irclog2html for #brlcad on 20050115

00:50.34 starseeker burp
01:07.27 a_b_normal the official thing for brlcad is bork
01:07.30 a_b_normal not burp
01:07.32 a_b_normal or poke
01:07.34 a_b_normal bork
01:07.40 a_b_normal :)
01:07.57 a_b_normal :(
01:10.43 starseeker Heh - just had dinner
01:10.52 starseeker hence burp :-)
01:11.43 starseeker learner: I don't suppose you guys are going to try and con some university engineering departments into funding work on brl-cad?
01:14.31 learner heh, funding? not directly anytime soon
01:14.38 learner perhaps indirectly through student contributions
01:15.05 learner brl-cad's been used in graphics and CAD classes at several universities over the years
01:15.12 learner it's particularly well suited to computer graphics classes
01:15.23 learner adding a new primitive, for example .. fun
01:15.32 starseeker Cool. Ah, so it's already got a good enough UI for the classes. Well, there goes that idea...
01:15.55 learner teaches CSG better than most any
01:16.05 starseeker Oh, I was going to ask you - you said the "wire mesh" 2D display didn't make much sense for brl-cad? or somebody did?
01:16.05 learner since most modelers have gone the brep path
01:16.14 starseeker brep?
01:16.27 learner boundary representation modeling
01:16.30 starseeker ah.
01:17.10 learner wire mesh 2d display? .. i said something related to that, but not really that :)
01:17.25 starseeker Oh - what did you say? Sorry
01:17.33 learner it's not well suited (yet) for drafting and CAM work until someone adds more functionality
01:17.46 starseeker Oh, OK.
01:17.56 starseeker But it can in theory be added?
01:18.00 learner so things like autocad sketchings don't really come over to solid modeling very well
01:18.10 learner sure, it can be added pretty easily
01:18.20 starseeker cool. I smell a grad student project ;-)
01:18.44 learner we just have no use for it in our domain, so it was never added
01:19.00 starseeker hehe - AutoCAD doomed by students - film at eleven
01:20.22 starseeker Although I guess Pro/E and Solidworks would be the more useful functionality to target?
01:21.20 learner better markets.. though drafting for cam purposes is interesting
01:22.38 starseeker In some ways it's too bad Solidworks appeared - with $70k a seat for something like Pro/E, it might have been rather easy to get companies to develop brl-cad
01:23.29 learner i don't mind the competition.. it was needed for the industry
01:23.55 learner though solidworks was just a union of several other cad companies trying to consolidate for profit and to fight the other guys
01:24.05 starseeker Ah.
01:25.01 starseeker Still, it would have been fun to see the Pro/E guys' reaction to the price for CAD suddenly dropping from $70k to 0
01:25.07 learner I want to take BRL-CAD in more directions than they are able to, though, and out-innovate them with new features
01:25.14 learner model life inside out
01:25.24 starseeker life ~= chaos
01:25.31 learner multi-user collaborations
01:26.13 starseeker Well, if there is any justice at all you should be swarmed
01:26.19 learner there's a lot of solidification and cumulation of existing features into a unified framework needed for starters, though
01:27.05 starseeker So you guys are still funded for continued development of brl-cad?
01:27.59 learner indeed
01:28.18 starseeker awesome
01:28.27 starseeker that puts you FAR ahead of most open source projects
01:28.37 learner this was never an abandonware event ala our other government friends
01:29.11 starseeker Incredible. That in itself is worth a news story
01:34.09 CIA-3 BRL-CAD: 03brlcad * 10brlcad/NEWS: special thanks regarding gentoo ebuild
01:35.18 starseeker uh - you're welcome? Did it work for you?
01:35.45 learner just preparing the sources for when i get to that step before I forget
01:35.59 starseeker Ah :-)
01:36.05 starseeker What's CIA-3?
01:37.30 learner CIA announces source code changes in real time to irc channels
01:37.41 starseeker Cool!!
01:37.45 learner also archives changes and calculates commit statistics to a web site
01:37.52 starseeker Very cool
01:38.17 learner friend of mine set that up.. most projects on freenode use it
01:38.30 learner cia.navi.cx
01:38.51 learner gentoo's been busy today :)
01:39.26 learner looks like the day just turned
01:39.40 starseeker how's that?
01:41.22 CIA-3 BRL-CAD: 03brlcad * 10brlcad/autogen.sh: oops, that dash wasn't intended or desired for head/tail
01:41.28 learner http://cia.navi.cx/stats/project/gentoo
01:41.33 starseeker Ah.
01:41.56 starseeker lordy lu
01:42.13 starseeker them keyboards are a-clickin
01:42.51 learner <PROTECTED>
01:42.58 learner for all projects
01:43.09 learner sometimes busy, sometimes not
01:43.54 starseeker Heh - looks like the US contingent is mostly out partying or something ;-)
01:44.00 starseeker whoop, there we go
01:44.13 learner it's pretty quiet for a friday night actually..
01:44.28 learner probably battlestar galactica premiere to be honest..
01:44.34 starseeker Hehehe
01:44.44 starseeker Ah, the joys of not having a TV...
01:45.02 learner ahh
01:45.08 learner i gave up tv for many many years
01:45.37 starseeker Heh - my girlfriend has mine, because the car was too full to fit the bugger when I moved.
01:45.37 learner now it's an occasional even for movies, stargate, farscape, and simpsons ;)
01:45.55 starseeker :-) The classics
01:46.21 starseeker Curiously enough, I'm not really a Simpson's fan, but I've got three of the four seasons of Futurama and want the fourth
01:47.01 learner good stuff
01:47.35 starseeker Yes, I've never figured out how Fox can produce such good cartoons and such bad news
01:47.44 starseeker it's like a split personality or something
01:48.50 starseeker Uh-oh, another openoffice-ximian commit today I see. Great, there goes fourteen hours or so of CPU time
01:49.12 starseeker It'll have to wait till after brl-cad 7.0.4 though
01:49.51 learner heh
01:50.41 starseeker I never, never read the change logs for openoffice-ximian or kde. I really don't want to know I just rebuilt everything for a minor update of something only needed on another platform or in rare useage
01:51.07 learner :)
01:51.44 starseeker :-) any new flags?
01:51.59 learner several
01:52.11 starseeker who regularly gets his butt eaten for lunch
01:52.16 learner wings is probably the most impacting
01:52.27 starseeker Yay :-)
01:52.31 learner you can turn while mid air and if the server is configured flap endlessly
01:52.39 learner so.. fly around in your tank
01:52.49 starseeker and die from GM :-)
01:53.14 learner several new bad flags
01:53.22 starseeker Hmm - brl-cad, used to model tanks. bzflag, tank combat game. Coincidence? ;-)
01:53.33 learner ;)
01:53.49 starseeker Funny, a lot of the people I run into on bzflag are gentoo guys
01:53.54 learner they really want me to write a g2bzw :)
01:54.07 learner could make brl-cad bzflag's primary modeler
01:54.15 starseeker sweet!
01:54.48 starseeker I don't suppose there is an armor flag that can bounce a few of the enemy's shots back at them?
01:59.33 learner heh
02:00.07 learner that's a harder one.. woul need means to graphically show that.. like a shield sphere
02:00.26 learner still easy.. but not a 10 minute tweak
02:00.45 starseeker true :-) Don't worry, that was just wishful thinking
02:01.28 learner careful what you wish for :)
02:03.03 starseeker Hehe. Uh oh
02:06.49 starseeker Well, as long as I'm making trouble - I've always thought a Targeter flag would be fun - press T or Enter or something and you are instantly flipped to be facing (i.e. targeting) your nearest opponent.
02:07.12 learner heh
02:07.18 starseeker But I think that one would be hard because the game would need to know your nearest opponent, check for stealth blocking, etc.
02:08.27 starseeker Sorry, wandering off topic
02:08.42 learner heh, don't worry about topic :)
02:12.04 starseeker Gotta admit though, I always like the old standby - Rapid Fire :-)
02:12.23 learner tis a powerful flag
02:12.35 starseeker particularly since it messes with people's timing
02:12.43 learner which in part inspired one of the new bad flags.. trigger happy
02:12.50 starseeker Uh oh :-)
02:12.51 learner tank unloads all your shots as fast as it can
02:12.54 starseeker hahaha
02:12.57 starseeker sweet
02:13.21 learner like running around with an uzi set to "on"
02:14.01 starseeker Gosh - that would turn that one 100 shot map into sheer madness
02:16.45 starseeker You know, a "Rigged" flag might be fun - if you shoot you kill yourself, but your explosion has a chance to take out anybody within SW radius - friend or foe
02:17.06 starseeker Alright, I should stop eating whatever I ate this evening
02:17.20 learner multiflags and configurable flags are what I'm planning next
02:17.29 starseeker COOOL :-)
02:17.46 learner so you could have shield.. with 10 shots for example
02:17.56 learner or laser and wings
02:18.02 learner etc
02:18.13 starseeker On wings, can you only shoot in the plain at which you are hovering?
02:18.24 learner have to disallow various combinations that would be too powerful like stealth+cloak
02:18.30 starseeker No kidding
02:18.41 starseeker Only Seer would stand a chance
02:19.31 starseeker You guys must be so sick of hearing flag ideas ;-)
02:19.40 learner shots always go out on the plane/elevation you shoot from unless inertial shots are turned on
02:19.50 learner so if you're falling, your shots do too
02:19.58 learner heh
02:20.12 learner yeah, we get a lot of flag suggeestions
02:21.15 starseeker I must say you guys have done a tremendous job
02:21.55 learner some past ideas: http://bzflag.org/wiki/FlagList
02:24.01 starseeker Should I go ahead and throw mine on there if they aren't already up?
02:24.09 learner sure, go ahead
02:26.09 starseeker Heh - too bad they said no to the dumb missle
02:29.01 learner some of the "no" were very contingent on the descriptions
02:29.13 learner and could be considerable under some situations
02:29.29 starseeker Ah. So the "never" are the big no-nos
02:30.01 learner yeah
02:30.10 learner the simple no's are often "i don't like that idea"
02:30.20 starseeker Ah :-)
02:30.48 learner there have been various shockwave blast round flags that have come up and been deemed possible
02:31.01 starseeker How does it work for server setup - if people want to add "custom" flags, do they need to do it at the source level?
02:31.42 learner you have to modify both the server and the client to add flags
02:31.56 starseeker Ah.
02:31.57 learner configuration of existing flags can be done any time
02:32.06 learner like making shockwave enormous and slow
02:32.15 starseeker :-)
02:33.02 starseeker I don't suppose it would be of interest to have a server "select" flag options on startup, and have the client download them? Or does the engine need modification for some flags to work?
02:34.48 learner they need to work together
02:34.55 learner otherwise there are protocol and versioning issues
02:35.04 starseeker Ah. Figured.
02:42.11 starseeker Heh - you guys need a bugzilla for flags
02:43.17 starseeker Fair number of duplicates
02:43.44 learner yeah, we only glance at it every few months when someone's motivated to actually add a flag
02:44.06 starseeker Do patches get more consideration that wiki suggestions?
02:44.14 learner there's so much other stuff that needs to get taken care of that the flags themselves are relatively low priority
02:44.36 starseeker Heh - responsible game developers :-). Whoda figured? ;-)
02:44.41 learner oh heck yes.. if it's implemented it's reviewed a little more quickly usually
02:44.55 learner unless it's a flag suggestion that would have been a "no/never"
02:45.27 starseeker Hmm. Just out of curosity, why not add them and then let the server activate/deactivate them?
02:45.38 starseeker unless they're something offensive or some such
02:47.10 learner because that ends up being code that has to be verified across all the platforms, changed when interfaces change (which happens a lot), maintained, etc.. all takes up time and effort
02:47.22 starseeker Ah.
02:47.40 learner and if it's a feature that's "destructive" to the nature of the game.. that's baggage that one generally doesn't want to carry
02:47.44 starseeker As you can tell, I don't do much game development ;-)
02:47.51 starseeker definitely true
02:48.06 learner you can't take everyone's idea .. simply too many various ideas
02:48.37 starseeker Very true
02:48.53 learner some would have it become quake, some would have it become pong, some would have it become a web browser.. with as wide a user base as it has, there are too many ideas so culling has to happen
02:49.00 starseeker Well, that's what open source is all about ;-)
02:49.14 learner yep
02:49.24 learner we hear enough people clammor for a feature.. it usually gets added
02:49.41 starseeker Sorry to bother ya with flag suggestions ;-)
02:50.45 learner nah, no harm in suggesting ;)
02:50.50 learner feedback is good
02:51.07 starseeker Clearly y'all know exactly what you're doing - you're the #1 open source game by a long shot
02:51.13 learner much holds true for brl-cad too .. there's lots of directions it can go now
02:52.05 starseeker hehe - there I'm pretty much TOTALLY unqualified. I'm not even a computer professional or design engineer
02:52.34 learner with a codebase as large as brl-cad, it's hard to go down a wrong path, though :)
02:53.02 starseeker Hehe.
02:53.12 learner only things it can't do without maintaining backwards support is the ray-intersection engine behavior in the raytrace library
02:53.45 starseeker in brl-cad?
02:54.04 starseeker I thought that was pretty stable
02:54.08 starseeker that part at least
02:54.16 learner it's very stable
02:54.37 learner that's what I mean.. new devs can't go modifying the way the raytracing behaves wontonly
02:54.43 starseeker Ah.
02:54.48 learner since it _must_ remain stable
02:54.54 starseeker I doubt they'll want to
02:55.13 learner it's used in many analytic codes that rely on it's behavior
02:55.28 starseeker The whole point of BRL-CAD is that it does a lot of the hard parts already
02:55.41 learner oh, you might want to modify the raytrace engine to make the pictures more pretty -- it can happen, just have to be careful to maintain certain aspects of behavior
02:56.21 starseeker Why not just make an export to povray? Use the internal render for "quick" stuff and send it to povray for the final "pretty" picture?
02:56.55 starseeker or would making things "prettier" involve fundamental changes?
02:57.12 learner i'd LOVE to see a g2pov and pov2g
02:58.10 starseeker I suppose it's a little harder than just writing the internal objects to povray format, and vice versa?
02:58.21 learner not really
02:58.42 learner would be one of the most simple export/imports .. pov supports many of the same primitives
02:58.48 learner and suports csg
02:59.09 starseeker Hmm. Can the brl-cad ray-tracer handle surfaces and transparency? (/me has vague recollections that these are two of the hard parts)
02:59.49 starseeker I mean surface textures
02:59.50 starseeker duh
03:01.15 learner what about textures?
03:01.33 learner brl-cad regions can do texture mapping
03:01.44 starseeker OK, so that'll map
03:01.47 starseeker cool.
03:02.21 starseeker Maybe once I get my thesis wrapped up I can take a look at a povray import/export as a "beginner" project
03:06.02 learner mm. what's the thesis on?
03:06.45 starseeker Behavior of CuIn(S,Se)2 thin films near compositional ratios of Cu/In ~= 1
03:06.56 starseeker possibly relevant to solar cells
03:08.19 starseeker Heh - I am sort of a weird person to be into all the open source science and engineering software - it stems mainly from my physics undergraduate days
03:09.21 starseeker Finishing a thesis seems to be the hard part of it, but once it's done I will hopefully have weekends free and clear to do open source coding
03:09.49 starseeker And perhaps then I can be of some real use to my favorite projects, and learn how to do Real Coding ;-)
03:11.18 learner hehe, interesting
03:12.01 starseeker Poor Tim Daly can't figure out why he's had a thousand downloads or more of Axiom - he's always wondering who those people are and what they are doing
03:12.27 starseeker I suppose a fair few are like me - not doing the hard science but checking out interesting software
03:12.33 learner so you've studied physics, materiels science, chemistry, etc
03:13.09 starseeker Not so much chemistry - my undergrad was physics with a focus on High Energy, my Masters was/is Materials Science with a focus on Thin Film Solar Cells.
03:13.29 learner got it
03:14.01 learner so how long before we can see solar efficiency reach 50%? ;)
03:14.04 starseeker My current job involves printer drum coatings, so that's the most chemistry oriented of the three
03:14.14 starseeker Look for flying pigs :-P
03:14.54 starseeker 50% would require several junctions and fine control over the band gaps - hard stuff
03:15.02 starseeker and currently expensive
03:15.49 starseeker Dunno. Tim probably would. Let me see if he's in IRC...
03:16.01 starseeker Snort. Not tonight. (naturally)
03:16.56 starseeker So how many honest to goodness engineers have you heard from so far, learner?
03:17.05 learner or even better.. a generic polynomial root solver -- even a simple linear-based newtonian estimation if it were fast and stable
03:17.39 learner engineers of what sort?
03:18.00 starseeker people who actually use programs like BRL-CAD to do real work :-)
03:18.31 learner ahh, that number is fairly large and active already
03:18.44 starseeker Oh, OK :-)
03:18.49 starseeker Good.
03:19.09 learner i've heard from 2 or 3 new interests since open source directly
03:19.19 starseeker Very good!
03:19.49 starseeker Interest is always key for open source projects, although in your case I guess you're already pretty well set up
03:20.37 learner there's probably about a dozen critical active codes more or less that use brl-cad today already for engineering analyses
03:20.42 starseeker I have no idea what the license is, but would this be of any interest? http://www.crbond.com/roots.htm
03:20.52 learner there have been hundreds cumulatively over the years.. but interest (and funds) come and go
03:21.25 starseeker Yep, know what that's like. (Solar cells + Bush administration ~= drought :-(
03:21.46 learner possibly.. people have various definitions of "high performance" :)
03:22.35 starseeker Probably the place to ask is the Maxima and Axiom lists. It can't hurt, and maybe the'll mention brl-cad to some friends in engineering ;-)
03:22.44 learner causes the reverse in our field, of course ;)
03:23.23 starseeker Well, at least someone benefits
03:25.03 learner that root solver page might be of use actually..
03:25.09 starseeker Cool!
03:25.48 learner if their bairstow implementation is stable, that would do well in replacing brl-cad's newtonian iterative solution
03:26.05 learner and it might make my superellipsoid primitive behave..
03:26.29 learner so it could become official
03:26.45 starseeker Cool! Did they stick a license in the zip file?
03:27.14 learner doesn't matter.. there are papers posted
03:27.32 learner i can implement off those if the license isn't suitable
03:27.35 *** join/#brlcad PrezKennedy (~Matthew@130.85.255.85)
03:27.48 starseeker :-). Life is good for smart people
03:29.25 learner you're researching thin film transmissive properties for solar energy and you don't consider yourself "smart people"? :)
03:29.26 starseeker It looks like GNU Octave may also have an implimentation of bairstow: http://academic.wsc.edu/faculty/jebauer1/numerical.html
03:29.56 learner that one's less interesting
03:30.15 learner it's a high level interface first, usually implying that it's not performance tuned
03:30.20 learner brl-cad's existing is fast..
03:30.31 learner another solver would need to be fast and more accurate
03:30.47 starseeker learner: Nah - the smart people are the ones who can actually model the materials, or understand why they're too complicated to model.
03:30.55 starseeker learner: Ah, OK.
03:31.21 starseeker If you ever want a headache, check out some papers by a guy called Zunger on CuInSe2
03:31.28 starseeker That's some impressive work.
03:32.50 starseeker My work was definitely more experimental - I saw something weird. I can't say what's causing it at a fundamental level
03:33.36 starseeker Experimentalists spot it. Then we give it to theorists to do the hard part ;-)
03:36.53 *** join/#brlcad noyb (~noyb@wbar2.lax1-4-8-213-215.dsl-verizon.net)
03:38.40 learner starseeker, that's cool still
03:39.37 starseeker thanks :-) It's a neat material growth effect, but it's ultimately something to avoid
03:39.57 starseeker the low band gap layer forms at the top, which isn't what we're after at all ;-)
03:40.39 starseeker If the first site falls through for the roots, you might check out the GNU scientific libraries
03:42.05 starseeker http://sources.redhat.com/gsl/
03:42.07 learner there are other solver systems I've read up on and fast math libs abound (brl-cad's is fast)
03:42.26 learner gsl i've seen. it's interesting on some levels
03:42.34 learner no real good numbers on performance though
03:43.21 starseeker Ah. You might want to ask the usnet groups - other than that, I'm not sure where to look. Axiom and Maxima haven't worried about really fast numerical solving, so I'm not really up on it.
03:43.32 starseeker 'course, Axiom as a commercial program had NAG
03:44.12 learner asking won't really help.. just have to try it some day :)
03:44.22 learner everyone will say it's "fast"
03:44.28 starseeker :-)
03:45.00 starseeker I can add one more (remote) possibility to the list - pari-gp seems to have a good reputation generally. But I don't know much about its abilities
03:45.13 starseeker http://pari.math.u-bordeaux.fr/
03:46.06 learner stability of the numerics on high order polynomials is more important even than speed
03:46.07 noyb I also think fast is related to what job you're asking the software to do. generally, nothing could be fast at every job.
03:46.44 learner there are numerical instabilities in a steiner surface I implemented that will be hard to solve
03:46.55 learner exactly
03:47.00 noyb for learning matlab-like skills, I use 'octave' on my mac, installed via fink
03:49.25 starseeker Indeed - apparently, in some cases the fastest algorithm is dependant on what hardware you have
03:49.43 starseeker Well, time for sleep - have long drive tomorrow.
03:50.05 starseeker Thanks again for your work on brl-cad (and bzflag!) learner
03:50.20 learner aiight
03:50.23 learner see you later
05:04.59 *** part/#brlcad stafa77 (~stafa77@d47-69-48-142.nap.wideopenwest.com)
05:17.01 *** join/#brlcad noyb (~noyb@wbar2.lax1-4-8-213-215.dsl-verizon.net)
05:39.49 tjyang learner,are you asleep yet ?
06:34.39 *** join/#brlcad brlcad (~brlcad@brlcad.bronze.supporter.pdpc) [NETSPLIT VICTIM]
06:34.39 *** join/#brlcad noyb (~noyb@wbar2.lax1-4-8-213-215.dsl-verizon.net) [NETSPLIT VICTIM]
06:34.39 *** join/#brlcad CIA-3 (~CIA@to.je.spocco.com)
06:34.39 *** join/#brlcad guu (guu@myth.gibbscam.com) [NETSPLIT VICTIM]
06:34.39 *** join/#brlcad a_b_normal (point@portablehole.net) [NETSPLIT VICTIM]
06:34.39 *** join/#brlcad EricWilhelm (~ewilhelm@adsl-68-93-192-127.dsl.tpkaks.swbell.net) [NETSPLIT VICTIM]
06:34.40 *** join/#brlcad learner (~brlcad@brlcad.bronze.supporter.pdpc) [NETSPLIT VICTIM]
06:34.40 *** mode/#brlcad [+oo brlcad learner] by irc.freenode.net
07:18.45 *** join/#brlcad fgq (fgq@203-190-200-022.dial.usertools.net)
07:19.30 *** join/#brlcad fgq (fgq@203-190-200-022.dial.usertools.net)
07:46.19 learner hello fgq
07:56.31 fgq hi learner
09:54.35 *** join/#brlcad fgq (fgq@203-190-195-082.dial.usertools.net)
10:36.43 *** join/#brlcad CIA-5 (~CIA@to.je.spocco.com)
12:17.59 *** join/#brlcad CIA-3 (~CIA@to.je.spocco.com)
13:21.07 *** join/#brlcad tjyang (~user@c-67-175-74-12.client.comcast.net)
13:53.17 *** join/#brlcad CIA-3 (~CIA@to.je.spocco.com)
16:42.32 *** join/#brlcad PhantomBantam (~phantomba@dialup-4.238.85.181.Dial1.Providence1.Level3.net)
17:20.13 *** join/#brlcad noyb (~noyb@wbar2.lax1-4-8-213-215.dsl-verizon.net)
17:42.50 *** join/#brlcad CIA-2 (~CIA@to.je.spocco.com)
18:47.09 *** join/#brlcad polyspin (~polyspin@pcp02462412pcs.chrchv01.md.comcast.net)
19:54.06 *** join/#brlcad CIA-4 (~CIA@to.je.spocco.com)
19:58.54 *** join/#brlcad noyb (~noyb@wbar2.lax1-4-8-213-215.dsl-verizon.net)

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.