree() FAIL on '/r.stl/s.stl' |
| 13:53.07 | alex_joni | nmg_fix_normals: nmg_classify_s_vs_s() failed for shells x103ac48 and x1043a40 |
| 13:53.08 | alex_joni | <PROTECTED> |
| 13:53.56 | d-lo | strangeness. |
| 13:54.29 | alex_joni | d-lo: given the conversion path above I'm not _that_ surprised ;) |
| 13:54.54 | d-lo | this is yet another need for a obj-g converter :/ |
| 13:57.52 | alex_joni | d-lo: after messing with the dxf I was able to load it in acad |
| 13:58.08 | alex_joni | it's now full of 3DFACES not POLYMESH (like it was initially) |
| 13:58.12 | alex_joni | but it's still no solid |
| 13:59.27 | alex_joni | (messing meaning hacking the dxf header) |
| 14:01.24 | d-lo | Hrm, well, you might wanna take a look at the DXF standard... maybe there is something you can edit in the dxf file that will set the objects to solids... then you can forego the whole bloody conversion path all together! |
| 14:04.53 | archivist | reminds me of the days of writing dxf fixing scripts |
| 14:24.05 | starseeker | sigh, and RamDebugger still won't run in bwish - after all that it's a different issue |
| 14:24.17 | starseeker | closes down the xterms and heads in to work... |
| 14:31.12 | *** join/#brlcad PrezKennedy (i=Matthew@whitecalf.net) |
| 14:59.26 | yukonbob | starseeker: what are you trying to fix w/ RamDebugger? |
| 15:40.35 | brlcad | yay, finally back on-line |
| 15:40.52 | brlcad | arrived home to dead internet service last night, now restored after a two-hour phone call |
| 15:43.11 | brlcad | terminology snafu -- "Archer and BRLCAD are very similar" .. Archer is a component of BRL-CAD, I believe he meant MGED instead of "BRLCAD" .. MGED and Archer are very similar |
| 15:43.18 | brlcad | ( alex_joni ) |
| 15:45.19 | brlcad | those external tools that had trouble with dxf weren't ours -- and that 1.5GB limit is almost certainly a 32-bit app trying to read in the entire file in and running out of memory |
| 15:45.31 | brlcad | (at least, running out of 32-bit addressable memory) |
| 15:52.08 | brlcad | alex_joni: also, fwiw -- depending on your version of AutoCAD (and I assume you mean autocad and not a tool from GD specifically called 'ACAD') -- they're even inconsistent on their interpretation of their DXF spec |
| 15:53.23 | brlcad | dxf spec itemizes illegal characters, none of which we use iirc, but then there are some versions of autocad that won't read various other characters (e.g., the slashes) |
| 15:54.29 | brlcad | and there was a long time where there was ambiguity on whether all faces were 4-points or whether 3-point faces (i.e., triangles) were valid .. so if you had a triangle and only 4-point 3dfaces were allowed, you had to repeat the last point per the spec |
| 16:08.59 | starseeker | yukonbob: In theory it would be helpful debugging pure tcl applications like Archer and rtwizard |
| 16:09.06 | PrezKennedy | brlcad, they're going to get me apple certified |
| 16:09.09 | PrezKennedy | how ironic eh? |
| 16:20.34 | CIA-40 | BRL-CAD: 03d_rossberg * r33846 10/rt^3/trunk/ (11 files in 2 dirs): |
| 16:20.35 | CIA-40 | BRL-CAD: - virtual constructor and assignment operator for the Object |
| 16:20.35 | CIA-40 | BRL-CAD: - some more functionality for the Combination |
| 16:20.35 | CIA-40 | BRL-CAD: - clean-ups |
| 16:20.54 | brlcad | PrezKennedy: ironic? no .. inspiring ;) |
| 16:33.17 | PrezKennedy | it only take 40 screws to pop that hard drive outta your laptop... ;) |
| 17:19.39 | *** join/#brlcad samrose (n=samrose@dhcp-x198w-162.mobile.uci.edu) |
| 17:38.23 | alex_joni | brlcad: ok, thanks for the insight |
| 17:38.44 | alex_joni | (although I'm not a long time user of BRLCAD, I do know what mged and Archer are ;) |
| 17:39.14 | alex_joni | and for the dxf part, I think you're right, this was using AutoCAD 2009, but I wouldn't be too surprised if things were different with an older version |
| 18:13.58 | starseeker | grrrrr |
| 18:14.08 | starseeker | come on pipe, figure it out |
| 18:15.16 | starseeker | apparently pipe won't automatically "smooth" the difference between slopes where two coiling patterns join |
| 18:16.20 | starseeker | starts scribbling math to calculate the tgc required to at least make the transition solid, if not truly "smooth" |
| 18:25.04 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1212 10/wiki/Main_Page: 3rd party utils |
| 18:25.58 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1213 10/wiki/Third_Party_Utilities: hex |
| 18:28.31 | brlcad | PrezKennedy: I don't have that laptop, only macbook pro's ftw |
| 18:31.34 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1214 10/wiki/Hex: New page: Supported items: * Hex head metric bolts: M1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14, 16, 18, 20, 22, 24, 27, 30, 33, 36, 39, 42, 45, 48, 52, 56, 60, 64 * Basic metric threaded nuts, * ... |
| 18:32.20 | brlcad | ~nslookup 77.58.247.103 |
| 18:32.41 | brlcad | smells sushi edits ;) |
| 18:34.10 | d-lo | good thing you like sushi eh? |
| 18:35.59 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1215 10/wiki/Hex: |
| 18:42.11 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1216 10/wiki/Hex: |
| 18:53.21 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1217 10/wiki/Hex: /* Installation */ |
| 18:55.40 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1218 10/wiki/Hex: /* Installation */ |
| 18:57.35 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1219 10/wiki/Main_Page: /* Projects */ |
| 18:58.19 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1220 10/wiki/Main_Page: /* Projects */ |
| 18:59.13 | *** join/#brlcad _sushi_ (n=_sushi_@77-58-247-103.dclient.hispeed.ch) |
| 18:59.25 | _sushi_ | I made a wiki page about the threade bolts tool http://brlcad.org/wiki/Hex |
| 18:59.34 | d-lo | we saw :) |
| 18:59.39 | d-lo | looks neat! |
| 18:59.49 | CIA-40 | BRL-CAD: 03johnranderson * r33847 10/jbrlcad/trunk/ (18 files in 6 dirs): Fixed a few bugs |
| 19:00.10 | starseeker | ah, OK. |
| 19:00.39 | starseeker | shoot - doing quarter turn tricks over constrains the coil when individual sections need to be "blended" |
| 19:00.54 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1221 10/wiki/Hex: /* Usage */ |
| 19:00.59 | starseeker | OK, guess I caln live with that |
| 19:03.22 | CIA-40 | BRL-CAD: 03starseeker * r33848 10/brlcad/trunk/src/proc-db/spring.c: |
| 19:03.22 | CIA-40 | BRL-CAD: Working example of a coil with a squared end. Will have to give up quarter turn |
| 19:03.22 | CIA-40 | BRL-CAD: increment control in editing in order to allow pipe to smoothly join sections - |
| 19:03.22 | CIA-40 | BRL-CAD: that's an acceptable tradeoff. There are other tricks that can be done |
| 19:03.23 | CIA-40 | BRL-CAD: (including manual pipe editing, for that matter) if something really weird is |
| 19:03.25 | CIA-40 | BRL-CAD: needed. |
| 19:04.40 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1222 10/wiki/Hex: |
| 19:05.30 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1223 10/wiki/Hex: |
| 19:06.12 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1224 10/wiki/Hex: |
| 19:07.11 | CIA-40 | BRL-CAD: 0377.58.247.103 07http://brlcad.org * r1225 10/wiki/Hex: |
| 19:10.23 | CIA-40 | BRL-CAD: 03johnranderson * r33849 10/jbrlcad/trunk/src/org/brlcad/samples/BrlcadJava.java: Addef a usage message |
| 19:15.02 | CIA-40 | BRL-CAD: 03bob1961 * r33850 10/brlcad/trunk/src/ (libged/gqa.c mged/cmd.c mged/cmd.h mged/setup.c): |
| 19:15.02 | CIA-40 | BRL-CAD: Added a -Ap option to libged's ged_gqa function for creating psuedo-solid |
| 19:15.02 | CIA-40 | BRL-CAD: wireframe of overlaps. Added a cmd_ged_gqa wrapper to MGED for using the |
| 19:15.02 | CIA-40 | BRL-CAD: currently displayed objects if no objects were specified on the command line. |
| 19:22.39 | _sushi_ | I am getting this error when compiling 7.14.0 /bin/sh: /home/clock/brlcad-7.14.0/src/other/tcl/unix/tclsh: No such file or director |
| 19:22.42 | _sushi_ | y |
| 19:24.23 | brlcad | hrm, need more of the build log |
| 19:25.10 | _sushi_ | seems I am getting this problem http://www.mail-archive.com/brlcad-tracker@lists.sourceforge.net/msg00113.html |
| 19:25.17 | _sushi_ | I am trying autogen.sh |
| 19:25.29 | _sushi_ | autogen.sh passed |
| 19:25.48 | brlcad | ah, yeah |
| 19:25.52 | brlcad | that's probably the problem |
| 19:26.33 | brlcad | _sushi_: can you post a full build log somewhere? |
| 19:26.52 | _sushi_ | now the ./configure is printing endless cycle of configure: Would you like to continue with /usr as the install prefix? [yes/no] |
| 19:26.54 | brlcad | don't have the problem here to reproduce it so a fix can be made |
| 19:26.55 | _sushi_ | configure: Please answer 'yes' or 'no' |
| 19:26.56 | _sushi_ | and ignores my keyboard input |
| 19:26.58 | _sushi_ | configure: |
| 19:27.03 | _sushi_ | Even if I paste "yes\n" in one mouse-click |
| 19:27.12 | _sushi_ | Before it worked without the autogen.sh |
| 19:27.33 | brlcad | you're installing into /usr ? .. yikes |
| 19:27.45 | _sushi_ | Yes the previous brl-cad was installed into usr |
| 19:27.59 | brlcad | k, still surprising :) |
| 19:28.13 | _sushi_ | I am actually reinstalling because of rendering bug in 7.10.4 |
| 19:28.56 | _sushi_ | 7.10.4 renders my holder incorrectly http://ronja.twibright.com/3d/plazmatron_0.png |
| 19:29.20 | _sushi_ | The yellow thing is interrupted in the middle you can see the wand with the ball at the end ends in the air it shouldn't |
| 19:29.35 | _sushi_ | And the blue part is having a "horn" sticking out which apparently cannot have any mechanical function |
| 19:29.54 | _sushi_ | Well is there a way I can get through the endless loop of questions? |
| 19:30.18 | brlcad | if it stays stuck, you can set BRLCAD_ROOT_OVERRIDE and it'll skip the yes/no question, though I'm curious how it got stuck in a loop |
| 19:30.30 | brlcad | post your configure file somewhere |
| 19:30.47 | brlcad | it's a simple loop that just does a read until it gets a yes or a no |
| 19:31.24 | _sushi_ | Good I hacked the yes test in configure |
| 19:31.58 | _sushi_ | replaced read with assignment =yes :) |
| 19:32.25 | _sushi_ | I am still getting the tclsh error |
| 19:32.43 | brlcad | you would have to rebuild cleanly |
| 19:33.20 | _sushi_ | http://twibright.com/typescript |
| 19:33.27 | _sushi_ | what should I do to rebuild cleanly |
| 19:38.19 | _sushi_ | The configure AFTER runjning autogen.sh is at http://twibright.com/configure |
| 19:39.54 | _sushi_ | Now it compiles suspiciously long. The problem may be gone. |
| 19:40.21 | _sushi_ | Actually isn't. Still pops up: /bin/sh: /home/clock/brlcad-7.14.0/src/other/tcl/unix/tclsh: No such file or directory |
| 19:41.32 | _sushi_ | trying --enable-all... |
| 19:41.36 | starseeker | brlcad: here's the bomb log from the pipe raytracing crash: http://pastebin.bzflag.bz/m35029930 |
| 19:44.37 | _sushi_ | When I do rm config.cache* then ./configure --prefix=/usr --enable-all then it crashes on a different error: |
| 19:44.50 | _sushi_ | ./src/htmltcl.c: In function 'configureCmd': |
| 19:44.50 | _sushi_ | ./src/htmltcl.c:1351: error: 'HTML_DEFAULT_CSS' undeclared (first use in this function) |
| 19:49.47 | brlcad | http://twibright.com/configure is a 404 |
| 19:50.12 | _sushi_ | ah connection refused |
| 19:50.17 | brlcad | ah, server is down |
| 19:50.24 | _sushi_ | it's a friend's apache tomcat |
| 19:50.40 | _sushi_ | when it crashes you usually wait for a short time and it's up again |
| 19:51.24 | brlcad | that backtrace is useless, it's for the wrong thread |
| 19:51.37 | brlcad | unfortunately |
| 19:51.45 | _sushi_ | brlcad: you are not talking about my log now are you? |
| 19:51.58 | brlcad | yep |
| 19:52.05 | _sushi_ | the typescript is useless? |
| 19:52.34 | brlcad | well, yeah, it's fairly useless too other than it confirmed that it's the same problem as the one you found on the mailing list |
| 19:52.50 | brlcad | but no, I mean http://pastebin.bzflag.bz/m35029930 isn't useful unfortunately |
| 19:53.22 | brlcad | the automatic bomb report only shows the main thread -- if it's ray-tracing, there are potentially lots of threads |
| 19:53.31 | brlcad | could try -P1 to force single cpu |
| 19:53.59 | brlcad | or attaching with a debugger while it's running to get a proper trace for any one of the threads |
| 19:55.49 | _sushi_ | brlcad: http://twibright.com/configure is up again |
| 20:00.28 | brlcad | don't see anything wrong with your configure script unless you have a bused 'read' command |
| 20:00.39 | brlcad | run this one-liner: |
| 20:00.40 | brlcad | echo "type yes or no" ; while true ; do read a ; case "x$a" in x*[yY][eE][sS]*) a=yes ; break; ;; x*[nN][oO]*) a=no; break; ;; x*) echo "answer yes or no" ; ;; esac ; done ; echo "you answered $a" |
| 20:01.50 | brlcad | that's the exact same logic it's using |
| 20:03.55 | _sushi_ | makes this - no loop: |
| 20:03.55 | _sushi_ | type yes or no |
| 20:03.55 | _sushi_ | yes |
| 20:03.55 | _sushi_ | you answered yes |
| 20:04.30 | _sushi_ | This one fortunately isn't critical while I can hack it easily |
| 20:04.48 | _sushi_ | But how to get through the one with tcl or the one with CSS I have no idea |
| 20:06.14 | _sushi_ | brlcad: can I produce some better typescript that tells you more about the cause/ |
| 20:10.50 | _sushi_ | brlcad: what's interesting if I don't run autogen.sh (using the configure from the tarball), it doesn't make the infinite yes loop. |
| 20:11.17 | _sushi_ | if my read were screwed up, then it would happen even in this situation |
| 20:25.48 | _sushi_ | Even symlinking the system-wide tcl into the required directory doesn't help - compiles for a long time then makes: //home/clock/brlcad-7.14.0/src/other/incrTcl/.libs/libitcl.so: undefined reference to `TclVarHashCreateVar' |
| 20:26.38 | CIA-40 | BRL-CAD: 03starseeker * r33851 10/brlcad/trunk/src/proc-db/spring.c: Successfully squared off both ends of a coil. |
| 20:27.19 | starseeker | brlcad: regression tests succeed for mged on my Mac with latest checkout |
| 20:30.18 | *** join/#brlcad BigAToo (n=BigAToo@64.74.225.82) |
| 20:32.49 | starseeker | brlcad: let's try this: http://pastebin.bzflag.bz/m1967e293 |
| 20:33.01 | _sushi_ | brlcad: any idea wha tI could temporarily do to get around the problem and compile a BRL-CAD that doesn't produce this raytracing error? |
| 20:33.14 | starseeker | winces at the sight of boolweave |
| 20:37.57 | CIA-40 | BRL-CAD: 03starseeker * r33852 10/brlcad/trunk/src/proc-db/ (Makefile.am coil.c spring.c): Let's get this renamed - it's coil, not spring - spring is too general in terms of possible geometries. |
| 20:39.21 | _sushi_ | I tried ./configure without --prefix, also crashes on tcl |
| 21:12.18 | *** join/#brlcad Elrohir (n=kvirc@p5B14DAEE.dip.t-dialin.net) |
| 21:39.40 | brlcad | cool, http://www.kenradio.com/content/view/1756/30/ |
| 21:39.46 | brlcad | thanks ken |
| 21:41.07 | brlcad | _sushi_: I know it's not critical, it's just really odd -- if you still have the busted configure script, add a print statement on the x*) case to print the $bc_answer |
| 21:41.27 | brlcad | otherwise we'll just ignore it |
| 21:42.10 | brlcad | by your 'read' being messed up, the difference is in which interpreter is getting used, whether read is external or built-in (shell) command, etc |
| 21:43.07 | brlcad | starseeker: oh, nice -- badmagic's in boolweave are pretty rare |
| 21:43.53 | brlcad | _sushi_: give me a shell somewhere and I could try to produce a build interactively |
| 21:44.30 | brlcad | otherwise, you could try using the latest sources -- tcl version is different there |
| 21:46.12 | brlcad | starseeker: wince all you like, but that is actually one of the most valuable bits of logic in librt |
| 21:46.40 | brlcad | highly optimized and tuned to exact behavior, tolerance corrections, interference detection, and performance |
| 21:48.01 | brlcad | I've attempted to rewrite it to be "more clean" about a half-dozen times and every single time I've either killed performance or changed behavior |
| 21:48.20 | brlcad | it'd be interesting to come up with a test harness that exercises every line of code in there |
| 21:49.37 | louipc | There's some spam here: http://brlcad.org/d/node/31#comments |
| 21:49.59 | brlcad | starseeker: once it's more complete, coil can move to src/shapes |
| 21:50.10 | brlcad | (like at the point that it gets documentation) |
| 21:50.30 | brlcad | louipc: huh, thanks .. that's odd |
| 21:50.38 | brlcad | no comments are supposed to post without moderation |
| 21:53.31 | louipc | hmm, probably better to lose the comments altogether eh? |
| 21:57.22 | brlcad | louipc: possibly, though there are a few that are useful |
| 22:00.56 | brlcad | there's so few that actually got through, makes me think that they're actually real people behind them |
| 22:01.05 | brlcad | 4 in total |
| 22:06.18 | brlcad | woot, http://dreamcss.blogspot.com/2009/02/10-free-and-useful-open-source.html |
| 22:07.59 | louipc | aww they could have put a more interesting screenshot |
| 22:09.16 | brlcad | yeah :) |
| 22:09.35 | brlcad | website needs another face lift |
| 22:29.49 | brlcad | 'read' is failing |
| 22:58.07 | *** join/#brlcad samrose (n=samrose@ip67-95-244-194.z244-95-67.customer.algx.net) |
| 23:04.13 | CIA-40 | BRL-CAD: 03brlcad * r33853 10/brlcad/trunk/m4/prefix.m4: if stdout is closed, read will fail on eof and loop infinitely. so we add some checks for failed read and basic sanity to only ask yes/no 10 times before giving up. |
| 23:08.26 | brlcad | that should take care of the infinite loop, now on to the next one later tonight |
| 23:08.31 | brlcad | (as I wait for this build) |
| 23:29.08 | starseeker | brlcad: not wincing at the boolweave code, but at the bug poping up there - there be dragons for debugging and code sensitivity |
| 23:58.35 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-71.sbndin.btas.verizon.net) |
