| 00:24.54 | CIA-52 | BRL-CAD: 03Sean 07http://brlcad.org * r2609 10/wiki/Google_Summer_of_Code: /* [[Google_Summer_of_Code/2009|GSoC 2011]] */ |
| 00:25.09 | CIA-52 | BRL-CAD: 03Sean 07http://brlcad.org * r2610 10/wiki/Google_Summer_of_Code: accepted |
| 00:29.33 | brlcad | in theory, you should be able to get an accelerated and non-accelerated context from both Qt and OGRE -- it's really just what toolkit does someone want to use and which one will integrate best (be easiest to maintain) |
| 00:30.21 | *** join/#brlcad Ralith (~ralith@d142-058-172-039.wireless.sfu.ca) | |
| 00:30.42 | brlcad | e.g., qt with an ogl context could be accelerated or it could be a pure raster context; ogre might be harder to get unaccelerated, but there's always software render ogl :) |
| 00:32.41 | starseeker | ah, ok - I had assumed we would want to go with something like OGRE for accelerated in order to support more advanced work to come |
| 00:37.59 | CIA-52 | BRL-CAD: 03starseeker * r43898 10/geomcore/branches/fossil/: Add a branch for some experiments |
| 00:38.43 | brlcad | ogl for the dm definitely makes the most sense |
| 00:38.55 | brlcad | something that has scenegraph management |
| 00:38.58 | starseeker | plain opengl? |
| 00:39.18 | brlcad | plain ogl with custom scenegraph management is a bitch |
| 00:39.33 | brlcad | that's where ogre becomes a big win over qt for the dm side |
| 00:39.41 | brlcad | for the fb side, either would be a win |
| 00:39.48 | brlcad | fossil scm? |
| 00:39.48 | starseeker | oh, you mean OGRE for the dm? |
| 00:39.54 | starseeker | yes |
| 00:39.59 | brlcad | heh, neat |
| 00:40.06 | starseeker | BSD licensed as of mid last year |
| 00:40.16 | brlcad | nods |
| 00:40.22 | starseeker | I didn't notice at the time - would have hopped on it quicker |
| 00:40.35 | brlcad | you know it's almost entirely built on sqlite? |
| 00:40.37 | starseeker | yep |
| 00:41.30 | starseeker | wasn't so worried about its backend storage mechanism as long as it doesn't cause us some kind of significant trouble |
| 00:42.06 | starseeker | more annoying is his fondness for global variables |
| 00:42.12 | starseeker | that's what I was grousing about earlier |
| 00:47.40 | CIA-52 | BRL-CAD: 03starseeker * r43899 10/geomcore/branches/fossil/ (17 files in 6 dirs): For now, clear other contents - keep the build logic simple during early phases. |
| 00:53.22 | CIA-52 | BRL-CAD: 03starseeker * r43900 10/geomcore/branches/fossil/ (48 files in 8 dirs): |
| 00:53.22 | CIA-52 | BRL-CAD: Add what work has been done so far - the first goal is to get db.c and |
| 00:53.22 | CIA-52 | BRL-CAD: everything it needs compiling without warnings about implicit definitions. No |
| 00:53.22 | CIA-52 | BRL-CAD: global 'g' structure with settings either - pass it properly, even if it does |
| 00:53.22 | CIA-52 | BRL-CAD: mean updating all the functions to do it. |
| 01:08.11 | CIA-52 | BRL-CAD: 03starseeker * r43901 10/geomcore/branches/fossil/ (CMakeLists.txt src/CMakeLists.txt): comment out dirs not active yet |
| 01:15.41 | CIA-52 | BRL-CAD: 03starseeker * r43902 10/geomcore/branches/fossil/ (3 files in 2 dirs): add a content.h header to manifest and diff, fix accordingly. |
| 01:56.57 | brlcad | ah, yeah |
| 01:57.20 | brlcad | it's endemic of fast but lazy coding |
| 01:57.54 | brlcad | great for getting things done, but crazy maintainability for new developers |
| 01:58.32 | starseeker | my favorite bit is he generates the headers based on what the .c files need |
| 01:59.40 | starseeker | brlcad: is sqlite a concern? |
| 02:00.02 | brlcad | nope |
| 02:00.25 | starseeker | breaths a sigh of relief |
| 02:00.31 | brlcad | implementation detail that might affect scalability, but not a problem until it's a problem |
| 02:01.16 | brlcad | have trouble seeing it scale to multiGB databases, but maybe |
| 02:01.33 | starseeker | yeah, huge meshes could be a problem |
| 02:01.50 | brlcad | single models could be a problem |
| 02:02.06 | brlcad | not to even consider hundreds or thousands |
| 02:02.45 | starseeker | well, ``Erik has some wicked test cases we can toss in |
| 02:03.06 | starseeker | brlcad: I was kind of thinking one sqlite file per .g namespace... |
| 02:04.23 | starseeker | under the hood, of course |
| 02:04.31 | brlcad | single db by itself is probably fine, thinking db with full revision history, hundreds of edits like you'd have if you start from scratch - that's where it'd get big |
| 02:04.39 | brlcad | yeah, possibly |
| 02:04.48 | brlcad | that'd definitely be better than one honking db |
| 02:05.14 | brlcad | but then merging/branching and linking across namespaces becomes really nasty, no? |
| 02:05.23 | starseeker | ah, yeah - was thinking a few rotates and translates at the toplevel plus lots of xpushing for edit testing |
| 02:05.26 | brlcad | with svn it was just across files using path convention |
| 02:05.36 | brlcad | all still within the revision system |
| 02:06.09 | brlcad | either way, interesting test |
| 02:06.09 | starseeker | brlcad: not sure how nasty it would be - I've been thinking about it, but nothing concrete to toss on the whiteboard yet |
| 02:06.48 | brlcad | get stuck with svn libs? |
| 02:07.04 | brlcad | or looking for better performance? |
| 02:07.14 | brlcad | or just having fun? :) |
| 02:07.37 | starseeker | heh - combination of the latter two |
| 02:08.42 | starseeker | I was hoping a "speak .g language" approach to the revision control would let us avoid the region breakout and still have reasonable performance, but looking at that level of subversion/apr just got downright scary |
| 02:09.22 | starseeker | fossil appealed because it's smaller, distributed and for a bonus has that web server stuff integrated |
| 02:09.50 | starseeker | thought it might be more practical to get it commiting individual geometry objects as binary blobs and checking out into .g files |
| 02:10.45 | starseeker | however screwed up its "yay global variables" approach is, once straightened out it should build portably and without fun like the apr libs |
| 02:12.49 | brlcad | be sure to consider the big picture management advantages/disadvantages too then, user management isn't nearly as well integrated, generic properties, proven stability across massive dbs, repository mirroring, efficient binary management, .. ;) |
| 02:14.06 | starseeker | actually it does seem to have some user management abilities, although I haven't looked hard at that |
| 02:14.30 | starseeker | I'm assuming it's got to be fairly good at mirroring, being a dvcs... |
| 02:14.31 | brlcad | if performance were the only consideration, that new libgit toy would be a viable candidate |
| 02:14.38 | brlcad | it has user management |
| 02:14.41 | starseeker | <snort> I took a look at that |
| 02:14.45 | brlcad | it's not just nearly as well integrated |
| 02:14.46 | starseeker | doesn't have the feature set yet |
| 02:14.56 | brlcad | svn's is pretty extensive |
| 02:15.19 | brlcad | there's a whole lib dedicated to it that can be coupled to the various remote connection mechanisms |
| 02:15.55 | starseeker | brlcad: is user management a big deal at the revision control level? I thought user management stuff was gonna live higher up the food chain, but maybe I was wrong |
| 02:16.13 | brlcad | initial stab was a complete punt, just let svn do what it does |
| 02:16.32 | brlcad | anything we do higher up is just going to suck |
| 02:16.38 | starseeker | ah |
| 02:16.52 | brlcad | that's trickty to get right, even stupid username+password management |
| 02:17.14 | starseeker | fossil does have at least some capability for username+password |
| 02:17.20 | brlcad | it has to :) |
| 02:17.39 | brlcad | it's an unintegrated home-grown solution iirc |
| 02:18.12 | starseeker | yeah, it's self-containment is one of its selling points |
| 02:18.23 | brlcad | yep |
| 02:18.30 | brlcad | it's a plus and a negative tradeoff |
| 02:18.42 | brlcad | super limited, but super simple |
| 02:19.42 | brlcad | to me spells "super inadequate" for long-term without hacking a completely separate user management layer on top, but that's not a concern for a while |
| 02:20.00 | brlcad | it's worth testing regardless |
| 02:20.33 | brlcad | anything we implement shouldn't be so tightly integrated that we cannot switch out to a different versioning engine without a couple weeks effort |
| 02:21.48 | starseeker | my plan was to get it refactored/reworked to the point where I could directly pass it librt's db byte arrays as objects, and check them out into a .g file instead of files |
| 02:22.05 | starseeker | then see what performance was like |
| 02:32.48 | CIA-52 | BRL-CAD: 03starseeker * r43903 10/geomcore/branches/fossil/src/libgeomvcs/ (CMakeLists.txt checkin.c): Still problems with this file, but can't continue tonight - check in what I've got. |
| 02:33.50 | brlcad | starseeker: the scrolledwidget ttk fix .. is that from upstream? |
| 02:34.07 | brlcad | worth pushing to upstream? |
| 03:50.22 | starseeker | brlcad: if you can call it a "fix" - just going for uniform ttk usage when possible |
| 05:28.52 | *** join/#brlcad Ralith (~ralith@S010600221561996a.vc.shawcable.net) | |
| 06:53.06 | *** join/#brlcad Stattrav (~Stattrav@111.93.134.142) | |
| 06:53.06 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) | |
| 07:21.59 | *** join/#brlcad Ralith (~ralith@S010600221561996a.vc.shawcable.net) | |
| 08:14.11 | *** join/#brlcad Stattrav (~Stattrav@111.93.134.142) | |
| 08:14.11 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) | |
| 08:38.49 | *** join/#brlcad d_rossberg (~rossberg@BZ.BZFLAG.BZ) | |
| 10:22.03 | *** join/#brlcad merzo (~merzo@193.254.217.44) | |
| 10:37.15 | dloman | Mernin all. |
| 10:50.16 | brlcad | so approximately 92 seconds of that 30min animation took up about 5GB of disk in image frames... think I'm going to need a bigger disk |
| 10:51.16 | brlcad | or I'm going to need to make a smaller animation (was aiming for 720p) |
| 10:52.23 | dloman | yikes. |
| 10:52.31 | dloman | that's totally raw image data i take it? |
| 10:53.12 | brlcad | basically, it's a ppm stream |
| 10:53.20 | dloman | at 5GB per 92 seconds, you're looking at about 100GB for the 30 min stretch. |
| 10:53.42 | dloman | I've got a 2TB external you can borrow if ya need :) |
| 10:54.18 | brlcad | yeah, which by itself isn't too shabby, except it's not exactly something I can upload to vimeo :) |
| 10:54.40 | brlcad | i'll figure something out |
| 10:56.22 | dloman | I have Adobe Premere elements and I *think* it can encode a ppm stream to something like mp4, avi, wmv, etc. |
| 10:56.26 | dloman | if ya need. |
| 10:57.33 | brlcad | that part I'm actually not worried about at all, plenty of ways to encode from the stream |
| 10:57.51 | brlcad | more the processing size and size of end result video |
| 10:58.33 | dloman | well, its not ideal, but you can always hack it into parts if its too big for upload. |
| 10:58.40 | brlcad | this should work.. just need to make some room :) |
| 10:59.14 | dloman | hehehe. |
| 10:59.45 | dloman | on that note: I am seriously considering picking up a pair of 2TB internals... time for some upgrading of the data server! |
| 11:00.28 | dloman | <Tangent> brlcad: Have you neard of a group called Improv Everywhere? |
| 11:06.49 | *** join/#brlcad sofichinu (~sofichinu@115.69.147.65) | |
| 11:15.10 | dloman | http://www.youtube.com/watch?v=lYJ9zOyzI4w&feature=player_embedded |
| 11:15.18 | dloman | That's the kinda stuff IE does :) |
| 11:15.25 | dloman | basically, hilarious stuff :) |
| 12:04.09 | CIA-52 | BRL-CAD: 03Tbrowder 07http://brlcad.org * r2611 10/wiki/Google_Summer_of_Code/Project_Ideas: |
| 12:15.06 | CIA-52 | BRL-CAD: 03Rossberg 07http://brlcad.org * r2612 10/wiki/Google_Summer_of_Code: probable a copy-and-paste error |
| 12:21.21 | CIA-52 | BRL-CAD: 03davidloman * r43904 10/geomcore/trunk/ (include/Config.h src/utility/Config.cxx): Added some zero length string checks for the UInt and UShort helper functions in Config class to keep from attempting to parse an empty string. Added a hasConfigValue() for performing simple mapping checks. |
| 12:27.14 | CIA-52 | BRL-CAD: 03davidloman * r43905 10/geomcore/trunk/src/utility/Config.cxx: Made getConfigValue() perform a check to see if the internal config map actually has the key prior to attempting to return the value. |
| 12:29.08 | CIA-52 | BRL-CAD: 03davidloman * r43906 10/geomcore/trunk/src/GS/geomserv.cxx: Simplify startup of geomserv by making Config do all the data conversions for the 'port' parameter. |
| 12:29.15 | starseeker | brlcad: I'll try and sync to stable today |
| 12:41.10 | CIA-52 | BRL-CAD: 03davidloman * r43907 10/geomcore/trunk/ (3 files in 3 dirs): Reordered GemetryService args to be more intuitive. Removed GeometryService's need to store a localnodename as it was redundant (it was being stored as the thread name anyways) |
| 12:43.54 | CIA-52 | BRL-CAD: 03davidloman * r43908 10/geomcore/trunk/src/GS/geomserv.cxx: Geomserv now correctly parses config file for listenAddy and listenPort. If not present, use defaults. |
| 13:12.40 | starseeker | here's a quick stab at a poster - I don't expect it'll pass muster but maybe it'll prod somebody: |
| 13:12.44 | starseeker | http://bzflag.bz/~starseeker/poster_2011_1st_draft.png |
| 13:13.16 | dloman | Im instantly grooving on the TRON font =D |
| 13:13.41 | ``Erik | yeah, int he lower right, that's cool |
| 13:14.04 | ``Erik | but the drop shadows? ortho view of a ginormous screenshot? (not saying I could do better, just being the peanut gallery) |
| 13:14.16 | starseeker | that's the GSoC logo - I'm not sure how to do that font |
| 13:14.32 | starseeker | ``Erik: that's why it's to prod, not proposed as a final |
| 13:15.10 | ``Erik | hm, matt wanted to do german today, you should call him up and go, bring up the poster, he or nikki might be interested in chattering about it, if not helping :) |
| 13:15.23 | starseeker | ``Erik: heh |
| 13:15.34 | starseeker | unfortunately, I doubt I'll be up there in time |
| 13:15.39 | ``Erik | we're programmers, not artists :D |
| 13:15.41 | ``Erik | ah, bummer |
| 13:15.44 | starseeker | amen |
| 13:15.59 | ``Erik | (or, rather, our art is code and math, not visual) |
| 13:16.06 | starseeker | the drop shadow effect was a stand in until someone figures out how to do that google font thing |
| 13:16.36 | ``Erik | looks like the tron font done in gimp with the glow and outline script-fu smacked on it? |
| 13:16.48 | starseeker | shrugs - maybe |
| 13:16.56 | ``Erik | script-fu is neat, it used to all be scheme, d'no if they've "fixed" that |
| 13:16.59 | starseeker | doesn't have strong script-fu |
| 13:17.21 | starseeker | ``Erik: if you want to hack around with it I can send you the xcf |
| 13:17.38 | ``Erik | nah, getting ready to shower and head down to union memorial |
| 13:17.52 | starseeker | a good visual for this might be a nice full-screen isst view of something cool |
| 13:18.55 | ``Erik | hm, what about a handful of keen images (archer, isst, stryker/rise, a few rt's) canted at like ~20 degrees in the z (so they're all parallelograms in outline) |
| 13:19.18 | starseeker | cool - can you do that? |
| 13:19.21 | ``Erik | semi-randomly placed |
| 13:19.23 | ``Erik | um, no? :D |
| 13:19.30 | ``Erik | just throwing ideas, dude |
| 13:19.32 | dloman | starseeker: you in today? |
| 13:19.37 | starseeker | later today, yes |
| 13:19.39 | dloman | kk |
| 13:19.40 | starseeker | dloman: what's up? |
| 13:20.09 | dloman | well brlcad is here already, so i figure the three of us + Photoshop CS5 can get hte job done relatively quickly. |
| 13:20.24 | starseeker | dloman: you don't need me for that |
| 13:20.31 | starseeker | you and brlcad can handle it fine |
| 13:20.36 | dloman | collective input |
| 13:20.47 | dloman | is stupid busy. |
| 13:21.04 | starseeker | nods - I won't speed up the process much, trust me |
| 13:21.36 | ``Erik | photo shop? that's the windows imitation of gimp, rgiht? :> |
| 13:21.56 | dloman | lol |
| 13:22.04 | dloman | i use both pretty extensively |
| 13:22.21 | dloman | and, this is the one area where the Commercial package spanks the FOSS version. |
| 13:22.46 | starseeker | ``Erik: I know - we could use a snapshot from brlcad's gource visualization :-P should be guaranteed to scare off the weak |
| 13:25.21 | ``Erik | hm, don't hvae gimp on my laptop, compiling it and it's chain would take too long, and sketchup looks wrong for what I want to say, damn |
| 13:25.36 | starseeker | inkscape? |
| 13:25.52 | starseeker | or xfig? :-P |
| 13:26.13 | dloman | ``Erik: you run a gui of any sorts? |
| 13:26.35 | ``Erik | dlo: mac. |
| 13:26.50 | dloman | ..theres no precompiled GIMP binaries? |
| 13:26.54 | ``Erik | probably are |
| 13:27.00 | dloman | kk |
| 13:27.09 | ``Erik | http://www.gimp.org/macintosh/ heh |
| 13:27.21 | CIA-52 | BRL-CAD: 03starseeker * r43909 10/brlcad/branches/STABLE/ (27 files in 17 dirs): Sync STABLE to trunk r43908 |
| 13:27.31 | starseeker | there we go |
| 13:27.36 | starseeker | gets moving |
| 13:27.47 | starseeker | lots to do, too few hours... |
| 13:30.37 | ``Erik | mebbe brlcad will go to prost and talk to matt about critiquing what we do or hints on how to not suck? |
| 13:31.20 | ``Erik | hits the rainlocker and heads off, hasta |
| 13:43.29 | *** join/#brlcad alex_joni (~alex_joni@emc/board-of-directors/alexjoni) | |
| 13:54.38 | CIA-52 | BRL-CAD: 03davidloman * r43910 10/geomcore/trunk/ (4 files in 2 dirs): Rework DataSource interface to better represent current repo design initiative. (plus cascading changes.) |
| 14:05.54 | CIA-52 | BRL-CAD: 03davidloman * r43911 10/geomcore/trunk/ (3 files in 2 dirs): Stub in SvnDataSource class |
| 14:06.55 | ``Erik | 10:06AM up 313 days, 16:43, 2 users, load averages: 0.05, 0.11, 0.14 |
| 14:07.18 | ``Erik | doh, wrong machine :D |
| 14:07.21 | dloman | thats good to know ``Erik , thanks! |
| 14:07.22 | ``Erik | 10:06AM up 414 days, 4:55, 17 users, load averages: 1.91, 2.03, 2.02 |
| 14:07.44 | ``Erik | starting to get heavy again, might break the old record |
| 14:08.25 | dloman | ``Erik: whats your FB post about? |
| 14:08.56 | ``Erik | just everything all happening at once, crazy couple of months |
| 14:09.03 | dloman | ah, got ya. |
| 14:09.35 | ``Erik | I mean, that whole area is lighting up like ww3 is about to start, natural disasters of unprecidented scale elsewhere, ... |
| 14:10.33 | dloman | maybe, just maybe, there will be some finality to the (seemingly) endless conflict in/around the middle easy. |
| 14:10.36 | dloman | lol |
| 14:10.37 | dloman | east |
| 14:11.19 | ``Erik | perhaps, if they modernize, but we're regressing on the flipside, so... |
| 14:11.48 | ``Erik | (mebbe that's the REAL pole shift to be scared about :D ) |
| 14:12.03 | dloman | =D |
| 14:12.55 | dloman | That would make a decent 'what if' scifi story: Earth's axial tilt gets knocked to near 90 degrees..... |
| 14:15.03 | ``Erik | ever read "lucifers hammer"? pournelle and niven, I think? |
| 14:15.15 | dloman | can't say that I have. |
| 14:15.28 | ``Erik | similar situation, amusing enough book :) |
| 14:16.07 | dloman | wonders if there is an audiobook for it..... |
| 14:16.47 | dloman | if pournelle can tame niven's habit for painful science detail, then it could be good :) |
| 14:17.13 | ``Erik | hehehe, I dig niven, I like the detail |
| 14:17.19 | dloman | the ringworld series was interesting, but it felt like i was reading a tech manual / textbook at times ;) |
| 14:17.36 | dloman | "Screw plot development, lets talk about cool geeky things" |
| 14:17.40 | ``Erik | needs to pick up pratchets discworld series |
| 14:17.40 | dloman | =D |
| 14:17.51 | dloman | heh, we have all of them |
| 14:17.54 | dloman | well, most of them. |
| 14:18.04 | dloman | he got really preachy in his latests books. |
| 14:18.34 | dloman | He's a faithfully devout Athiest and started to use his books to evangelize :/ |
| 14:18.55 | ``Erik | hm |
| 14:19.00 | dloman | that was kinda annoying, but the first, oh, 8-10 books in the series were great. |
| 14:19.09 | ``Erik | also need to restock on harry harrison and mercedes lackey O.o |
| 14:19.20 | ``Erik | does robert asprin still write? is he still alive? O.o |
| 14:19.23 | dloman | I have a few audio books for pratchet if you want. |
| 14:19.31 | ``Erik | oh, he died in '08 :( |
| 14:19.33 | dloman | no idea. Who's that? |
| 14:19.46 | ``Erik | the 'myth' series and 'phules company' |
| 14:19.54 | ``Erik | camp comedy scifi |
| 14:20.15 | ``Erik | http://en.wikipedia.org/wiki/MythAdventures was his big series |
| 14:20.54 | ``Erik | kinda mel brookes goofy |
| 14:21.30 | dloman | hrm, looks interesting. |
| 14:21.37 | dloman | bookmarks. |
| 14:21.52 | dloman | so I was wondering... |
| 14:21.57 | ``Erik | the blue pill |
| 14:22.05 | dloman | i found a website pertaining to the whole Ringworld thing |
| 14:22.22 | dloman | some guy, back in the 80's, tried to do a simulation of the Ringworld. |
| 14:22.34 | dloman | but hit a wall due to computer limitations |
| 14:22.44 | dloman | ....think we could model it in brlcad? |
| 14:22.46 | ``Erik | hm, we do have a hair more computational power these days, I think |
| 14:22.50 | dloman | simply of course |
| 14:23.06 | ``Erik | model it? sure, um |
| 14:23.12 | dloman | although, an algo that would generate terrain in addition to the ring would be neat. |
| 14:23.15 | ``Erik | accuracy will get gimpy due to the sizes |
| 14:23.26 | ``Erik | I thought we had a few of those |
| 14:23.37 | dloman | ....but on that scale? :) |
| 14:23.44 | ``Erik | fractal noise generators that create dsp's, etc |
| 14:23.53 | ``Erik | sure, big enough fractal, it's just a memory constraint |
| 14:24.12 | dloman | ponders |
| 14:24.22 | dloman | down to what resolution though.... |
| 14:24.31 | dloman | 1km ? 10m ? |
| 14:24.36 | ``Erik | could alter one to do 'bound' fractals to chain them along by preseeding one side and generate a metric buttload of dsps' to cover the inner surface |
| 14:25.04 | dloman | i think its funny how you can spew tech stuff and still work the word 'buttload' into it. |
| 14:25.10 | dloman | =D |
| 14:25.39 | ``Erik | 10m should be okish, I'd start getting nervous about 1m... um, the estimated orbital distance is about jupiters orbit, and we store purely in mm |
| 14:26.11 | ``Erik | I'm under the impression that jupiter is (let me figure a better word) a whole metric assload of mm's from the center of the sun |
| 14:26.28 | ``Erik | so double precision will start stretching out a fair bit |
| 14:26.43 | dloman | right, but we have 64 bits of storage to work with right? And with doubles, we get what.... 15 decimals of precision? |
| 14:26.54 | ``Erik | 15 decimals is a falacy |
| 14:27.09 | dloman | ...hence the question mark :P |
| 14:27.18 | ``Erik | it's a logorithmic encoding, with the most accuracy right at 1.0ish |
| 14:27.34 | ``Erik | if you get too small or too big, the discernable difference increases |
| 14:27.44 | dloman | well forget float then, why not go with pure integer? |
| 14:28.36 | ``Erik | I was actually trying to come up with a system in the late 90's, wanting to right an xwing/wingcommander type game for an entire solar system, did the math and double was 'good enough', after being mocked for trying to come up with a tiered integer system |
| 14:29.25 | ``Erik | but that was a space fighter sim, the speeds and distances were assumed to be significant, so mapping between model and world space ... I don't recall :/ |
| 14:29.27 | dloman | what was your minimum accuracy with that system? 1m 10m? |
| 14:29.58 | ``Erik | I want to think a full double was down to a handful of cm, but I really don't remember |
| 14:30.17 | ``Erik | that's 64b packed, 80b computation, the norm for intel |
| 14:31.10 | ``Erik | of course, a ringworld model could hold translation matrices all up the chain to gain accuracy, ya just might get weirdness if you xpush'd |
| 14:31.23 | dloman | hehehe |
| 14:31.43 | ``Erik | might be a fun diversion in april :D |
| 14:31.53 | dloman | that'd be a pretty awesome ISST demo |
| 14:32.00 | dloman | april == offsite? |
| 14:32.29 | ``Erik | nah, offsite should be serious, but we all have too much to worry about this month, can't have diversions like that |
| 14:33.23 | ``Erik | (mebbe we shoulda made geomserv the offsite, it was postponed because geomserv was more important, from what I understand) |
| 14:42.09 | dloman | well now, work on the Infinity engine has certainly come along nicely. Eye candy: http://www.youtube.com/watch?v=h7eREddMjt4 |
| 14:52.54 | CIA-52 | BRL-CAD: 03davidloman * r43912 10/geomcore/trunk/ (5 files in 4 dirs): Modify GeometryReqMsg to carry 'Path' and 'Recurse' vars instead of 'ReqType' and 'Path' vars |
| 15:08.55 | *** part/#brlcad sofichinu (~sofichinu@115.69.147.65) | |
| 15:09.17 | *** join/#brlcad KimK (~Kim__@ip174-71-95-176.om.om.cox.net) | |
| 16:23.07 | CIA-52 | BRL-CAD: 03starseeker * r43913 10/brlcad/branches/cmake/ (5 files in 4 dirs): MFC r43912 |
| 16:57.06 | CIA-52 | BRL-CAD: 03davidloman * r43914 10/geomcore/trunk/src/GS/ (geomserv.config geomserv.cxx): Clean up and standardize Datasource declarations in the .config file |
| 17:02.45 | CIA-52 | BRL-CAD: 03davidloman * r43915 10/geomcore/trunk/ (include/DataManager.h src/GS/DataManager.cxx): Simplify DataManager by only allowing it to handle a single DataSource.... for now. |
| 17:42.57 | CIA-52 | BRL-CAD: 03starseeker * r43916 10/brlcad/trunk/src/ (libbu/booleanize.c libged/red.c): Two things to fix red behavior - need blank and not space in one place in the regex matching, and recognize (null) as a boolean 0 in bu_str_true |
| 17:48.01 | starseeker | pwd |
| 17:48.10 | starseeker | whoops |
| 17:50.41 | CIA-52 | BRL-CAD: 03starseeker * r43917 10/brlcad/branches/STABLE/src/ (libbu/booleanize.c libged/red.c): Sync STABLE to trunk r43916 |
| 17:52.48 | CIA-52 | BRL-CAD: 03starseeker * r43918 10/brlcad/branches/cmake/src/ (libbu/booleanize.c libged/red.c): MFC r43917 |
| 18:09.14 | CIA-52 | BRL-CAD: 03davidloman * r43919 10/rt^3/trunk/ (27 files in 25 dirs): Purge out all GeometryService related stuff. This now lives in the GeomCore module. |
| 18:10.16 | *** join/#brlcad Nohla (~Nohla@64.76.19.227) | |
| 18:11.37 | CIA-52 | BRL-CAD: 03starseeker * r43920 10/brlcad/trunk/src/libged/color.c: edcolor was eating the colors - check for the right number of digits in the scan line |
| 18:14.38 | CIA-52 | BRL-CAD: 03starseeker * r43921 10/brlcad/trunk/NEWS: edcolor was removing old color tables and not accepting new ones - fixed scan logic to expect the actual number of digits being scanned for. |
| 18:16.50 | CIA-52 | BRL-CAD: 03starseeker * r43922 10/brlcad/branches/cmake/ (NEWS src/libged/color.c): MFC r43921 |
| 18:17.45 | CIA-52 | BRL-CAD: 03starseeker * r43923 10/brlcad/branches/STABLE/ (NEWS src/libged/color.c): Sync STABLE to trunk r43921 |
| 19:23.31 | *** join/#brlcad bhinesley (~bhinesley@adsl-99-125-83-101.dsl.bkfd14.sbcglobal.net) | |
| 19:44.19 | bhinesley | Hello everyone, I'm a student participating in the GSoC. I would appreciate any suggestions for project proposals. I have about 5 years of professional experience in 3d CAD design (designing commercial plumbing systems in AutoCAD). I have minimal experience with C++ and Java, but I have coded dozens of VBA programs for the companies I have worked for. I use Linux all the time, personally, and I also have about 3 years of |
| 19:44.19 | bhinesley | experience administering a Linux server on a ~25 workstation network. |
| 19:45.49 | bhinesley | Also, I am somewhat familiar with Python, and I believe I could get up to speed quickly. |
| 19:50.50 | *** join/#brlcad yukonbob_ (~bch@20-144.wireless.kamloops.net) | |
| 19:50.53 | starseeker | do you have particular interests? |
| 19:51.56 | starseeker | bhinesley: if you have experience with modeling interfaces, you might want to take a look at our ideas for GUI enhancement |
| 19:52.42 | starseeker | We're using Tcl/Tk, which is a bit different from Python, but not tremendously difficult |
| 19:54.35 | starseeker | http://brlcad.org/wiki/Google_Summer_of_Code/Project_Ideas |
| 19:55.00 | bhinesley | I don't have any particular interests, that I can think of yet. I don't have experience with modeling interfaces. |
| 19:55.15 | starseeker | well, you used AutoCAD yes? |
| 19:55.17 | bhinesley | Learning a new language isn't really a problem |
| 19:55.23 | bhinesley | yes |
| 19:55.29 | starseeker | that's experience :-) |
| 19:56.00 | starseeker | bhinesley: the first thing I'd suggest is compiling BRL-CAD and taking a look at MGED and Archer |
| 19:56.11 | starseeker | they're where the current GUI development action is |
| 19:56.31 | bhinesley | I think I misunderstood you, I thought you meant programmer-modeling user interfaces |
| 19:56.51 | bhinesley | okay |
| 19:56.55 | starseeker | Was your pipe design work in 2D or 3D (drawings, or three space)? (there's no wrong answer) |
| 19:57.01 | bhinesley | 3d |
| 19:57.07 | bhinesley | for fabrication |
| 19:57.09 | starseeker | nods |
| 19:57.42 | bhinesley | although, obviously I am proficient in 2d drawings as well |
| 19:57.55 | starseeker | If you're not afraid to get down and dirty with Tcl/Tk, I'd suggest taking a look at the sketch editing task and possibly the Ayam task |
| 19:58.05 | starseeker | do you have any mathematical background? |
| 19:58.29 | bhinesley | yes, I have finished calculus 2 (out of 3; I'm on a semester system) |
| 19:58.48 | starseeker | OK - Ayam is related to NURBS, which is pretty heavy duty in the mathematical department |
| 20:00.29 | bhinesley | interesting... I am assuming it involves more mathematics than I am familiar with(?) |
| 20:01.42 | starseeker | It depends... the mathemathical requirements may not be absolutely essential for mapping Ayam nurbs editing to our nurbs editing, but you'll have to become familiar with NURBS structures |
| 20:03.12 | starseeker | Our sketch editor needs help rather badly, so if you're more comfortable with 2D drafting you can take a look at our sketch editor, at what TkCAD can do, and where to go from there |
| 20:04.33 | starseeker | http://brlcad.org/wiki/MGED_Sketch_Editor_Migration_and_Enhancement |
| 20:04.44 | starseeker | http://brlcad.org/wiki/Ayam_Editor_Feature_Integration |
| 20:05.13 | starseeker | those will have links to get you started |
| 20:05.18 | bhinesley | thank you |
| 20:05.44 | bhinesley | not sure if this helps, but here is a picture of some of my work: http://stashbox.org/manage_file/1087048/pipe |
| 20:05.48 | starseeker | remember those are just suggestions - be sure to pick something you find interesting |
| 20:06.12 | bhinesley | certainly |
| 20:06.49 | starseeker | I base those suggestions on the fact that you've worked at modeling tasks before, and thus have background interacting with graphical modeling tasks |
| 20:07.40 | bhinesley | that's great, that is exactly what I was looking for |
| 20:08.16 | starseeker | If you're familiar with Python Tcl/Tk shouldn't seem too alien, although Archer in particular makes heavy use of Itcl/Itk |
| 20:09.31 | starseeker | bhinesley: you'll want to get BRL-CAD built first and foremost, since everything follows from that, and get MGED and Archer running |
| 20:09.47 | bhinesley | yes, I'll do that right now. |
| 20:10.17 | starseeker | then explore the sketch editor (I'll offer a shortcut - when you have mged running, create a basic sketch and then edit it to get the gui) |
| 20:11.00 | starseeker | from the MGED command line: |
| 20:11.03 | starseeker | make sketch.s sketch |
| 20:11.14 | starseeker | sed sketch.s |
| 20:11.51 | starseeker | (remember to compile using --with-ogl to get Archer working) |
| 20:12.19 | starseeker | bhinesley: I can't seem to see that stashbox.org link |
| 20:14.05 | bhinesley | Hmm, I was afraid of that. I'm trying to find an alternative. |
| 20:16.22 | bhinesley | try this: https://picasaweb.google.com/lh/photo/_dWpWLr1esGb16X7_4DlNHMyrgI048JfNwPx7Dl9cn0?feat=directlink |
| 20:17.11 | bhinesley | that doesn't show much solids modeling, but I've done that too |
| 20:17.37 | bhinesley | let me clarify: a tool was used for sizing the pipe |
| 20:17.58 | bhinesley | all routing/placement was done manually |
| 20:18.36 | bhinesley | I modeled all of the pumps/tanks to spec though |
| 20:21.05 | starseeker | cool |
| 20:26.45 | bhinesley | so what kind of stuff do you do for BRL-CAD, if you don't mind me asking/ |
| 20:26.49 | bhinesley | *? |
| 20:29.29 | starseeker | bug fix, development, support |
| 20:30.11 | starseeker | the guy to really listen to is brlcad, he'll probably appear in channel later |
| 20:30.38 | bhinesley | oh okay. founder? |
| 20:30.58 | starseeker | he's the lead of the open source project |
| 20:31.08 | starseeker | the actual founder of BRL-CAD itself is Mike Muuss |
| 20:31.43 | starseeker | it's a very old project: http://brlcad.org/d/about |
| 20:32.05 | starseeker | is a newbie, relatively speaking :-) |
| 20:33.06 | starseeker | bhinesley: if you want to ask questions, go ahead and post them - it may be a few hours before anyone responds, but that's normal - we read backlogs |
| 20:33.36 | bhinesley | hey, that's cool |
| 20:33.37 | bhinesley | is a newbie, in absolute terms |
| 20:33.51 | bhinesley | thanks for your help |
| 20:34.00 | starseeker | no problem - any of that look interesting? |
| 20:34.58 | bhinesley | I haven't looked yet, I'm working on getting the source |
| 20:35.09 | starseeker | nods |
| 20:35.25 | starseeker | remember not to check out the whole thing - you only want the latest development sources: |
| 20:35.38 | starseeker | svn co https://brlcad.svn.sf.net/svnroot/brlcad/brlcad/trunk brlcad |
| 20:35.53 | bhinesley | oh thanks, I wasn't sure which to checkout |
| 20:37.04 | starseeker | the toplevel files have instructions for building - ideally, ./autogen.sh && ./configure --enable-all --with-ogl && make && make install will do it |
| 20:39.10 | CIA-52 | BRL-CAD: 03starseeker * r43924 10/geomcore/branches/fossil/ (4 files in 3 dirs): Get checking building, albeit with a lot of implicit warnings |
| 20:39.59 | starseeker | bhinesley: be sure you review the checklist - do you have a sourceforge account? |
| 20:41.38 | bhinesley | I will in about 30 seconds |
| 20:51.33 | CIA-52 | BRL-CAD: 03starseeker * r43925 10/geomcore/branches/fossil/ (4 files in 2 dirs): Add find_option to basic, but this does not belong in a library (none of the argc/argv stuff does) and will be part of a restructuring |
| 20:56.01 | CIA-52 | BRL-CAD: 03starseeker * r43926 10/geomcore/branches/fossil/src/libgeomvcs/checkin.c: update find_option |
| 21:31.51 | *** join/#brlcad Emma (~Emma@p5.eregie.pub.ro) | |
| 22:30.35 | CIA-52 | BRL-CAD: 03starseeker * r43927 10/geomcore/branches/fossil/ (10 files in 2 dirs): Declare some more functions. |
