IRC log for #brlcad on 20080609

00:17.54 *** join/#brlcad andrecastelo (n=chatzill@189.71.15.124)
00:19.19 andrecastelo howdy brlcad, ``Erik and pacman87
00:20.03 pacman87 hi andrecastelo
00:20.09 poolio I resent that andrecastelo ;)
00:20.26 andrecastelo howdy poolio ! heheh :D
00:55.42 yukonbob evening, cadheads
01:20.55 pacman87 hi yukonbob
01:41.31 yukonbob hey pacman87 -- how're things?
02:05.19 PrezKennedy howdy brlcad!
02:26.05 *** join/#brlcad ChanServ (ChanServ@services.)
02:26.05 *** mode/#brlcad [+o ChanServ] by irc.freenode.net
03:20.00 *** join/#brlcad pacman87 (n=timothy@71.170.63.120) [NETSPLIT VICTIM]
04:10.28 *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1177872280.dsl.bell.ca)
04:11.45 *** join/#brlcad pacman87 (n=timothy@71.170.63.120) [NETSPLIT VICTIM]
04:26.33 *** join/#brlcad pacman87 (n=timothy@71.170.63.120) [NETSPLIT VICTIM]
04:26.37 *** part/#brlcad pacman87 (n=timothy@71.170.63.120)
04:26.37 *** join/#brlcad pacman87 (n=timothy@71.170.63.120)
04:26.55 *** part/#brlcad pacman87 (n=timothy@71.170.63.120)
04:27.04 *** join/#brlcad pacman87 (n=timothy@71.170.63.120)
04:37.34 brlcad ahh
04:38.52 starseeker notices the science overlay has an ebuild for salome
04:41.59 brlcad starseeker: are you on the news mailing list?
04:43.04 *** join/#brlcad pacman87 (n=timothy@71.170.63.120) [NETSPLIT VICTIM]
04:43.04 brlcad or better yet, has anyone received a message today from the brlcad-news mailing list?
04:43.05 starseeker checks...
04:43.13 *** join/#brlcad pacman87 (n=timothy@71.170.63.120) [NETSPLIT VICTIM]
04:43.19 starseeker no, just the devel list
04:43.20 brlcad pacman87: welcome back :)
04:43.23 starseeker moves to correct this...
04:43.25 brlcad hrmph
04:43.34 brlcad oh, you're only on devel, got it
04:45.05 starseeker is subscribed now, but doubts that helps with the immediate problem...
04:46.17 starseeker chuckles as gentoo pulls opencascade from freebsd.org
04:47.40 starseeker brlcad: IIRC the opencascade license is somewhat odd?
04:48.05 *** join/#brlcad dtidrow (n=dtidrow@c-69-255-182-248.hsd1.va.comcast.net)
04:50.33 brlcad yeah, rather horribly worded license
04:51.00 brlcad somewhat sketchy terms
04:51.32 starseeker wonders if they ever publicly justified the reasons for the one-off license?
04:55.16 starseeker gets "welcome to brlcad-news" email, if that helps any...
04:56.29 *** join/#brlcad homovulgaris (n=cb6ef51f@bz.bzflag.bz)
04:57.15 *** join/#brlcad pacman87 (n=timothy@71.170.63.120) [NETSPLIT VICTIM]
04:57.15 homovulgaris hi Sean, u there ?
04:57.22 brlcad nope
04:57.37 homovulgaris huh :O
04:57.52 brlcad ~ask
04:57.53 ibot hmm... ask is Questions in the channel should be specific, informative, complete, concise, and on-topic. Don't ask if you can ask a question first. Don't ask if a person is there; just ask what you intended to ask them. Better questions more frequently yield better answers. We are all here voluntarily or against our will.
04:58.09 starseeker he must have been killed and replaced by the AI bot he was working on...
04:58.19 starseeker :-P
04:58.24 homovulgaris :P
04:58.37 homovulgaris anyways.. i have been living without internet for around 5 days now..
04:58.40 homovulgaris it is a disaster
04:58.58 starseeker ow
05:00.12 brlcad apparently so
05:00.33 *** join/#brlcad homovulgaris (n=cb6ef51f@bz.bzflag.bz)
05:01.14 brlcad well sorry to hear that.. glad to see you were able to find the intarwebs eventually ;)
05:01.39 homovulgaris brlcad: i will commit tomorrow hopefully when the network comes up :(
05:02.10 homovulgaris literally worst five days i have had in a long time .. maybe i am a net addict :O --> self-realization
05:02.36 homovulgaris i dont have all the data i have been working with but i think i can still discuss it if you are free
05:03.24 homovulgaris basically i started a new directory libpc which contains most of the code, a new pc.h header containing the structures in include
05:03.54 homovulgaris inside librt i started a new file pc_constraint.c which implement pc_constraint_export and pc_constraint_import
05:04.03 homovulgaris the import part i havent written
05:04.42 homovulgaris then i have made a simple mk_constraint function for now inside libpc
05:05.43 brlcad now libpc instead of libpg?
05:05.49 homovulgaris Inside libpc i have also defined generate_parameters function , the idea of which basically is.. if you pass a directory pointer to it , it generates a parameter set of all possible handles
05:06.26 homovulgaris yeah i thought parametrics and constraints are significant .. i mean Parametrics & Constraints rather than parametric geometry
05:07.30 homovulgaris and a write_parameter_set function which basically writes the parameter set as an (attribute value) array to the db using update_attributes
05:07.46 homovulgaris i hope i am not messing up any of the function names from memory :P
05:07.59 brlcad it all sounds good and well without seeing the code structure itself ;)
05:08.10 homovulgaris :) i know
05:08.55 homovulgaris hey sean, regarding librt integration.. the way i have implemented it presently is , so that we still use the librt functions to write stuff to db
05:08.59 brlcad the exact function names aren't significant at this point other than consistently prefixing the calls with the library prefix
05:09.25 brlcad that's good
05:09.36 brlcad librt should be the gatekeeper for all db io
05:09.59 homovulgaris i am finding using the attribute system for storing parameters a little bit too verbose
05:11.03 brlcad the question we'd talked about earlier was whether to create a new non-geometric db object or using attributes
05:11.10 homovulgaris but i am not able to think of a better way to handle it without extending the database system to a new version or defining a new object type just to hold the "handles" or parameters of the existing geometry object
05:11.21 brlcad no doubt that using attributes is more verbose
05:12.00 brlcad i thought we'd settled on using a new db object actually after the last discussion
05:12.11 homovulgaris for storing the equations or constraints we do need a non-geometric object which i have defined.
05:12.25 brlcad okay, then what are you using attributes for then?
05:12.55 homovulgaris for each geometric type there are a certain list of variables or parameters u can change right.. for denoting them
05:14.20 homovulgaris like for example in the case of a sphere ( radius,center-x,center-y,center-z) whether they are parametric or explicit, is there some domain restriction ( 0<radius < fastf_t_MAX )
05:15.26 homovulgaris adding a new object for each and every variable of a geometric object seems overkill
05:16.24 brlcad there are implicitly a list of vars/params.. it'll take a bit of effort to make that list explicit through a generic interface
05:16.36 homovulgaris so i was thinking we could use the attribute system for storing that data.. since in a sense it is the attribute of the sphere after all. but the verbosity is making me think about shifting that infomration as well to an object type somehow
05:17.04 brlcad that list of variables/parameters is in the library itself, not persisted to the db
05:17.41 homovulgaris didnt get it
05:18.33 brlcad so say you want to make a sphere .. it has radius and position values that are stored already
05:18.41 homovulgaris ok
05:18.42 brlcad that much doesn't need to change
05:18.50 homovulgaris right
05:19.10 brlcad but say now we want to add a constraint that the radius must be > 100mm
05:19.24 homovulgaris ok
05:19.55 brlcad that requires knowledge that "radius" is something that can be constrained on a sphere, which is something the sphere primitive itself needs to declare
05:20.28 brlcad as part of librt, i.e. src/librt/primitives/sph would have some logic that says that radius and position are two things that can be parameterized
05:20.53 brlcad now we need to create the actual constraint
05:21.07 homovulgaris yeah that is the ideal way to do it :) so we should change the geometry description u men
05:21.10 brlcad for that, there's a new db constraint object
05:21.54 brlcad that db constraint object will refer to something like sph.radius or some other defined nomenclature for saying it applies to a given sphere's radius
05:22.00 homovulgaris if the sphere is able to declare the fact that it has two variables for example , then things should be simple
05:22.14 homovulgaris the question is, how do we represent which of the parameters are variable or parametric and which are constant
05:22.22 brlcad I think it needs to
05:22.41 homovulgaris like in this case we need to say that only radius is the only variable and position is not
05:22.57 homovulgaris that data needs to be stored in db right
05:23.43 brlcad saying that radius is variable and position is sound like two constraints to me
05:23.58 brlcad s/constraints/parameters/
05:25.11 brlcad so by default, all parameters defined by the primitive could be changed like they can now -- you'd create a constraint object to restrict that further (e.g. radius > 100 && position == immutable)
05:25.22 homovulgaris ok so we put two constraints ( position fixed) and (radius >100) ?
05:25.30 brlcad yeah
05:25.44 homovulgaris hmm
05:25.58 homovulgaris this way we wont need generate_parameters and similar functions
05:26.27 brlcad hasn't seen the code, so agrees :)
05:26.55 homovulgaris :D
05:27.14 brlcad now what isn't clear, maybe you've sorted out already, is how to create generic equations that can be applied
05:27.54 brlcad and stored/referenced in the db as distinct generic function objects
05:28.06 brlcad or if it's even worth dealing with that right now
05:28.14 homovulgaris to store the equations i was thinking of using a stack with postfix or prefix which refers to these parameters.
05:28.46 homovulgaris basically for the whole solution part i was planning on using c++
05:29.37 brlcad that should be fine, you'll just have to have a well-defined abstraction into librt so you don't infect it type-wise :)
05:29.54 homovulgaris hopefully :)
05:30.10 brlcad necessarily
05:30.20 homovulgaris obviously :)
05:30.46 homovulgaris regarding storage
05:31.09 homovulgaris what if i just store the stack as a string in the .g file ?
05:31.40 homovulgaris and when you import it, it generates the constraint_network graph
05:31.49 brlcad that should be fine
05:32.04 brlcad especially since the objects are referred to by string references necessarily anyways
05:32.06 homovulgaris the edges in the graph are these equations ( like sph.1_r>1000)
05:32.28 brlcad there may be value, though, in storing the operators and values in binary
05:32.34 brlcad especially the values
05:32.45 brlcad for precision
05:33.08 homovulgaris hmm.. so names i store as string and values and operators i convert to binary
05:33.51 brlcad yeah, probably ideal
05:35.24 brlcad maybe something like STR sph.1_r INT 1000 OP >
05:35.33 brlcad where str, int, and op are some binary code
05:36.11 brlcad so you push values onto a ministack when importing, and perform operations as needed for each equation
05:36.49 brlcad I don't mean that exact string, only "sph.1_r" would be a chars in that example, the rest would be network encoded
05:36.50 homovulgaris for representing the solution of any individual parameter i was thinking of struct parameter_soln { int n_domains; struct domains * d[] } ; where struct domain { double min; double max}
05:37.13 homovulgaris makes sense :)
05:38.13 *** join/#brlcad pacman87 (n=timothy@71.170.63.120) [NETSPLIT VICTIM]
05:38.36 homovulgaris hi pacman87
05:38.55 brlcad a fixed array of domains?
05:40.29 *** join/#brlcad homovulgaris (n=cb6ef51f@bz.bzflag.bz)
05:40.53 homovulgaris well a dynamic array of domains
05:41.01 brlcad you could use a simple bu_list of domains
05:41.05 *** join/#brlcad clock_ (n=clock@217-162-111-232.dclient.hispeed.ch)
05:41.33 homovulgaris bu_list is a linked list right ?
05:41.38 brlcad they're a pretty simple container
05:41.43 brlcad yeah, doubly linked lists
05:41.48 homovulgaris ok.
05:42.02 homovulgaris i still have to figure out how to solve equations like r<1000
05:42.05 brlcad you create your own data type (i.e. struct domain), and just add bu_list as the first element
05:42.25 brlcad that automatically makes it a linkable structure
05:42.33 homovulgaris kewl :)
05:42.48 brlcad there's an example in bu.h
05:43.40 homovulgaris ok
05:43.58 brlcad you're basically implementing a simple lamba calculus for evaluating the expressions
05:44.21 brlcad the same evaluation that lets you do a == b should work for c > d, etc
05:44.39 brlcad giving truth values for each result that tell if the parameters hold
05:45.15 brlcad at least that handles all constraints, evaluating equations is a bit different of course
05:46.09 homovulgaris hmm.. ( need to do more reading :))
05:47.55 homovulgaris pacman87: sorry, but my svn is 6 days old :) did u rename the tentry in table.c to lesser than 8 characters. i was getting a compiler warning, initializer string too long.. you are the one wokring on hyperboloid right ?
05:48.28 brlcad homovulgaris: it's since changed, it's less
05:49.01 homovulgaris and then again there is the case of generic equations which can themselves be complex
05:49.06 brlcad it was increased, breaking binary compatibility, then reverted/renamed to something less than 8 chars to avoid that (and it wasn't necessary)
05:49.12 brlcad it's a 'hyp' primitive now
05:49.28 homovulgaris like 2nd degree polynomials in 2 variables and so on.
05:49.36 homovulgaris oh.. i realy need to update :)
05:50.40 homovulgaris pacman87: the uv mapping looks ultra cool :)
05:52.38 homovulgaris i will go think about how to make geometry declare their variables to the world :)
05:52.45 brlcad very cool: http://brlcad.org/gallery/s/renderings/primitives/hyp.png.html :)
05:53.47 homovulgaris see you tonight again hopefully on my computer hopefully :)
05:54.59 brlcad i think each primitive's list of their variables is close to the class of entity types you listed
05:55.38 homovulgaris hmmm.. true
05:56.59 brlcad points, lines, planes, axes, corners, circles, curves
05:58.30 brlcad e.g. sphere probably only has a point and an axis, whereas an arb8 would have 8 points, 6 planes, 12 lines, and an axis
06:00.45 brlcad was talking with a professor from IIT Kanpur just last week at a conference
06:01.09 homovulgaris hehe.. was he a nice guy ?
06:01.20 brlcad oh sure
06:01.29 brlcad it's a small highly technical conference
06:01.47 homovulgaris on cad ?
06:01.51 brlcad we talked for quite a while about open source
06:01.53 brlcad and gsoc
06:02.02 homovulgaris :) nice
06:02.04 brlcad on solid and physical modeling
06:02.44 homovulgaris what is his name ?
06:03.08 brlcad i actually don't remember atm, I'd have to look up his paper
06:03.39 homovulgaris :) lots of good professors in india at IITs
06:03.50 homovulgaris lots of not so good ones also .. but still :D
06:05.01 brlcad mm.. would be really interesting to have the primitives also define their constraints, and then use the library to ensure they're held
06:05.20 brlcad that might even make a good first step before trying to sort out the db io
06:05.44 brlcad since all primitives had implicit constraints
06:05.45 homovulgaris hmmm.. define their constraints in the sense the self constraints right ?
06:05.50 brlcad right
06:06.03 homovulgaris like for the sphere ? the implicit restrictions on radius ?
06:06.05 brlcad a sphere can't have a negative or infinite radius, for example
06:06.46 brlcad and more specifically, since the parameters for a sphere are technically the same as that of an ellipsoid, there's an A and B vector for parameters
06:07.07 homovulgaris basically this declaration would involve adding new functions to each geometry type right
06:07.08 brlcad and the constraint is that |A| == |B| for it to be a sphere
06:07.40 homovulgaris i meant the declaration of constraints and parameters to the world
06:07.58 homovulgaris like new functions in g_ell or g_arb8
06:08.02 brlcad one callback function that returns a table should suffice
06:08.17 homovulgaris yeah new function sorry :)
06:09.56 homovulgaris and where would the code which calls the library to ensre the constraint holding be located ?
06:35.04 CIA-21 BRL-CAD: 03brlcad * r31347 10/brlcad/trunk/src/librt/primitives/hyp/hyp.c: add some basic sanity checks on the uv range, mapping is still slightly off with a rotation offset on the u values and incorrect count
07:31.30 *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch)
07:38.23 brlcad hmms and thinks thinks he just needs to keep going
09:01.13 *** join/#brlcad docelic (n=docelic@78.134.207.155)
09:26.52 *** join/#brlcad docelic_ (n=docelic@78.134.205.144)
09:34.48 brlcad quad charts suck
09:35.02 clock_ brlcad: what's a quad chart?
09:35.36 brlcad an imposed format that tries to categorize information into 4 specific bins
09:35.43 clock_ lol
09:35.51 clock_ some kind of mind mapping tool?
09:36.07 brlcad nah, just business junk
09:36.08 clock_ categorizes information into one specific dustbin
09:36.18 clock_ brlcad: :)
09:40.01 *** join/#brlcad elite01 (n=elite01@dslb-088-071-034-129.pools.arcor-ip.net)
09:47.15 *** join/#brlcad mafm (n=mafm@elnet-111.lip.pt)
09:47.40 mafm hello
09:52.12 brlcad howdy mafm
09:52.20 brlcad buenos dias!
10:00.27 mafm hi
10:00.32 mafm you have strange sleep habits, don't you? :D
10:00.41 poolio mornin'
10:00.41 clock_ brlcad: you can speak spanish?
10:00.49 poolio brlcad: way to ... sleep
10:01.56 clock_ A friend is designing his own beds, lamps etc. and using some commercial software I don't remember the name
10:02.12 clock_ Should I suggest BRL-CAD? Does it run under Windows good enough for an end user as well?
10:03.29 brlcad mafm: heh, you could say that ..
10:04.12 brlcad doesn't generally like sleeping
10:05.10 clock_ finds falling asleep boring
10:05.14 mafm I don't either, but I have to
10:06.21 brlcad nonsense! just have to train your body to expect less :)
10:07.00 brlcad body is exceptionally resiliant and adaptive if you force it
10:07.32 brlcad clock_: it runs just fine under windows -- if they're as smart and adaptive as you are, then sure
10:07.46 clock_ they are smart but not sure if they are adaptive :)
10:07.53 brlcad in general though, you are a bit exceptional as for a modeler getting up to speed quickly and adapting to your needs
10:08.14 clock_ they use windows and they paid for commercial program maybe they will want to find psychological reasons to not admit their bad financial decision :)
10:08.51 clock_ brlcad: you mean I am exceptional in adapting to BRL-CAD? Ordinary user would advance slower?
10:33.50 *** join/#brlcad hippieindamakin (i=cbc85f82@gateway/web/ajax/mibbit.com/x-fca99bf11cf4a3d0)
10:33.59 hippieindamakin hey guys
10:34.12 clock_ hippieindamakin: hey
10:34.33 hippieindamakin may i have the link to the svn of brlcad java trunk ?
10:36.21 clock_ hippieindamakin: BRLCAD in java?
10:36.54 hippieindamakin yeah there are some programs written in java man
10:37.33 hippieindamakin wanted to get a reference and for some reason sourceforge doesnt openup
10:39.21 hippieindamakin nevermind i reached there..
10:41.29 hippieindamakin clock_: http://brlcad.svn.sourceforge.net/viewvc/brlcad/jbrlcad/ this is wat i meant
10:41.53 mafm brlcad: I've already had problems for not getting enough sleep
10:41.59 brlcad clock_: something like that, you're pretty much the ideal/intended user type that dives in, isn't afraid to script tools, can tie things together, etc
10:42.26 brlcad hippieindamakin: just like the main repo, but jbrlcad instead of brlcad
10:42.28 brlcad ~cadsvn
10:42.31 ibot To obtain BRL-CAD from Subversion: svn checkout https://brlcad.svn.sourceforge.net/svnroot/brlcad/brlcad/trunk brlcad
10:42.48 brlcad so this: https://brlcad.svn.sourceforge.net/svnroot/brlcad/jbrlcad/trunk jbrlcad
10:42.53 hippieindamakin hey mafm and brlcad
10:44.00 mafm hi hippieindamakin
11:08.10 *** join/#brlcad d_rossberg (n=rossberg@bz.bzflag.bz)
11:13.08 starseeker has finally encountered something worse than BRL-CAD for gentoo ebuild creation...
11:13.19 starseeker stares in awe at the SALOME gentoo bug history...
12:03.48 *** join/#brlcad thing0 (n=ric@123.208.245.43)
12:14.13 *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6)
12:26.41 *** join/#brlcad mafm (n=mafm@elnet-111.lip.pt) [NETSPLIT VICTIM]
12:35.10 *** join/#brlcad mafm (n=mafm@elnet-111.lip.pt)
12:44.16 clock_ The dating all boils down to the looks, and the looks boil down to the gym ;-)
12:44.28 clock_ brlcad: was it you who attends gym or was it ``Erik?
12:57.55 mafm uh
13:12.42 *** join/#brlcad thing0 (n=ric@58.171.180.96)
13:19.04 *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch)
13:26.29 *** join/#brlcad Elperion (n=Bary@p54877F6F.dip.t-dialin.net)
13:37.18 *** join/#brlcad dtidrow (n=dtidrow@c-69-255-182-248.hsd1.va.comcast.net)
13:47.57 ``Erik brlcad is the gym monkey, my idea of exercise is the 12oz curl :D
14:07.18 *** join/#brlcad Axman6_ (i=root@210-9-141-161.netspeed.com.au)
14:07.54 *** join/#brlcad Axman6_ (i=Axman@210-9-141-161.netspeed.com.au)
15:04.19 pacman87 morning, all
15:04.28 mafm morning
15:35.18 ``Erik neato, bzflag has shadows, tracks, and textures now O.o
15:35.38 pacman87 ``Erik: um... what was the last version you tried?
15:35.59 ``Erik d'no, about 5 years ago
15:36.09 pacman87 that explains it ;)
15:36.53 ``Erik dodging is difficult with no 'strafe' type motion
15:37.26 pacman87 what server are you on?
15:37.36 ``Erik and being able to aim seperate from trvel would be neat (like keys for movement and mouse for turret)
15:37.47 ``Erik uh, some 'BRL.something' one?
15:37.59 ``Erik desert something-or-rather
15:38.44 ``Erik i was dicking with macports, noticed it in the games dir, so *shrug* threw it on and jumped on one... couldn't find the name of one on brlcad's machine
15:50.29 *** join/#brlcad prasad_ (n=psilva@h-67-103-183-185.mclnva23.covad.net)
16:19.00 *** join/#brlcad Axman6_ (n=Axman6@210-9-140-165.netspeed.com.au)
16:35.34 mafm brlcad: when I submit next patches, should be the whole file or related with the patches yesterday?
18:04.18 mafm ping!
18:15.26 mafm omfg I killed the channel!!!!111!!!
18:17.47 louipc pong
18:19.26 mafm :)
18:52.06 *** join/#brlcad PrezKennnedy (i=Matthew@208.43.126.194)
18:59.43 pacman87 do the triangular faces for tess() need to be (counter)clockwise, or does the direction not matter?
19:27.30 *** join/#brlcad clock_ (n=clock@77-56-91-99.dclient.hispeed.ch)
20:33.16 mafm @brlcad and maybe other gsoc mentors: I'm waiting on what to do with my patches to either create new patches based on them, or submit new ones with the full content of the files, etc
20:33.21 mafm in the meantime: http://wainu.ii.uned.es/~mafm/brlcad/brlcad_rbgui_20080609-1.png
20:33.57 mafm heading home now, see you tomorrow
20:42.46 *** join/#brlcad docelic (n=docelic@78.134.205.144)
20:46.19 *** join/#brlcad docelic_ (n=docelic@78.134.204.94)
22:24.16 poolio ``Erik: where were you today? :)
22:26.54 *** join/#brlcad CIA-21 (n=CIA@208.69.182.149)
22:41.42 *** join/#brlcad CIA-22 (n=CIA@208.69.182.149)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.