| 00:53.47 | CIA-85 | BRL-CAD: 03starseeker * r38061 10/brlcad/trunk/src/libgcv/obj/obj_grammar.y: Add a little more output to the testing of obj_grammar.y. |
| 01:13.28 | starseeker | blinks - subversion is now "apache subversion?" |
| 01:13.37 | Ralith | O.o |
| 01:27.36 | starseeker | brlcad: do the newest subversion releases get us any closer to the idea of being able to merge different repository histories? "merge tracking" sounds promising but I can't really say for sure... |
| 01:27.55 | starseeker | (would sure simplify our sync instructions, if it does work...) |
| 01:38.46 | brlcad | that's fantastic news (regarding subversion incubation in the ASF) |
| 01:39.55 | brlcad | happened earlier in the winter, no? |
| 01:40.06 | starseeker | I believe so |
| 01:40.09 | starseeker | Feb? |
| 01:40.15 | brlcad | thought it was last year |
| 01:40.35 | starseeker | 2010-02-17 Subversion becomes Apache Subversion |
| 01:40.54 | brlcad | http://www.apache.org/foundation/press/pr_2009_11_04.html |
| 01:40.54 | starseeker | oh, incubator nov last year |
| 01:40.59 | starseeker | nods |
| 01:41.05 | starseeker | sorry, didn't spot that |
| 01:41.59 | brlcad | merge tracking would probably reduce a step or two when syncing branches, but it really just sort of auto-commits from defined points |
| 01:42.26 | starseeker | oh, OK - so it doesn't solve the hard problem |
| 01:42.50 | starseeker | (in the sense of distributed VCS merging) |
| 01:43.36 | brlcad | "the hard problem" being? |
| 01:43.55 | brlcad | it solves the problem of having to look up the last sync revision |
| 01:44.01 | starseeker | preserving the revision history of the branch in the resulting merged trunk? |
| 01:44.15 | brlcad | instead of manually managing that revision by looking at the log and comments |
| 01:44.53 | brlcad | ah, you mean if you had 20 commits to a branch, having those 20 commits reflected on trunk when merged back? |
| 01:44.59 | starseeker | right |
| 01:45.14 | brlcad | not sure if merge tracking would address that, maybe bidirectional merge tracking |
| 01:45.46 | brlcad | that's a fairly new feature, haven't read up on all the details |
| 01:46.04 | starseeker | nods |
| 01:49.18 | brlcad | the core benefiting feature for the geometry service would be offline commits |
| 01:50.24 | brlcad | which is effecitvely *repository* branching (cloning) with pushed merges |
| 01:50.42 | starseeker | hmm. http://subversion.wandisco.com/component/content/article/1/44.html |
| 01:51.00 | brlcad | but the trick being how to do that without carrying around the entire repository |
| 01:51.59 | brlcad | yeah, they've been talking about doing offline commits for a couple years |
| 01:52.22 | brlcad | just a matter of resources to do the impl and validation |
| 01:54.16 | brlcad | hadn't heard about wandisco's announcement though |
| 01:58.39 | brlcad | fogel, sussman, and fitzpatrick have talked extensively about those plans over the past few years (svn devs, google folk) |
| 01:58.55 | starseeker | cool |
| 01:59.05 | brlcad | this is a pretty nice followup to torvalds talk from fogel: http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=898498 |
| 02:03.11 | starseeker | interesting read, and sounds encouraging |
| 02:03.35 | starseeker | be nice if they got the offline commit thing working in time for us to make use of it :-) |
| 02:04.16 | brlcad | "The working copy should really be a repository, even if |
| 02:04.16 | brlcad | it's not always going to store all the history available on the server |
| 02:04.16 | brlcad | side (with some projects, you really can't, it's too big). |
| 02:04.48 | brlcad | " -- that comment summarizes a lot of good understanding of the issues |
| 02:04.49 | starseeker | nods - that's kinda us in a nutshell |
| 02:06.45 | brlcad | that and "most users don't know or care whether a |
| 02:06.45 | brlcad | system is centralized or decentralized -- their ideal system is one |
| 02:06.47 | brlcad | they don't notice. " |
| 02:06.58 | starseeker | grins |
| 02:07.02 | brlcad | sound familiar? :) |
| 02:07.07 | starseeker | indeed :-) |
| 02:22.57 | brlcad | I have access to the new tkhtml repository now |
| 02:23.15 | starseeker | sweet! |
| 02:23.54 | starseeker | is that the sourceforge one or the fossil one? |
| 02:24.10 | brlcad | the fossil one |
| 02:24.19 | starseeker | ah, neat! |
| 02:24.20 | brlcad | already had/have access to the sf one |
| 02:26.14 | starseeker | how does fossil look to work with? |
| 02:27.37 | brlcad | don't know, but I'm willing to give it a go |
| 02:28.36 | brlcad | want to make sure we can get access to the web files and tracker data before getting too deep, but can at least clone the entire history now, and could probably manually dump to svn if we had to now |
| 02:29.07 | starseeker | excellent |
| 04:05.24 | brlcad | starseeker: http://fr.wikibooks.org/wiki/Initiation_?_BRL-CAD/Prototypage_rapide_avec_Archer |
| 04:06.25 | brlcad | Archer being documented in real-time... too bad a lot of that is in flux already :) |
| 04:11.55 | starseeker | hah, cool |
| 04:12.38 | starseeker | yeah, that'll be a lot of work if they want to keep up over the next couple months |
| 04:12.54 | starseeker | fear the Bob |
| 04:25.51 | brlcad | thinks interlay should go away |
| 04:29.33 | brlcad | the complexity it adds (however minor) doesn't seem to provide enough value to be warranted as a GUI option |
| 04:30.30 | brlcad | more familiar would probably be a "Send to back" and "Bring to front" context menu option |
| 04:31.56 | brlcad | that way, the embedded framebuffer just acts like a proper display layer |
| 04:38.23 | brlcad | starseeker: I sure hope you realized that "char *argv" was a bug .. ;) |
| 04:44.30 | brlcad | int main(int ac, char *av[]); is main's signature per the standard |
| 04:45.15 | brlcad | unless you're talking about C++ in which case the two parameters are optional (but it still must return int) |
| 05:39.54 | CIA-85 | BRL-CAD: 03brlcad * r38062 10/brlcad/trunk/src/conv/obj-g.c: print the zero-face message regardless of the object name. |
| 05:42.25 | starseeker | brlcad: yeah, I knew it was probably a bug. :-P |
| 05:42.57 | CIA-85 | BRL-CAD: 03brlcad * r38063 10/brlcad/trunk/src/conv/obj-g.c: fix header, was started in 2009 |
| 05:43.09 | starseeker | I just wasn't sure if someone had horribly hacked it somehow to work that way, but no reference to av at all cleared it |
| 05:44.33 | brlcad | it'd still work just because it's a pointer (until someone tried to do pointer arith on an argv with more than one element) |
| 05:45.18 | brlcad | but it was outright wrong per the standard regardless of any intent |
| 05:45.25 | starseeker | nods. Shows the use of clang's colorful error messages - I'm sure gcc must have been complaining about that and I never saw it |
| 05:46.27 | brlcad | there's even a subtle difference between char **argv and char *argv[], though as a parameter the difference is mostly academic pedantry |
| 05:47.32 | starseeker | supposes we could ditch interlay... it's already in Archer though, I'd suggest leaving it until it's established as the new MGED and then ditching it, unless you think it's better to nuke it now |
| 05:48.05 | brlcad | I was even referring to MGED |
| 05:48.13 | starseeker | oh, OK |
| 05:48.17 | brlcad | not this release, but next minor |
| 05:48.31 | brlcad | soon as the dang mac bug gets fixed |
| 05:48.42 | brlcad | is going to try to squash it this week |
| 05:48.55 | starseeker | there is one situation I know of where it might help - a superdense wireframe within which one is trying to do an oed edit |
| 05:50.42 | starseeker | on the whole though, I'd be delighted to get rid of it - it would undoubtedly simplify the Tk framebuffer embedding |
| 05:51.57 | starseeker | reflects that the "correct" fix for the overly dense wireframes is probably level-of-detail drawing anyway.... |
| 05:52.25 | brlcad | the window could still have the interlay ability, it's a matter of how many options are exposed through the GUI |
| 05:52.58 | brlcad | most open source suffers from massive option overload because it's so simple to "add one more checkbox" or button or whatever |
| 05:53.07 | starseeker | urm. How would we use the capability if it's not GUI exposed? |
| 05:53.10 | starseeker | true... |
| 05:53.14 | brlcad | but the resulting complexity seriously can hurt usability |
| 05:53.39 | brlcad | they need to pull their weight in terms of merit (which is understandably hard to gauge and subjective without stats) |
| 05:54.24 | brlcad | anything via the GUI should be exposed via a command, so if the framebuffer has even overlay/underlay options, there should be some command-line mechanism for setting it |
| 05:55.20 | brlcad | either accessing some framebuffer object/command or accessing some graphics window object/command or via basic system preferences (e.g. rset), etc |
| 05:55.49 | starseeker | thinks it might make sense as an rset variable |
| 05:56.17 | starseeker | what was the original motivator for overlay mode? Just "cause we can?" |
| 05:57.33 | starseeker | er interlay rather |
| 05:57.43 | starseeker | should probably get some sleep here soon... |
| 06:10.56 | brlcad | no, there's a case where you have really complex geometry being rendered, want to raytrace and still see part of the wireframe |
| 06:11.45 | brlcad | more a failing of super-stupid wireframe drawing, though |
| 06:28.59 | CIA-85 | BRL-CAD: 03brlcad * r38064 10/brlcad/trunk/src/conv/obj-g.c: fix a memory failure. the bot ip is released during wdb_export, so be sure to null it out so we don't try to free it again (or access it). |
| 06:29.37 | CIA-85 | BRL-CAD: 03brlcad * r38065 10/brlcad/trunk/src/librt/primitives/bot/bot.c: be more careful about releasing memory and add even more sanity nulls after memory is released. |
| 06:57.19 | *** join/#brlcad talcite (~matthew@69-165-143-29.dsl.teksavvy.com) |
| 07:14.58 | CIA-85 | BRL-CAD: 03brlcad * r38066 10/brlcad/trunk/src/librt/primitives/bot/bot.c: more sanity checking for faces with missing vertices. fixes plot crash on such bots. |
| 08:44.45 | *** join/#brlcad dnk-88 (~dnk-88@217.21.40.13) |
| 09:28.51 | *** join/#brlcad mafm (~mafm@193.153.52.54) |
| 10:34.16 | *** join/#brlcad mafm (~mafm@193.153.52.54) |
| 10:48.21 | *** join/#brlcad parigaudi (~quassel@pd95b7f5e.dip0.t-ipconnect.de) |
| 10:52.23 | CIA-85 | BRL-CAD: 03brlcad * r38067 10/brlcad/trunk/src/conv/obj-g.c: (log message trimmed) |
| 10:52.23 | CIA-85 | BRL-CAD: good grief, messy wild assumptions going on in here regarding faces. fix a |
| 10:52.23 | CIA-85 | BRL-CAD: processing bug that was causing a crash due to the parsed face and vertex data |
| 10:52.23 | CIA-85 | BRL-CAD: (which was being stashed in a bot internal) getting released during export. |
| 10:52.23 | CIA-85 | BRL-CAD: instead, stash the vertex/face data into its own container that is memory |
| 10:52.24 | CIA-85 | BRL-CAD: managed separately. since face ids accummulate as the file is procesed, keep |
| 10:52.25 | CIA-85 | BRL-CAD: them all even as we write out groups. this in-core approach blindly and |
| 10:53.49 | brlcad | that should be an improvement, but there seems to be something wonefully busted with tessellation that needs to be regression tested |
| 10:59.10 | CIA-85 | BRL-CAD: 03brlcad * r38068 10/brlcad/trunk/NEWS: |
| 10:59.11 | CIA-85 | BRL-CAD: fixed a bug in obj-g where it was crashing if there were multiple objects in the |
| 10:59.11 | CIA-85 | BRL-CAD: obj file. the vertex/face data was getting released when a bot is written out, |
| 10:59.11 | CIA-85 | BRL-CAD: leaving much room for platform/environment-spefic crashness. improved memory |
| 10:59.11 | CIA-85 | BRL-CAD: management with a new data container though still more work is needed. |
| 11:00.09 | *** join/#brlcad alex_joni (~alex_joni@emc/board-of-directors/alexjoni) |
| 11:13.43 | CIA-85 | BRL-CAD: 03brlcad * r38069 10/brlcad/trunk/NEWS: janine gettier continues to exhuberate awesomeness with another 32 manual pages converted into Docbook format and added to the build by cliff. |
| 11:16.02 | CIA-85 | BRL-CAD: 03brlcad * r38070 10/brlcad/trunk/sh/orbit.sh: pretty sure test/plain is not a valid mime type erik. using text/x-sh instead. |
| 11:26.58 | CIA-85 | BRL-CAD: 03brlcad * r38071 10/brlcad/trunk/src/liboptical/photonmap.c: %% is a literal % character (e.g., 'Irradiance Cache Progress: 30% Approximate...') .. man 3 printf ftw. |
| 11:31.44 | brlcad | goes to bed, taking in last day of vacation |
| 11:54.41 | *** join/#brlcad poolio (~poolio@BZ.BZFLAG.BZ) |
| 12:24.55 | ``Erik | heh, woops, typo'd one of them annoying propsets |
| 12:25.05 | ``Erik | there's a lot more wrong with that file, though |
| 12:26.18 | ``Erik | ('exhuberate'? mebbe exuberant? O.o ) |
| 13:09.54 | *** join/#brlcad Phurl_ (~mdupont@cl-1773.dus-01.de.sixxs.net) |
| 13:33.38 | *** join/#brlcad ``Erik (~erik@c-69-140-109-104.hsd1.md.comcast.net) |
| 13:39.39 | *** join/#brlcad Phurl_ (~mdupont@ip-81-210-228-126.unitymediagroup.de) |
| 13:52.39 | CIA-85 | BRL-CAD: 03bob1961 * r38072 10/brlcad/trunk/src/tclscripts/lib/RtControl.tcl: Added the following options: -fb_enabled, -fb_enabled_callback. |
| 13:55.35 | CIA-85 | BRL-CAD: 03bob1961 * r38073 10/brlcad/trunk/src/tclscripts/archer/Archer.tcl: Added code to toggle the image of the framebuffer enable button when the state changes. |
| 16:22.16 | *** join/#brlcad parigaudi (~quassel@pd95b7f5e.dip0.t-ipconnect.de) |
| 17:49.24 | CIA-85 | BRL-CAD: 03bob1961 * r38074 10/brlcad/trunk/include/tclcad.h: Added go_rt_end_callback member to struct ged_obj. |
| 17:57.37 | CIA-85 | BRL-CAD: 03bob1961 * r38075 10/brlcad/trunk/include/ged.h: Added an abort parameter to the gd_rtCmdNotify member of struct ged_drawable. |
| 18:00.41 | CIA-85 | BRL-CAD: 03bob1961 * r38076 10/brlcad/trunk/src/libged/ (rt.c rtcheck.c): Added a parameter to the gd_rtCmdNotify call. |
| 18:02.44 | CIA-85 | BRL-CAD: 03bob1961 * r38077 10/brlcad/trunk/src/libtclcad/ged_obj.c: Added code to get an indication of when a raytrace finishes and possibly pass that off to Tcl land. |
| 18:03.30 | CIA-85 | BRL-CAD: 03bob1961 * r38078 10/brlcad/trunk/src/tclscripts/lib/Ged.tcl: Added the rt_end_callback method to cawidgets::Ged. |
| 18:05.04 | CIA-85 | BRL-CAD: 03bob1961 * r38079 10/brlcad/trunk/src/tclscripts/archer/Archer.tcl: Consolidated the raytrace and abort toolbar buttons. |
| 18:21.12 | CIA-85 | BRL-CAD: 03bob1961 * r38080 10/brlcad/trunk/include/raytrace.h: Added a missing parameter to the declaration of rt_nmg_mc_realize_cube. |
| 19:26.10 | CIA-85 | BRL-CAD: 03erikgreenwald * r38081 10/brlcad/trunk/src/librt/primitives/nmg/nmg_tri_mc.c: fix lame macro booboo |
| 21:23.00 | CIA-85 | BRL-CAD: 03erikgreenwald * r38082 10/brlcad/trunk/src/librt/primitives/nmg/nmg_tri_mc.c: shoot rays across instead of using midpoints. |
| 22:49.32 | *** join/#brlcad dnk-88 (~dnk-88@217.21.40.13) |
| 23:04.43 | *** join/#brlcad roberthl (~robert@2001:ba8:1f1:f03d::2) |
| 23:04.43 | *** join/#brlcad roberthl (~robert@silentflame/member/roberthl) |
| 23:13.02 | *** join/#brlcad Aeamus (~Enigma@unaffiliated/r0b0t1) |
| 23:45.44 | ``Erik | huh, looks like I forgot to commit a file ... quite a while ago O.o and no one mentioned it heh |
| 23:50.50 | CIA-85 | BRL-CAD: 03erikgreenwald * r38083 10/brlcad/trunk/include/raytrace.h: this func changed signature. |