| 00:18.18 | CIA-28 | BRL-CAD: 03brlcad * r34316 10/brlcad/trunk/src/librt/primitives/pnts/pnts.c: |
| 00:18.18 | CIA-28 | BRL-CAD: simplify the export calculations for the point data size: use masks so we can |
| 00:18.18 | CIA-28 | BRL-CAD: avoid error-prone magic numbers. also remove the unnecessary void pointer |
| 00:18.18 | CIA-28 | BRL-CAD: casting redundancy in _describe(), BU_LIST_FOR gives us the right iterator type. |
| 00:19.00 | CIA-28 | BRL-CAD: 03brlcad * r34317 10/brlcad/trunk/src/librt/primitives/pnts/pnts.c: ws |
| 00:25.52 | CIA-28 | BRL-CAD: 03brlcad * r34318 10/brlcad/trunk/src/librt/primitives/pnts/pnts.c: cleanup and prepare to fix ifree() |
| 00:27.55 | CIA-28 | BRL-CAD: 03brlcad * r34319 10/brlcad/trunk/src/librt/primitives/pnts/pnts.c: print the type if we encounter something unexpected to help with debugging |
| 01:02.53 | b0ef | jdoliner: not really sure what you mean by hybridization between B-REP and CSG; you will work on NURBS in other words?;) |
| 01:30.59 | *** join/#brlcad rbfish2k (n=chatzill@pool-71-125-225-196.nycmny.east.verizon.net) | |
| 02:06.58 | CIA-28 | BRL-CAD: 03brlcad * r34320 10/brlcad/trunk/src/librt/primitives/pnts/pnts.c: |
| 02:06.58 | CIA-28 | BRL-CAD: restore an _ifree() implementation that still 'cheats' by just leveraging the |
| 02:06.58 | CIA-28 | BRL-CAD: fact that all of the point structure types have a bu_list element that we can |
| 02:06.59 | CIA-28 | BRL-CAD: iterate over. we can save a little time by avoiding to dequeue the points too |
| 02:06.59 | CIA-28 | BRL-CAD: since the list is being wiped out anyways. stub in the _print() callback |
| 02:07.01 | CIA-28 | BRL-CAD: pretending that rt_pnts_internal is what we're going to use for the solid |
| 02:07.03 | CIA-28 | BRL-CAD: specific structure. |
| 02:47.14 | *** join/#brlcad Ralith (n=ralith@216.162.199.202) | |
| 02:52.04 | CIA-28 | BRL-CAD: 03brlcad * r34321 10/brlcad/trunk/src/librt/primitives/pnts/pnts.c: one more void* elimination along with consolidating the buf size to one place. |
| 02:57.34 | *** join/#brlcad Ralith (n=ralith@216.162.199.202) | |
| 03:11.48 | CIA-28 | BRL-CAD: 03brlcad * r34322 10/brlcad/trunk/src/librt/primitives/pnts/pnts.c: make the point iteration consistent with the other routines -- don't use a separate head pointer. add a couple magic sanity checks too |
| 03:16.58 | *** join/#brlcad Ralith (n=ralith@216.162.199.202) | |
| 03:53.38 | *** join/#brlcad Ralith (n=ralith@216.162.199.202) | |
| 03:59.14 | *** join/#brlcad Ralith (n=ralith@216.162.199.202) | |
| 05:20.22 | starseeker | hrm. |
| 05:20.40 | starseeker | appears to have exceeded the memory capabilities of his system |
| 05:29.28 | Ralith | haiku runs nicely on my laptop :D |
| 05:29.36 | starseeker | cool! |
| 05:29.40 | Ralith | I should install it to disk and see if it's actually responsive from there >_> |
| 05:29.53 | starseeker | if it supports the laptop hardware |
| 05:30.07 | Ralith | hm? |
| 05:30.12 | Ralith | it's running pretty well right now... |
| 05:30.20 | Ralith | full resolution, functional network, mouse, keyboard |
| 05:30.22 | starseeker | wait, are you running in virtual machine? |
| 05:30.28 | Ralith | nope |
| 05:30.31 | starseeker | !! |
| 05:30.31 | Ralith | booted it from a USB stick |
| 05:30.32 | starseeker | nice |
| 05:30.43 | Ralith | pretty shiny. |
| 05:30.58 | Ralith | kinda dissapointed that ff2 came up though |
| 05:31.01 | Ralith | was hoping for a native browser |
| 05:31.07 | Ralith | that is, custom browser |
| 05:31.11 | starseeker | ah |
| 05:31.19 | starseeker | thinks that will come after a stable release |
| 05:31.28 | starseeker | probably use webkit or gecko though |
| 05:33.20 | Ralith | yeah, I remembered reading about some effort to transfer existing work to webkit |
| 05:41.20 | Ralith | 300fps spinning teapot in software |
| 05:41.21 | Ralith | not bad |
| 06:33.19 | *** join/#brlcad elena (n=ebautu@92.86.0.28) | |
| 06:38.46 | *** join/#brlcad mafm (n=mafm@223.Red-83-49-86.dynamicIP.rima-tde.net) | |
| 06:38.58 | mafm | hi |
| 06:44.37 | Ralith | hi |
| 06:46.33 | *** join/#brlcad LarsG (n=lars@nusnet-228-5.dynip.nus.edu.sg) | |
| 06:51.29 | elena | hello |
| 07:09.17 | madant | howdy mafm |
| 07:09.43 | madant | Ralith: how's the gui planning coming up :) |
| 07:16.02 | *** part/#brlcad elena (n=ebautu@92.86.0.28) | |
| 07:22.00 | *** join/#brlcad _clock_ (n=_sushi_@77-58-147-167.dclient.hispeed.ch) | |
| 07:52.36 | *** join/#brlcad mafm_ (n=mafm@223.Red-83-49-86.dynamicIP.rima-tde.net) | |
| 08:46.25 | *** join/#brlcad elena (n=ebautu@89.136.118.141) | |
| 08:58.50 | brlcad | wanders off to practice |
| 09:01.27 | brlcad | Ralith: that's pretty cool, so next step is get a compiler installed and work on a minimum brl-cad compilation? :) |
| 09:03.11 | Ralith | maybe! |
| 09:03.18 | Ralith | madant: slower than I'd like |
| 09:07.34 | madant | :) one thing i learned last time is :D most things take longer than you anticipate :) |
| 09:22.15 | *** part/#brlcad elena (n=ebautu@89.136.118.141) | |
| 09:22.23 | *** join/#brlcad elena (n=ebautu@89.136.118.141) | |
| 09:50.26 | *** part/#brlcad elena (n=ebautu@89.136.118.141) | |
| 09:54.06 | *** join/#brlcad elena (n=ebautu@89.136.118.141) | |
| 10:53.20 | d-lo | nornin all! |
| 10:58.57 | ``Erik | yargh, matey, shiver me timbers, or somethin' |
| 11:09.41 | *** join/#brlcad madant_ (n=d@117.196.133.20) | |
| 11:24.34 | *** join/#brlcad Elrohir (n=kvirc@p5B14DF88.dip.t-dialin.net) | |
| 11:43.37 | *** join/#brlcad Elrohir (n=kvirc@p5B14DF88.dip.t-dialin.net) | |
| 11:55.14 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-67.sbndin.btas.verizon.net) | |
| 12:04.04 | *** part/#brlcad elena (n=ebautu@89.136.118.141) | |
| 12:12.51 | *** join/#brlcad Elrohir (n=kvirc@p5B14DF88.dip.t-dialin.net) | |
| 12:25.59 | *** part/#brlcad LarsG (n=lars@nusnet-228-5.dynip.nus.edu.sg) | |
| 12:27.28 | *** join/#brlcad Elrohir (n=kvirc@p5B14DF88.dip.t-dialin.net) | |
| 13:22.28 | *** join/#brlcad Elrohir (n=kvirc@p5B14DF88.dip.t-dialin.net) | |
| 13:22.58 | ``Erik | cooks up some corned beef hash and eggs O.o |
| 13:23.22 | d-lo | Now *THAT* sounds good... Mmmmmm |
| 13:23.52 | ``Erik | the gf thought I was batshit insane for buying a can of corned beef hash |
| 13:23.54 | ``Erik | it's good shit! |
| 13:24.03 | ``Erik | fat kid shit, but good! |
| 13:24.52 | ``Erik | I think I'll bring eds lappie back to him tomorrow |
| 13:25.17 | ``Erik | in th emean time, this ucw action vs function crap not acting like call/cc should is strusfrating me. |
| 13:25.50 | brlcad | that's good stuff |
| 13:26.34 | ``Erik | (it's my rdo, I think I've dropped more lines of code this morning than in the last fortnight) |
| 13:28.43 | brlcad | just had a delectable plate of his own creation. yum! |
| 13:29.02 | ``Erik | think you guys could spare a corner with a workstation for me? I'm getting clobbered with questions left and right that burn time and even lead to philosphical debates :( that he actually stated that he was doing me a favor by not asking me a question for a WHOLE FUCKING HOUR |
| 13:29.12 | ``Erik | I have shit to do, I need to hide to do it :( |
| 13:30.45 | ``Erik | indianlarry takes most of it for me as he's not on a crunch for anything, but when he's not there or doesn't have the answers... *grouse* |
| 13:31.14 | ``Erik | I might be turnin' those new suns into my chair real soon |
| 13:31.24 | brlcad | um, maybe don't get suckered into philosophical debates |
| 13:32.04 | ``Erik | hehehe, it's hard, he has ideas that he needs to be heared, and when ya disagree, it must be because you didn't understnad him |
| 13:32.05 | ``Erik | :/ |
| 13:32.14 | d-lo | lol |
| 13:32.16 | brlcad | i'm sure that's not all the questions, but that one you can certainly do something about |
| 13:32.16 | d-lo | lock the door |
| 13:32.30 | ``Erik | uhhh, he has a key? :D |
| 13:33.02 | ``Erik | waterfall/perfection vs agile/chaotic |
| 13:33.05 | ``Erik | *sigh* :D |
| 13:33.09 | d-lo | oh, THAT one. I thought you were talking about the previous owner of cliffs seat. |
| 13:33.23 | ``Erik | no, he's totally controllable, not an issue at all |
| 13:34.50 | ``Erik | last friday, I said "ok, if you think that's how it should be, go do it. If it works, people will follow" way too many times |
| 13:35.29 | ``Erik | <-- was seriously close to flipping out, wasn't planning on going to lunch, but HAD to get out of that environment O.o |
| 13:35.44 | ``Erik | gave me a damn headache :D |
| 13:35.48 | ``Erik | aaaanyways |
| 13:36.31 | ``Erik | I think I might have to look for a place to squat in to crank some productivity, dark and music... O.o |
| 13:36.50 | d-lo | You like headphones? |
| 13:39.50 | ``Erik | I can't use headphones for very long, they tend to push on my jaw just below my ears and give me a headache |
| 13:43.04 | *** join/#brlcad Elrohir (n=kvirc@p5B14DF88.dip.t-dialin.net) | |
| 13:44.02 | starseeker | tries to gear up argouml and discovers that a major python version update on gentoo == pain |
| 13:46.20 | ``Erik | d-lo, do you tubesperms do mess duty, or do ya'll get a specialist? |
| 13:46.42 | ``Erik | s/allothta/pasttesnse()/ |
| 13:48.13 | d-lo | Mess duty only when you are new to the boat and can't stand any Watchstation. |
| 13:48.20 | d-lo | NUB = Non Useful Body |
| 13:48.26 | ``Erik | :D |
| 13:48.31 | d-lo | NUB = Mess duty. |
| 13:48.45 | ``Erik | this somehow reminds me of navy mess... *cookcoookcook* |
| 13:49.19 | d-lo | If you aren't pulling out 15 day old leftovers, you aren't doing it right imho. =D |
| 13:49.39 | ``Erik | hehehe |
| 13:51.49 | ``Erik | noms his brunch |
| 13:54.56 | CIA-28 | BRL-CAD: 03erikgreenwald * r34323 10/brlcad/trunk/include/bn.h: stub the notion of a point |
| 13:58.56 | brlcad | enum ftw |
| 14:00.22 | ``Erik | public thinking *shrug* when I get around to the actual draw routines, mebbe I'll consider applying it |
| 14:01.01 | ``Erik | mmm salty, but good, I should buy corned beef hash more often :D |
| 14:03.29 | ``Erik | heads out to check up on his car, renew his rental contract and stop by the grocery store O.o |
| 14:06.06 | *** join/#brlcad madant (n=d@117.196.145.148) | |
| 14:07.55 | d-lo | ...just don't go out for pizza! *ducks* |
| 15:01.13 | *** join/#brlcad mafm (n=mafm@83.49.86.223) | |
| 15:01.22 | mafm | hi |
| 15:15.23 | brlcad | howdy mafm |
| 15:28.37 | *** join/#brlcad elite01 (n=omg@unaffiliated/elite01) | |
| 15:31.07 | madant | hi mafm |
| 15:41.46 | *** join/#brlcad hippieindamakin8 (n=hippiein@202.3.77.38) | |
| 15:59.11 | CIA-28 | BRL-CAD: 03brlcad * r34324 10/brlcad/trunk/include/nmg.h: alignment cleanup while looking where data could be attached |
| 15:59.14 | brlcad | hippieindamakin8: hola! any progress on that nmg routine? |
| 15:59.31 | hippieindamakin8 | brlcad, hey |
| 15:59.50 | hippieindamakin8 | not yet. just back from 7 hrs of continous examinations. |
| 16:00.27 | hippieindamakin8 | i ll do that on 30th as soon as i am done with the endsems . got 2 more on wednesday and i ll be done. |
| 16:02.52 | hippieindamakin8 | brlcad, i ll look through the nmg_junk and the other conversions to nmg and rewrite it then |
| 16:12.26 | brlcad | hippieindamakin8: oh yeah, how'd those go? |
| 16:12.31 | brlcad | your exams that is |
| 16:12.40 | CIA-28 | BRL-CAD: 03brlcad * r34325 10/brlcad/trunk/include/nmg.h: more ws and comment cleanup |
| 16:13.30 | brlcad | hippieindamakin8: when you do look, keep in mind that nmg_junk alludes to the notion that the way it does it for polys from file isn't necessarily the 'right' way |
| 16:13.56 | brlcad | didn't investigate exactly what was meant by that comment but I think it has to do with the fact that the nmg is manually constructed instead of using functions |
| 16:24.33 | ``Erik | no no, I seem to get accosted by curbs when I go out for pizza |
| 17:23.19 | *** join/#brlcad elite01 (n=omg@unaffiliated/elite01) | |
| 17:38.54 | ``Erik | grooves to regina spektor |
| 17:43.39 | *** join/#brlcad mafm (n=mafm@223.Red-83-49-86.dynamicIP.rima-tde.net) | |
| 17:46.16 | *** join/#brlcad mafm (n=mafm@223.Red-83-49-86.dynamicIP.rima-tde.net) | |
| 17:58.01 | CIA-28 | BRL-CAD: 03brlcad * r34326 10/brlcad/trunk/include/nmg.h: looks like indent is messing with some of the defines, reindent. also a few other minor tweaks for consistency and null out the NMG_FREESTRUCT pointer after it's free for sanity too. |
| 18:06.19 | *** join/#brlcad elite01 (n=omg@unaffiliated/elite01) | |
| 18:27.28 | *** join/#brlcad dreeves (n=c752f348@bz.bzflag.bz) | |
| 18:27.52 | dreeves | waives at brlcad and starseeker |
| 18:29.14 | dreeves | ~logs |
| 18:29.15 | ibot | All conversations are logged to http://ibot.rikers.org/channel, where "channel" is replaced by the URL-encoded channel name, such as %23freenode for #freenode. Lines starting with spaces are not logged. |
| 18:54.34 | brlcad | howdy dreeves |
| 18:58.16 | dreeves | I have been very preoccupied with a project but should be freeing up in the next day or two then maybe we can make some more progress I see you all have added some more test data thanks!! |
| 18:58.27 | CIA-28 | BRL-CAD: 03brlcad * r34327 10/brlcad/trunk/include/raytrace.h: align the closing scope with the define macro |
| 19:19.45 | starseeker | hey dreeves |
| 19:33.11 | starseeker | brlcad: Are there any worthwhile open source SIMD abstraction libraries out there? |
| 19:49.49 | *** join/#brlcad elena (n=ebautu@89.136.118.141) | |
| 19:50.21 | elena | hi. |
| 19:50.57 | starseeker | howdy |
| 19:51.04 | elena | how are you? |
| 19:51.12 | starseeker | well. yourself? |
| 19:51.18 | elena | fine. |
| 19:51.59 | elena | do you have some spare time now? |
| 19:52.04 | starseeker | sure |
| 19:52.42 | elena | can we talk about the project plan? |
| 19:53.12 | starseeker | yep. |
| 19:53.18 | elena | great! |
| 19:53.39 | elena | where do we start? |
| 19:54.02 | starseeker | do you have a wiki page with your proposal? |
| 19:54.07 | elena | yes. |
| 19:54.09 | elena | 1 sec. |
| 19:55.04 | elena | http://brlcad.org/wiki/User:EBautu |
| 19:57.49 | starseeker | I guess the main question at the moment is what features are needed/desired? |
| 19:59.36 | elena | ``Erik mentioned a rating system. |
| 19:59.46 | ``Erik | wakes up |
| 20:01.09 | ``Erik | oh, yeah, I figure the point is a community, there'll be both spammers as well as well meaning folk who simply produce crap... |
| 20:01.21 | elena | basiclly, browse for models, search, and upload models. |
| 20:02.38 | starseeker | right - there are some fairly obvious core capabilities without which there is no site. The most important is the model -> web-ready information capabilities |
| 20:03.37 | starseeker | I would say that the user interaction modes are orthogonal to the "handle CAD models" core functionality |
| 20:04.00 | brlcad | personally (and open to debate) I'd like the site to effectively counter abuse in a manner similar to most wikis |
| 20:04.04 | brlcad | that is .. make it easier to remove the content than it is to add it, use recaptcha for anonymous submissions, and provide a strong feedback loop (like cia and/or e-mail notifications on changes) |
| 20:04.19 | starseeker | agrees |
| 20:04.28 | starseeker | recaptcha is a must |
| 20:04.47 | ``Erik | no, recaptcha is NOT a must |
| 20:04.55 | ``Erik | a strong community is a must |
| 20:05.17 | elena | i think we should not allow anonymous submissions. |
| 20:05.30 | starseeker | ``Erik: I would contend that a strong basis of good quality models will attract a community |
| 20:05.45 | elena | the least you can do is create an account if you want to submit something. |
| 20:06.02 | starseeker | my concern with anonymous submissions is being sure that they aren't (say) something from a company's commercial files |
| 20:06.17 | brlcad | elena: put a lot of thought into that particular topic -- there's a reason to allow it |
| 20:06.17 | ``Erik | hackernews (http://news.ycombinator.com) has an aggressive moderation community, reddit.com does not... slashdot uses a designated coop |
| 20:06.40 | brlcad | that's not to say unchecked submissions -- but to allow them with extra effort |
| 20:06.41 | ``Erik | this is a topic that has a lot of random experimentation and little proof |
| 20:08.03 | brlcad | specifically with regards to starseeker's point, *any* submission should have a cited source and the submission process allowing that source to be specified |
| 20:08.54 | starseeker | I would suggest solving the technical problems of CAD model -> web page is the first order of business - how those tools are used is (obviously) more complicated from a social standpoint |
| 20:09.29 | ``Erik | argues that there is no technical solution to a social issue |
| 20:09.48 | brlcad | if the claim is "I made this", then anonymous wouldn't be allowed -- but if some random user found the openmoko model and wanted to add it to the db, nothing should prevent that contribution presuming it's properly cited (and probably manually reviewed) |
| 20:09.58 | starseeker | ok - but we're doing a gsoc project here, so shouldn't the focus be on concrete functionality? |
| 20:10.18 | starseeker | brlcad: ok, that makes sense |
| 20:10.54 | ``Erik | but is the essense o fthe project to create a pile of code? or is it ot begin developing a community? |
| 20:11.07 | starseeker | I would say it's the former |
| 20:11.43 | starseeker | without the former, the latter isn't practical |
| 20:11.49 | brlcad | I'd rather the system be set up as open as possible, and just make it easy to correct and respond to bad use (think wikipedia not journal submission) |
| 20:11.56 | ``Erik | *shrug* I personally believe it's a critical discriminating point |
| 20:12.14 | brlcad | ``Erik: developing the community is our job :) |
| 20:12.39 | ``Erik | I kinda think the point of any software is to support a community, to develope a piece of software with no appreciation of a community is ... pointless |
| 20:12.43 | brlcad | the project is about writing code and becoming a contributor |
| 20:13.05 | brlcad | those aren't orthogonal goals |
| 20:13.09 | ``Erik | and now; I walk like an egyption. *dance* |
| 20:13.23 | starseeker | blinks |
| 20:13.48 | brlcad | elena: so I mentioned this earlier when you weren't around but just in case... |
| 20:13.54 | brlcad | please do keep in mind a few things: 1) conversations should be public, 2) mentoring should similarly be public especially technical discussions, and 3) your listed mentor is predominantly for logistically tracking your progress -- i.e., they're not your "go-to" person for technical discussions -- those really should be open forums (here and/or irc) where any/other devs can at least listen in |
| 20:13.54 | ``Erik | whoa ohh whoahh ohh woohhha ohh wohhh ohh whoaahhhh |
| 20:14.29 | brlcad | mm, and that should have read (here and/or mailing list) ;) |
| 20:14.31 | elena | ok. I know that. |
| 20:14.37 | brlcad | elena: okay, just making sure :) |
| 20:14.50 | brlcad | sometimes a critical piece of information slips through.. makes all the difference |
| 20:14.51 | elena | thank you. |
| 20:15.02 | brlcad | trying to make sure everyone gets the same information this year :) |
| 20:15.17 | brlcad | the channel /topic is useful too ;) |
| 20:15.56 | starseeker | has a question - is the intent to recognize the file format being provided, or to rely on the user to identify the format? |
| 20:16.41 | elena | I didn't do it before, but we could try to recognize the file format. |
| 20:16.49 | elena | using some magic bytes maybe? |
| 20:17.32 | elena | or simpler, it's file extension (not sure how general is that)... |
| 20:17.36 | brlcad | to me, the heart of that project from a task perspective is the ability to upload a .g file (for starters) with annotated metadata, and then visualize that model categorized through the website (raytrace renderings along with more fundamental information like file types, object counts, sizes, dates, license, etc) |
| 20:18.37 | brlcad | it doesn't need to recognize -- there's only a limited set of types |
| 20:18.52 | brlcad | have a selector on upload that specifies the type, then it can be named anything |
| 20:19.11 | elena | ok. that's even simpler :) |
| 20:19.14 | brlcad | could still have the uploader require a consistent naming convention too |
| 20:20.45 | brlcad | elena: what would your preference be in terms of scope -- more focus on conversion capability or more focus on the site index itself? |
| 20:21.25 | elena | what do you mean by "site index"? |
| 20:23.03 | brlcad | the site browsability, the web interface itself |
| 20:23.26 | elena | aaha. |
| 20:23.40 | brlcad | as much as there are lots of potential ideas where this can/will go, there's simply not enough time, so narrowing that scope would be good |
| 20:24.08 | elena | usability is importanta, but i'd like to focus on conversion, too. |
| 20:24.34 | elena | conversion will require me to learn more about brlcad. |
| 20:24.40 | elena | not you, the software :D |
| 20:24.40 | brlcad | e.g. making the front-end framework really easy to use and well organized vs having support for tracking versions of files vs having support for importing/exporting various formats cleanly vs rendering views vs rating/karma system vs robust metadata support, etc |
| 20:24.53 | brlcad | the dash makes all the difference ;) |
| 20:24.56 | brlcad | brlcad vs brl-cad ;) |
| 20:26.33 | elena | i'd go with support import/export, then rating, then rendering, then easy to use, metadata, etc. |
| 20:26.59 | brlcad | now the all-important question -- why? :) |
| 20:26.59 | elena | not sure about the last one. it might be higher weighted. |
| 20:27.36 | elena | import/export is important. without it we have no content. |
| 20:27.39 | brlcad | I can say that clear licensing is a top-priority regardless and that has a little bit to do with metadata |
| 20:28.08 | brlcad | without the licensing sorted out, most of the other features are moot |
| 20:28.21 | brlcad | not complicated, but critical |
| 20:28.22 | elena | rating is important (Erik is right that people may upload bad stuff, by mistake or willingly). |
| 20:28.44 | starseeker | barring a flood of content site admins can address that manually |
| 20:29.20 | starseeker | free CAD models are currently fairly rare |
| 20:29.25 | brlcad | a review queue solves that, similar to drupals comment moderation system |
| 20:29.29 | elena | then I'd say that licensing implies registred users for submissions. |
| 20:29.40 | elena | not necesary to say I made this. |
| 20:29.48 | elena | but to say I put this here. |
| 20:30.21 | brlcad | depends entirely on the source and the license |
| 20:30.25 | elena | it's not necesary for flood. |
| 20:30.40 | elena | but let's say I find 10 models of a cup. |
| 20:30.58 | brlcad | okay |
| 20:31.14 | elena | I'd like to get an idea of other though about them before I start downloading them. |
| 20:31.47 | brlcad | no comprendo |
| 20:31.49 | elena | i, myself, I'd always start with the highest rated ones. |
| 20:32.15 | starseeker | oh, you're talking about a user sorting site content |
| 20:32.32 | elena | yes. sorry, was related to rating and flooding. |
| 20:33.33 | brlcad | rating is simply a means to filter/sort site content -- with the right browsing and searching mechanisms, those aren't necessarily that important; moreover using filtering/searching on top of bad browsing is still usually a horrible user experience |
| 20:34.07 | brlcad | the unfiltered browsing needs to be well organized by itself regardless of searching/filtering aids |
| 20:34.31 | brlcad | a really good example site related to this is freshmeat.net |
| 20:34.42 | starseeker | sourceforge's system of filters might also be worth mentioning here - you can (for example) filter out all search results that don't match a particular license |
| 20:35.10 | elena | I was thinking more of http://extensions.services.openoffice.org/ |
| 20:35.28 | brlcad | freshmeat has evolved into a rather feature-filled system, but it's core browsability remains efficient even with tens of thousands of projects |
| 20:36.18 | starseeker | elena: the trick with rating systems like that is they tend to highlight the "coolest" items |
| 20:36.23 | brlcad | yeah, that's not bad -- pretty similar |
| 20:36.50 | starseeker | for technical searching, my first order of business is to limit the search space to the "must haves" - license being the obvious one |
| 20:36.53 | elena | yes. coolest is not always the best. |
| 20:37.19 | brlcad | my point earlier was scope though -- there's not nearly enough time to "support import/export, then rating, then rendering, then easy to use, metadata, etc" .. the point is (at least for this summer), what to leave out |
| 20:37.34 | elena | btw, what drupal version is on brl-cad.org? |
| 20:37.56 | brlcad | I'd say rating/sorting is one to leave out .. searching and basic browsing would come first |
| 20:38.06 | brlcad | elena: a version that needs to be upgraded :) |
| 20:38.06 | starseeker | brlcad: yeah |
| 20:38.25 | starseeker | agreed - that's why I figured code dealing directly with the cad files was first up |
| 20:38.51 | starseeker | which I guess is import/export and metadata |
| 20:38.53 | elena | agree. |
| 20:39.28 | brlcad | import/export for any geometry type or just .g files (for now)? |
| 20:39.39 | starseeker | would say .g files... |
| 20:40.09 | elena | how about import for g, export multiple types? |
| 20:40.24 | elena | i played just a little with the conversion tools. |
| 20:40.48 | elena | it looked to me that I could used them to convert one to another. |
| 20:40.50 | starseeker | probably want to go the other way, actually - ending up with the .g file, at least BRL-CAD can read it :-) |
| 20:40.56 | elena | but you know better. |
| 20:41.33 | starseeker | the conversion tools can do a lot, but they have their limitations |
| 20:41.45 | elena | yeah. it make more sense that way :) |
| 20:41.51 | brlcad | elena: the issue is that almost every time you convert geometry from format A to format B, it's a lossy conversion (nearly universal to the point of being universal to the CAD industry) |
| 20:42.19 | brlcad | our importers often preserve everything or nearly everything, but our exporters most certainly do not for many formats |
| 20:42.34 | elena | yes. I know. but how lossy is it? |
| 20:42.35 | brlcad | and many formats simply cannot represent various forms of geometry |
| 20:42.40 | brlcad | night and day |
| 20:42.51 | brlcad | fundamentally different representation schemes |
| 20:42.58 | brlcad | not like jpg vs png |
| 20:43.20 | starseeker | elena: sometimes very lossy. that's why you want to always preserve the original submitted format |
| 20:43.39 | elena | got it. |
| 20:43.45 | brlcad | difference is more like text file vs screenshot of text file |
| 20:44.02 | starseeker | even something like Pro-E -> IGES is likely to lose information - Pro-E -> dxf is guaranteed to since dxf can't hold a lot of what Pro-E uses |
| 20:46.09 | elena | then, it's export/import .g or export .g/import many+keep original? |
| 20:46.23 | elena | what's your oppinion? |
| 20:46.23 | brlcad | consider a simple sphere -- we could represent that same sphere in at least five different ways: 1) as a point+radius, 2) as a 2D outline of a circle that is rotated about an axis, 3) as a collection of polygons approximating the surface, 4) as one or more spline surfaces, 5) as a volumetric data set, ... |
| 20:46.59 | elena | yes. you're right. |
| 20:47.14 | brlcad | and brl-cad actually supports all five of those particular representations... :) |
| 20:47.23 | elena | and I've seen this when created an .g object database and exported to triangles. |
| 20:47.39 | brlcad | yet a given export format usually does not -- many export formats only support polygons or triangles for example |
| 20:47.42 | brlcad | exactly |
| 20:47.53 | starseeker | start with export/import .g, imho. Once that's working, more formats can be added |
| 20:49.03 | starseeker | the original format of the model is an important fact to preserve in the metadata (probably the second most important after license) |
| 20:49.16 | elena | ok. |
| 20:49.35 | starseeker | but you should preserve that information for .g files too ;-) |
| 20:49.41 | starseeker | so no harm starting there |
| 20:50.31 | brlcad | and the users can always install our tools and convert to .g if they really want to :) |
| 20:50.44 | brlcad | but that does bring up another point about the site interface |
| 20:51.06 | brlcad | I'd really like it to talk about what format the geometry is or at least what types of geometry are contained |
| 20:51.18 | brlcad | so that if it's 100% triangles, the site will say so |
| 20:51.21 | elena | "users can always install our tools and convert to .g if they really want to" correct! |
| 20:51.51 | brlcad | likewise, a special flag to note whether it's a solid model or not |
| 20:52.27 | ``Erik | convert from, rather? |
| 20:52.42 | brlcad | for .g, the answers to that metadata is pretty simple (there's only 3 primitive types we support that aren't solid) |
| 20:54.56 | elena | is there a tools that gives this information? I don't recall one. |
| 20:55.09 | elena | like some kind on statistics. |
| 20:55.16 | brlcad | mged can |
| 20:57.55 | brlcad | another thing I think will be critical is queueing -- uploaded geometry gets queued for addition, views are queued for rendering, conversions are queued, etc |
| 20:58.31 | brlcad | so that the system can be taken down or manually prodded without interrupting jobs or without locking up resources |
| 20:59.01 | elena | I think so, too. |
| 20:59.30 | ``Erik | if a tool does not exist, it will shortly after notification that it should :) |
| 20:59.37 | brlcad | geometry uploads are going to be generally be very big (possibly hundreds of megs), renderings and conversions are going to take potentially minutes or hours |
| 21:00.03 | *** join/#brlcad jdoliner (n=jdoliner@c-68-51-76-57.hsd1.il.comcast.net) | |
| 21:00.05 | brlcad | yeah, we can make tools as needed to in order to facilitate the site |
| 21:00.32 | elena | really?! I didn't expect geometry files that big. :( |
| 21:00.55 | brlcad | they can be anywhere from a few bytes to a GB or more, depending on the model |
| 21:01.03 | brlcad | we'll just have to manually limit the sizes |
| 21:01.05 | elena | probably that says a lot of my experience :) |
| 21:01.12 | ``Erik | we deal with files that gobble mulpiple gigs on drive :( |
| 21:02.03 | brlcad | not that any of *those* would ever be uploaded, but it's not inconceivable that the system will grow to support large models (they are often the most interesting after all) |
| 21:02.15 | elena | uploading those might be a problem. |
| 21:02.20 | elena | for the user. |
| 21:02.32 | elena | aha. |
| 21:02.34 | brlcad | hence the need for a queue and (ideally), multiple upload methods |
| 21:02.41 | ``Erik | gotta think towards future-rpoofing, though |
| 21:02.47 | starseeker | nods |
| 21:03.44 | brlcad | e.g., could easily support anonymous ftp uploads ala sf.net's former file release system |
| 21:04.16 | brlcad | you upload your model via anon ftp, then select your file from a list |
| 21:04.35 | brlcad | could do similar things with sftp/scp |
| 21:04.42 | ``Erik | I mean, starseeker is making tires that cork hard drives, yet he still can't get my tread right :D |
| 21:04.44 | starseeker | have to be careful about restricting that so it isn't abused, but yes that could work |
| 21:05.03 | starseeker | ``Erik: whatda mean? my tires are tiiiiny |
| 21:05.12 | starseeker | 27k for the default with tread, iirc |
| 21:05.18 | ``Erik | where's my effin' tread? |
| 21:05.24 | brlcad | "sphflake -d8" is enough to make a model that will fill most hard drives :) |
| 21:05.47 | brlcad | tread's off by default |
| 21:05.51 | ``Erik | much less brlcad's! |
| 21:06.01 | starseeker | waiting on a fix for the major/minor axis assumptions in that weird dimensional case, iirc |
| 21:06.29 | starseeker | yeah - I regard tire -p 1 as the "default" treaded example |
| 21:06.32 | ``Erik | I'm still not sure if my fronts are being replaced by all seasons or not |
| 21:07.00 | starseeker | elena: anyway - is this helpful in narrowing your focus? |
| 21:07.16 | elena | yes. very. |
| 21:07.30 | ``Erik | well, damn, I'm doing my best to UNnarrow here :D |
| 21:07.38 | elena | however, I'll copy it and read it again tomorow to filter it out. |
| 21:07.50 | elena | :) |
| 21:07.54 | starseeker | heh |
| 21:08.15 | elena | is this chat logged? |
| 21:08.16 | ``Erik | mannies does soem good lasagna |
| 21:08.20 | ``Erik | yeah, uh |
| 21:08.23 | starseeker | ~log |
| 21:08.23 | ibot | methinks log is as piece of wood, or a record, or the opposite of exponentiation, or http://ibot.rikers.org/%23uphpu/ |
| 21:08.53 | brlcad | ~logs |
| 21:08.53 | ibot | All conversations are logged to http://ibot.rikers.org/channel, where "channel" is replaced by the URL-encoded channel name, such as %23freenode for #freenode. Lines starting with spaces are not logged. |
| 21:08.58 | starseeker | showoff |
| 21:10.08 | starseeker | Ralith: heh - speaking of native Haiku web browsing - looks like a gsoc project will integrate webkit into the native Haiku browser. You might get your updated browsing experience fairly soon :-) |
| 21:10.47 | brlcad | BeOS's old web browser was pretty hilarious -- which is where 'Haiku' got its name |
| 21:11.20 | starseeker | What was that called. Netcrawler or something? |
| 21:11.39 | brlcad | if a page wasn't there, it would give you a haiku |
| 21:11.41 | brlcad | netpositive |
| 21:11.46 | starseeker | ah :-) |
| 21:11.50 | starseeker | oh yeah, I remember that |
| 21:12.07 | ``Erik | there is no page here |
| 21:12.07 | ``Erik | I tried real hard to find one |
| 21:12.07 | ``Erik | but there was none here |
| 21:12.07 | starseeker | got a lot of those things - getting browsing up for a virtual machine is still a trick for me |
| 21:12.08 | brlcad | http://8325.org/haiku/ |
| 21:13.11 | starseeker | wishes in some ways that Microsoft had bought BeOS after Apple dropped it and turned it into the next version of Windows |
| 21:13.23 | brlcad | <PROTECTED> |
| 21:13.26 | starseeker | would have been a way to get a nice Windows |
| 21:13.33 | ``Erik | some of thos eare brilliant |
| 21:13.54 | brlcad | rewrote low-level threading support so that it would cleanly support 10,000 threads multiplex seamlessly |
| 21:14.00 | starseeker | jeez |
| 21:14.19 | starseeker | does Haiku manage any of that yet? |
| 21:14.20 | brlcad | hilarious dev articles |
| 21:14.32 | brlcad | a few things, but not much yet |
| 21:14.43 | brlcad | they've had a mountain to climb just to reproduce the old api |
| 21:14.52 | brlcad | s/reproduce/reimplement/ |
| 21:14.56 | starseeker | nods |
| 21:14.59 | brlcad | without optimizations |
| 21:15.09 | starseeker | who bought the old BeOS code? |
| 21:15.13 | starseeker | Palm was it? |
| 21:15.22 | brlcad | basically |
| 21:15.32 | starseeker | doesn't understand why they're still sitting on it |
| 21:15.33 | brlcad | there were some proprietary licensed portions too |
| 21:15.37 | starseeker | ah |
| 21:16.40 | brlcad | even if they could give away some of the code, they'd be suceptible to shareholder lawsuit |
| 21:16.52 | starseeker | mmm |
| 21:16.54 | starseeker | that sucks |
| 21:17.03 | brlcad | long threads about the be code |
| 21:17.12 | brlcad | lots of folks wanted it |
| 21:17.24 | starseeker | remembers a few abortive attemps to buy it ala blender style, but those never went anywhere either |
| 21:17.29 | brlcad | one even tried (illegally apparently), which is where Zeta OS came from |
| 21:18.35 | starseeker | oh, yeah. I had forgotten them |
| 21:19.02 | starseeker | never made any sense to me - it was clear if the OS was going to have any chance at gaining mind share it would have to be as open source |
| 21:20.22 | starseeker | brlcad: oh, bty - know of any portable open source SIMD abstraction layers? |
| 21:22.09 | brlcad | there are a few projects related to abstraction, but it's pretty low level effort |
| 21:22.43 | starseeker | ah. So, not likely to be terribly useful? |
| 21:22.51 | brlcad | more work designing the algorithms themselves and they're very closely tuned to cache sizes, the algorithm at hand, and the capabilities of the unit |
| 21:29.29 | brlcad | there's projects like gcc's automatic vectorization and other more spansive projects like OpenGL |
| 21:29.36 | brlcad | or the variety of GPGPU libraries |
| 21:30.28 | brlcad | cuda, libsh, brookgpu |
| 21:30.32 | starseeker | was just asking because these raytracing guys seem to like SIMD if they want decent performance on anything... |
| 21:31.33 | brlcad | guys? |
| 21:31.41 | brlcad | context? |
| 21:31.53 | starseeker | the Direct raytracing of NURBS paper |
| 21:31.59 | brlcad | ah, right |
| 21:32.08 | brlcad | very much required to get the performance they were getting |
| 21:32.25 | brlcad | some of that was stubbed in for sse2 |
| 21:32.53 | brlcad | include/vector*.h |
| 21:33.24 | starseeker | OK, so we will need to do it ourselves then |
| 21:34.31 | brlcad | pretty much |
| 21:34.38 | starseeker | needs to dig into this "create a tree of bounding boxes" step |
| 21:34.53 | starseeker | not clear why the performance was so bad on the openbook model |
| 21:34.56 | brlcad | the closest you could probably get as a general system would be something like opencl, but it's not clear what that would imply dependencywise |
| 21:35.23 | brlcad | would think you'd need to understand that before making a decision on how to improve it ;) |
| 21:35.34 | starseeker | exactly |
| 21:36.02 | starseeker | you're referring to Khronos Group's OpenCL? |
| 21:36.42 | brlcad | yes |
| 21:36.53 | CIA-28 | BRL-CAD: 03brlcad * r34328 10/brlcad/trunk/include/raytrace.h: more consistency cleanup |
| 21:37.04 | ``Erik | hronos is still around? |
| 21:37.35 | brlcad | very much so |
| 21:37.37 | starseeker | well, at least the license looks reasonable |
| 21:38.00 | brlcad | have a pretty big presense at siggraph each year |
| 21:38.15 | ``Erik | wow, I thought they'd died off in the late 90's |
| 21:38.56 | brlcad | opencl is also slated to be included in Mac OS X 10.6 so there are expectations that library may take off in a big way |
| 21:40.02 | brlcad | starseeker: dependencywise, I meant more what they take to be managed as a dependency, like build issues, how well it actually works, whether it's a special compilation pass or simply a library link, it's longevity and portability, etc |
| 21:40.14 | brlcad | it'd certainly require algorithm/code rewrites regardless |
| 21:40.16 | starseeker | ah |
| 21:51.29 | elena | i'll go now. thank you all for your help. bye. |
| 21:51.48 | *** part/#brlcad elena (n=ebautu@89.136.118.141) | |
| 21:55.03 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-67.sbndin.btas.verizon.net) | |
| 22:29.53 | brlcad | was just the victim of a hit n' run |
| 22:30.21 | madant | brlcad: seriously ? |
| 22:30.27 | brlcad | dammit, really need to have video cameras installed |
| 22:30.29 | brlcad | madant: yeah |
| 22:30.47 | madant | damn |
| 22:30.59 | brlcad | minor damage, but seriously infuriating |
| 22:31.05 | brlcad | as I think it was intentional |
| 22:31.15 | madant | you are ok right ? |
| 22:31.28 | brlcad | yeah, it's to the car, not me |
| 22:31.41 | madant | aah. ok.. |
| 22:32.21 | madant | no way to find out what happened or who did it ? |
| 22:32.24 | brlcad | I heard a large truck backing up (the sort that beeps) and then take off fast, which made me look out the window .. and I saw the damage |
| 22:33.25 | madant | hmm. that sucks .. insurance would cover the repairs etc. right ? |
| 22:33.31 | brlcad | yeah |
| 22:33.56 | madant | some people are just evil :D |
| 22:34.25 | madant | just woke up all of a sudden at 4 am |
| 22:38.30 | madant | would the police be able to track them ? you are filing a complaint right ? i have no idea how the system works in US . |
| 22:42.43 | Ralith | starseeker: that must have been what I read to lodge the idea in the first place |
| 22:43.19 | Ralith | brlcad: I'm kind of sad that that sort of error's mostly absent in the modern world of software |
| 22:43.44 | Ralith | yeah, it's only really practical when the user understands and can intuit what's wrong anyway, but it just sounds fun. |
| 22:44.44 | brlcad | madant: already filed and have the report, but nothing they can really do about it given no eye witness |
| 22:45.24 | Ralith | brlcad: oh damn :/ |
| 22:45.29 | Ralith | glad the damage is minor. |
| 22:45.34 | Ralith | and insured. |
| 22:53.33 | madant | hmm. yeah you should implement the camera idea.. |
| 23:00.21 | brlcad | Ralith: what sort of error? |
| 23:00.43 | brlcad | a little out of context and not thinking clearly at the moment to follow that response |
| 23:01.31 | Ralith | haiku 404s |
| 23:01.36 | Ralith | er, DNS failures |
| 23:01.39 | Ralith | or whatever that error was. |
| 23:01.52 | Ralith | netpositive's |
| 23:02.21 | brlcad | ahh |
| 23:02.28 | brlcad | okay, yeah, that makes more sense now :) |
| 23:03.13 | brlcad | it was generally an alternative (or in addition) to returning "Error: 404 Page not found" .. which is arguably just as cryptic :) |
| 23:06.51 | Ralith | yeah, but familiar, and standard. |
| 23:07.00 | Ralith | in addition would be cool. |
| 23:07.16 | Ralith | is of the opinion that the software world would be a nicer place if humor abounded more freely in released products |
| 23:41.10 | starseeker | Ralith: you mean Windows isn't a joke? |
| 23:41.26 | Ralith | :P |
| 23:41.43 | Ralith | it is, but sarcasm so extreme that it fools most of the world isn't my kind of humor. |
| 23:41.51 | starseeker | brlcad: crap - sorry to hear about the car :-( |
| 23:41.58 | starseeker | Ralith: ah |
| 23:42.17 | Ralith | "Oh yeah, *this* is a real OS! It's *exactly* what you need!" |
| 23:44.44 | brlcad | eep, /tmp: out of inodes |
| 23:44.55 | Ralith | that sounds bad |
| 23:48.19 | brlcad | yeah.. especially since I don't think it's 'actually' out of inodes |
| 23:48.42 | Ralith | :| |
| 23:48.45 | Ralith | that sounds *really* bad then |
| 23:49.47 | brlcad | hmm.. that's not a root filesystem though.. might be able to fix that one |
| 23:50.51 | Ralith | well, it is /tmp; couldn't you just boot a livecd and nuke it? |
| 23:51.56 | brlcad | eh, this is a dedicated server sitting down in florida |
| 23:52.01 | brlcad | maybe after a 20 hour drive :) |
| 23:52.13 | madant | heh :P |
| 23:55.19 | Ralith | that does indeed make things harder. |
| 23:55.38 | Ralith | perhaps some cleverness with a temporary symlink to a dir on the root fs while you unmount and nuke it? |
| 23:56.21 | Ralith | I can imagine lots of things going wrong in the instant between deletion and symlink creation, though |
| 23:56.33 | Ralith | then again, they're probably going wrong anyway if the FS is borked. |