| 00:11.19 | CIA-9 | BRL-CAD: 03brlcad * 10brlcad/include/ (bn.h bu.h): have the headers include the headers they respectively require |
| 00:26.25 | *** join/#brlcad SWPadnos (n=Me@dsl245.esjtvtli.sover.net) |
| 01:34.29 | Maloeran | http://www.rayforce.net/galleon.png - The killer of ray-tracing acceleration structures |
| 01:34.47 | Maloeran | With all these wires and ropes in the way :) |
| 01:36.22 | Maloeran | That will look better with a touch of texturing... that and global illumination |
| 01:39.23 | Maloeran | I challenge these kd-tree people to get good performance in this scene, eheh |
| 01:53.46 | *** join/#brlcad dan_falck (n=danfalck@pool-71-111-76-8.ptldor.dsl-w.verizon.net) |
| 01:53.58 | brlcad | nice model |
| 01:54.26 | brlcad | where'd it come from? |
| 01:56.23 | Maloeran | A friend of mine made it, as I mentionned to Lee. Technically, I'm paying the guy |
| 01:56.51 | Maloeran | It's much more impressive with texturing and so on |
| 01:57.21 | Maloeran | http://mcarp.earthstorm.com/reel/rotter-lowe/index.php?showimage=rotter-lowe-070.jpg&screenwidth=1600 |
| 01:59.23 | brlcad | unlimited rights? willing to share? :) |
| 01:59.28 | brlcad | would make for a great test model |
| 01:59.52 | Maloeran | Exactly my idea, it's the perfect model for raytracing benchmarks |
| 01:59.59 | brlcad | is it a full on poly model only? |
| 02:00.01 | Maloeran | It will be available for non-commercial use soon |
| 02:00.32 | Maloeran | "Full on poly model only"? It's made of triangles exclusively, and all volumes are closed to run vulnerability analysis or whatever else you guys do |
| 02:02.25 | brlcad | yeah, a triangle/poly-only model |
| 02:02.38 | Maloeran | *nods* Right |
| 02:02.40 | brlcad | instead of say a patches model of some sort |
| 02:02.54 | brlcad | solid would have been my next question, cool |
| 02:03.38 | Maloeran | And it was made from real blueprints, so this should be nice. The author really is impressing me, he's a hobbist |
| 02:03.54 | Maloeran | Well, besides the pitiful amounts I pay him to tweak his models for my needs |
| 02:04.26 | brlcad | :) |
| 02:05.37 | brlcad | is he looking for work? I'd pay for some models myself |
| 02:06.03 | Maloeran | Oh, he definitely is, he has no contacts to get in the industry |
| 02:06.24 | Maloeran | Hence why I'm trying to promote his work a bit, because he's getting really good |
| 02:07.44 | Maloeran | If you are interested, his nickname is mcarp on efnet, otherwise mcarp@mcarp.org |
| 02:12.52 | brlcad | hangs out anywhere in particular? |
| 02:13.16 | Maloeran | On efnet's #asm, among the crazy assembly programmers like myself |
| 02:13.53 | brlcad | aiight |
| 02:15.07 | brlcad | well I am definitely interested, there's been a couple models I've wanted for several years |
| 02:16.01 | brlcad | though I would like to have it constructed of primitives and CSG so I can do comparisons |
| 02:16.33 | brlcad | either implicits or breps/nurbs, but CSG-driven design to match the original |
| 02:16.43 | Maloeran | Excellent news. I don't know if he's used to anything else than triangles as final product, but feel free to ask him |
| 02:17.00 | Maloeran | He certainly is used to nurbs, I wouldn't know about CSG though |
| 02:17.02 | brlcad | he's a student? |
| 02:17.15 | Maloeran | He's a ~40 years old tech support guy |
| 02:17.20 | brlcad | ah, cool |
| 02:18.41 | Maloeran | Feel free to share how it goes if you chat with him, I have been willing to get some attention on his skills for some time |
| 03:07.23 | Maloeran | Erik, the autoconf/automake stuff breaks when trying to compile a 32 bits executable on amd64. It appends a ton of -L paths but these are for 64 bits libraries |
| 03:08.38 | Maloeran | Ah, it forces -L/usr/lib/libSDL.so instead of letting ld find it by itself |
| 03:17.37 | *** join/#brlcad IriX64 (n=Who@bas3-sudbury98-1168055376.dsl.bell.ca) |
| 06:20.07 | *** join/#brlcad IriX64 (n=Who@bas3-sudbury98-1168050695.dsl.bell.ca) |
| 07:37.01 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) |
| 08:25.31 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) |
| 11:15.48 | *** join/#brlcad digitalfredy (n=digitalf@200.71.62.161) |
| 11:16.29 | *** join/#brlcad mcarp (n=mcarp@24-178-112-16.dhcp.crtn.ga.charter.com) |
| 13:44.21 | *** join/#brlcad Twingy (n=justin@74.92.144.217) |
| 15:36.25 | mcarp | does anyone have an estimate of how long it took to create the tank in this screen shot: |
| 15:36.26 | mcarp | http://sourceforge.net/dbimage.php?id=36174 |
| 15:37.17 | clock_ | mcarp: you mean how long it took to develop Stryker ICV? Or how long the guys in factory assemble it together? Or how long it took to make the model? Or how long it renders? |
| 15:37.31 | Maloeran | Make the model surely |
| 15:37.48 | mcarp | how long did it take to make the model from primitives in brl-cad |
| 15:38.35 | mcarp | approx man hours |
| 15:39.28 | clock_ | mcarp: I'm sure this is a classified information, the information that the information is classified is classified too, and by just thinking about it you are automatically becoming a terrorist and will be dispatched to Syria upon your next visit to an airport :) |
| 15:39.36 | Maloeran | There's a possibility it might have been provided by the company producing the vehicule. It was also rendered with ADRT which only deals with triangles, so perhaps it wasn't made of BRL-CAD CSG primitives |
| 15:39.52 | mcarp | cute |
| 15:40.28 | Maloeran | Are you on the brl-cad team or an user of the software, clock_? |
| 15:40.43 | clock_ | Maloeran: user |
| 15:41.23 | mcarp | i fail to understand how keeping the number of man hours to build that model secret has any relevence to anything in reality |
| 15:41.37 | Maloeran | Right. mcarp : brlcad or ``Erik are likely to know more |
| 15:41.58 | clock_ | mcarp: the whole idea o classified information doesn |
| 15:42.04 | clock_ | t have any relevance to reality either |
| 15:42.18 | clock_ | and also the one with terrorism ;-) |
| 15:42.23 | mcarp | incorrect. classified information protects us national security sir |
| 15:43.05 | clock_ | Much more people die on highways than from terrorism... |
| 15:43.16 | mcarp | point? |
| 15:43.36 | clock_ | it's just a masquerade |
| 15:43.57 | clock_ | Basically "we want power, but we have to hook it up on something to make it look plausible" |
| 15:44.25 | mcarp | that is irrelevent to my question of required man hours for a model of this complexity |
| 15:44.59 | clock_ | mcarp: I don't know it and I just wanted to push my irrelevant opinion :) |
| 15:58.11 | brlcad | mcarp: it's a hybrid model, put together in various systems and involves lots of people |
| 15:59.02 | brlcad | it's a mix of csg and poly, converted to full poly for that particular rendering for adrt |
| 15:59.39 | clock_ | brlcad: adrt is part of brlcad? |
| 15:59.56 | brlcad | it's shipped as part of brl-cad, yes |
| 16:00.16 | clock_ | brlcad: is rt for CSG and adrt for polygons? |
| 16:00.28 | brlcad | pretty self-contained codebase though, it could just as easily be it's own project (and is/was) |
| 16:01.34 | brlcad | clock_: pretty much, though it's more like adrt for triangles and rt for anything (csg, poly, brep, etc) |
| 16:01.48 | brlcad | adrt gets its performance by specializing on processing triangles quickly |
| 16:01.53 | clock_ | brlcad: is adrt faster or what is the advantage? And what does the ad mean? |
| 16:02.15 | brlcad | advanced distributed ray tracer |
| 16:02.32 | brlcad | adrt is faster processing triangles than rt is |
| 16:03.30 | brlcad | rt supports much more of a variety of geometry, lighting models, run-time options, cross-platform support, etc |
| 16:05.18 | mcarp | well as far as man hours on that model are we talking about weeks months what |
| 16:06.03 | brlcad | mcarp: the stryker model is also part of a much larger effort that involves several organizations, engineering teams, analysis teams all with different focus points and is a constantly changing model as new armor is tested, different systems are tried, etc |
| 16:07.02 | mcarp | im trying to get an idea what kind of workflow time wise the average modelers are getting in this |
| 16:07.32 | brlcad | as far as man hours, if you were to try to model what's already there and had all the design specifications (which isn't reasonably feasible), we'd be talking several weeks, maybe a couple months |
| 16:07.35 | clock_ | brlcad: how is it from moral point of view if someone doesn't want to contribute to arms race, but wants to work on GPL project? Is working on BRL-CAD moral or immoral then? |
| 16:07.58 | brlcad | and that estimate is using BRL-CAD or Unigraphics or Pro/E or Solidworks |
| 16:08.38 | brlcad | as it is, the model actually involves what's probably more like several man-years of effort as, again, it's a constantly changing model |
| 16:09.00 | brlcad | and brl-cad only touches one small aspect of it, the vulnerability/lethality analysis side |
| 16:09.03 | mcarp | are there any non-classified examples of high detail available |
| 16:09.25 | brlcad | mcarp: not of any of the good models |
| 16:09.30 | clock_ | brlcad: I did lethality analysis on Ronja, but the resulting lethality is very low :) |
| 16:10.12 | mcarp | as an independent contractor i have to get a grasp on some method of fee calculation |
| 16:10.19 | brlcad | i've been trying to get a simplified version of a particular russian vehicle released as open source for several years without success (a very well known model that you could probably go buy on the market if you really wanted) |
| 16:10.29 | mcarp | which is directly perportional to likely time spent on complex geometry |
| 16:11.26 | brlcad | mcarp: understood, it's a hard problem to estimate time on an unknown task |
| 16:11.44 | mcarp | havoc.g for instance crashes brlcad on raytrace so i cant get a good idea of the final results. castle.g will ratrace |
| 16:11.49 | mcarp | tank-car.g crashes |
| 16:11.58 | brlcad | clock_: i don't personally see contributing to brl-cad as contributing to an arms race |
| 16:12.13 | clock_ | brlcad: because anyone can use brl-cad for any purpose, right? |
| 16:12.14 | mcarp | is there anything about "draw *" that could contribute to a crash of ratrace attempts |
| 16:12.17 | brlcad | havoc.g crashes? .. that shouldnt' happen |
| 16:12.36 | brlcad | eek |
| 16:12.36 | clock_ | brlcad: and if the army pays some money for development, it will become weaker by the money removed, and arms race is actually slowed down :) |
| 16:12.43 | brlcad | yeah.. "draw *" is .. bad :) |
| 16:13.06 | brlcad | gets back to all of the expert knowledge that you have to know for basic operation ;) |
| 16:13.16 | mcarp | sure of course |
| 16:13.27 | brlcad | * expands to every object at every level of the model hierarchy |
| 16:13.31 | mcarp | what is the recommended level of drawing to get a view of the final results |
| 16:13.33 | clock_ | brlcad: we implemented a segfault recovery into Links - maybe you could borrow the code |
| 16:13.33 | brlcad | which is certainly not what you wanted |
| 16:13.51 | mcarp | right of course |
| 16:13.53 | clock_ | brlcad: we use it for the final demonstration as university project - to make sure it doesn't segfault in front of the comission :D |
| 16:13.53 | brlcad | start with "tops" to get a list of the top level objects, and then "draw some_object" |
| 16:14.15 | mcarp | isnt there a draw everything thats going to be looked at type command? |
| 16:14.47 | clock_ | brlcad: then the debugging was like "PerM, you should really do something with your javascript interpreter, my browser got already 7 segfaults..." |
| 16:15.16 | brlcad | no, brl-cad geometry files often contain multiple models, not just one like you often find in other modeling systems .. there is no "default" object to display in several instances |
| 16:15.39 | brlcad | we try to only have one in the public test databases, to minimize confusion, but that's not the common practice |
| 16:15.48 | brlcad | for havoc.g, the primary object is "havoc" |
| 16:16.19 | brlcad | clock_: what license is links? |
| 16:16.31 | *** join/#brlcad danfalck (n=dan@pool-71-111-76-8.ptldor.dsl-w.verizon.net) |
| 16:16.47 | Maloeran | Ah, I wish I had access to reasonably complex models myself ;), though I'm really fond of mcarp's superb frigate |
| 16:17.22 | Maloeran | I finally have a model with which I can work on collection of multiple intersection points per ray, segment construction for traversal of materials |
| 16:17.55 | brlcad | clock_: i also don't get the "it will become weaker by the money removed", you mean the army? that's hardly the case |
| 16:18.21 | mcarp | is there a method of discovery for what is the primary object |
| 16:18.22 | brlcad | governments are not corporations, don't act like them and certainly don't involve finances like them |
| 16:18.32 | brlcad | mcarp: yes, tops |
| 16:19.02 | mcarp | ah, sorry i misunderstood that |
| 16:19.02 | mcarp | ty |
| 16:19.02 | clock_ | brlcad: GPL |
| 16:19.02 | brlcad | it's a hierarchy (a directed acyclic graph), so there are only a few trees |
| 16:19.14 | brlcad | clock_: ah, then no use to me :) |
| 16:19.24 | clock_ | brlcad: why? BRL-CAD is BSD? |
| 16:19.31 | brlcad | need to remain lgpl/bsd clean |
| 16:19.43 | clock_ | brlcad: we could give you an exception |
| 16:19.56 | brlcad | brl-cad is a mix, different portions under different licenses |
| 16:20.04 | clock_ | brlcad: we already made exception for OpenSSL |
| 16:20.43 | brlcad | all of the GPL code is going to be converted to LGPL here shortly due to an issue with integrating the GPL portions |
| 16:21.12 | clock_ | GPL that's like a woman - issues, issues, issues... |
| 16:21.30 | mcarp | thats certainly a much improved display (havoc) as compared to castle.g or truck.g |
| 16:21.53 | mcarp | that being the only way to see the geometry final results is a big hit to workflow tho |
| 16:22.07 | brlcad | havoc is a fairly small model, but probably the most complex CSG example openly available |
| 16:22.35 | brlcad | rt -F/dev/ogll (assuming you're on Windows) will render into a window |
| 16:22.36 | clock_ | mcarp: if you want some example small models, http://ronja.twibright.com/3d |
| 16:22.43 | mcarp | i'm only seeing orthagonal view modes are there not full perspective views? |
| 16:23.19 | clock_ | mcarp: if you subject me to hypnosis I'll be maybe even recall how much manhours it was |
| 16:23.49 | brlcad | mcarp: hit "3", is that what you mean? otherwise, go to Misc->Perspective for a default high-degree perspective |
| 16:24.01 | mcarp | do you guys make bolts with threads or are all bolts non-threaded? |
| 16:24.04 | brlcad | using the shift-grips key bindings, you can spin/pan/zoom the view |
| 16:24.13 | brlcad | mcarp: depends entirely on the model |
| 16:25.37 | mcarp | yes ive had no problems using shift grips |
| 16:25.59 | mcarp | '3' appears to show an orthagonal view with no diminishing point |
| 16:26.02 | brlcad | for a vulnerability/lethality analysis, threads are generally not of interest on bolts |
| 16:26.21 | brlcad | mcarp: okay, just making sure you really meant perspective ;) |
| 16:26.31 | mcarp | yes i REALLY meant perspective |
| 16:26.35 | brlcad | you're looking for misc->perspective :) |
| 16:27.04 | mcarp | ah, better |
| 16:27.15 | mcarp | is this display ogl accellerated? |
| 16:27.20 | clock_ | mcarp: I do without threads because I don't know how to do with threads |
| 16:28.12 | brlcad | mcarp: depends which platform -- in general it usually is compiled with opengl support, though it's not really going to "accellerate" a wireframe view |
| 16:28.42 | brlcad | mged supports various "display managers", one of them is an opengl view |
| 16:29.19 | mcarp | i dont currently have available a bsd or linux system with useful graphics capability so i'm stuck in windows |
| 16:29.36 | mcarp | not that 'stuck in windows' is bad for me |
| 16:29.39 | brlcad | the windows binary should be opengl |
| 16:29.57 | brlcad | that doesn't get you shaded views, but that's an entirely different/long discussion |
| 16:30.43 | brlcad | implicit geometry vs explicit, design history, .. |
| 16:31.09 | mcarp | flat shaded in viewport would be a big problem? |
| 16:33.21 | brlcad | as mged is currently architected, it's a non-trivial problem |
| 16:33.53 | mcarp | its not non-trivial for visualization of what you're building |
| 16:34.04 | brlcad | brl-cad on the whole doesn't really care, but the geometry is fundamentally stored in implicit form which doesn't get you triangles to feed opengl |
| 16:34.42 | mcarp | so the primitives arent drawable per se in ogl? |
| 16:35.16 | brlcad | opengl *only* deals with triangles, so what ever you have, you have to convert to triangles to display via opengl |
| 16:35.34 | brlcad | depending on your base model geometry types, that involves different sorts of conversions |
| 16:36.12 | brlcad | they can be converted, all of them, and we do for some purposes (e.g. adrt is a triangle-only ray-tracer) |
| 16:37.03 | mcarp | has anyone there done any boat hulls? |
| 16:37.20 | brlcad | guys at the navy |
| 16:37.36 | mcarp | nurbs? |
| 16:38.22 | brlcad | there's no direct nurbs edit support in mged, only rendering support where the models come in through conversion or through programmatic creation |
| 16:38.58 | mcarp | how about suspension springs |
| 16:39.05 | mcarp | single helix |
| 16:39.10 | brlcad | oh sure, lots of those |
| 16:39.38 | mcarp | i havent gone deep enough into the tutorials to have a clue how you'd go about doing that with primitives |
| 16:40.00 | brlcad | those can be rather easily constructed using torus sections or using the pipe primitive (which is basically a path-swept sphere with constraints) |
| 16:40.00 | mcarp | which is why i asked about the bolt threads b4 |
| 16:40.51 | mcarp | ok, so there are some specialized deformative type parameters then |
| 16:41.13 | brlcad | csg construction methods don't "come naturally" if you're coming from different design philosophies |
| 16:41.47 | brlcad | it's a different way at looking at modeling problems |
| 16:41.51 | mcarp | well i think its not the csg issues that im in the dark about its the brl-cad issues that im really in the dark on |
| 16:42.02 | brlcad | some things lend themselves very nicely to CSG, others not so nicely depending on the primitive set |
| 16:42.14 | mcarp | sure its all about the primitive set |
| 16:42.28 | mcarp | just as with any system its all about what parameters can be applied |
| 16:45.43 | brlcad | yep |
| 16:46.11 | brlcad | some things with the brep method become complex, there certainly is a tradeoff in the techniques |
| 16:46.33 | brlcad | which would be why most systems have become rather highly hybrid (the CAD systems that is, MAX doesn't count) ;) |
| 16:47.37 | brlcad | solid modeling in general has a variety of topological constraints that aren't easily verifiable in non-CAD systems |
| 16:50.54 | mcarp | yeah thats completely obvious when you start talking about things like pullies and ropes reguarding overlap |
| 17:08.59 | mcarp | /dev/ogll could be much more efficient if it draws only invalidated regions rather than the entire window |
| 17:09.35 | mcarp | but i am able to ratrayce all the scenes i crashed originally with draw * |
| 17:10.24 | Maloeran | Anything better looking than the donut-wheeled truck? :) |
| 17:10.38 | mcarp | yeah havoc looks very nice |
| 17:10.48 | mcarp | tank_car is reasonable as well |
| 17:10.55 | Maloeran | *nods* Good |
| 17:11.01 | mcarp | i havent explored all the examples yet tho |
| 17:11.54 | mcarp | donut tires tho...*shakes head*....i wouldnt show those, not good for recruiting business |
| 17:13.10 | Maloeran | Ah! Donuts, is there anything they can't do? |
| 17:13.56 | brlcad | mcarp: none of the example models are intended for recruiting business .. can't generally provide the models that would be good for recruiting business |
| 17:14.10 | brlcad | they're mostly intended as test models for development |
| 17:14.18 | brlcad | made by non-modelers often too |
| 17:14.59 | brlcad | havoc is perhaps the only exception, and maybe the barbie jeep toy |
| 17:15.28 | mcarp | which database is that? |
| 17:15.53 | brlcad | the latter is a good example of what was possible after the introductory training and a couple hours modeling |
| 17:16.12 | brlcad | it's toyjeep.g |
| 17:16.25 | brlcad | probably not in your version, you'd have to pull it from cvs |
| 17:16.31 | mcarp | yeah its not here |
| 17:17.05 | brlcad | a *really* nice one that I like is the motorolla walkie-talkie that was used in the advanced modeling tutorial |
| 17:19.58 | brlcad | but i've not put that into cvs yet, but it's a nice modeler by one of the best modelers; it's got a lot of meta resources, textures, extrusions, and even a few articulations iirc |
| 17:20.21 | mcarp | okeydoke |
| 17:22.50 | brlcad | and instead of the "couple hours" investment in the toyjeep, it's got a couple weeks of care and detail put into it |
| 17:23.15 | mcarp | got a screen shot? |
| 17:23.23 | brlcad | mcarp: you can get toyjeep from here: http://brlcad.cvs.sourceforge.net/brlcad/brlcad/db/ |
| 17:23.35 | brlcad | download the .asc and run asc2g to convert it |
| 17:24.12 | mcarp | k |
| 17:25.23 | brlcad | there are a couple screenshots in http://ftp.brlcad.org/VolumeIII-Principles_of_Effective_Modeling.pdf |
| 17:25.35 | brlcad | though none of the talkie as a full model |
| 17:25.46 | mcarp | yeah im actually flipping through that document at the moment but havent gotten to it |
| 17:25.47 | mcarp | yet |
| 17:26.06 | brlcad | it just covers a piece of the antenna and the circuit board |
| 17:34.58 | mcarp | odd, raytrace lingering window exits when i try to raytrace that |
| 17:35.17 | mcarp | tops |
| 17:35.18 | mcarp | all/ |
| 17:35.20 | mcarp | draw all |
| 17:36.21 | mcarp | at least it didnt crash mged :P |
| 17:52.35 | mcarp | i can raytrace individual objects but there is something under all/ that cant be drawn |
| 17:54.00 | mcarp | or rather, cant be raytraced |
| 17:55.42 | brlcad | hmm |
| 17:56.13 | mcarp | lot of stuff under there i'd hate to have to figure out what object(s) shouldnt be drawn |
| 17:56.17 | mcarp | the only thing in tops is all/ |
| 17:57.38 | mcarp | ah |
| 17:57.39 | brlcad | it ray-traces just fine here, there may be some issue with the windows version, or asc2g on windows |
| 17:57.48 | mcarp | um, error! fabs(y)=0 |
| 17:58.19 | mcarp | bn_mat_inv: error! fabs(y)=0 |
| 17:58.55 | brlcad | yeah, something wrong with the model |
| 17:59.19 | brlcad | see if this .g works: http://ftp.brlcad.org/~sean/toyjeep.g |
| 17:59.40 | mcarp | a sec |
| 18:00.20 | mcarp | exact same behavior |
| 18:00.36 | mcarp | bn_mat_inv: error! fabs(y)=0 |
| 18:00.37 | mcarp | MATRIX singular matrix: |
| 18:00.37 | mcarp | <PROTECTED> |
| 18:00.37 | mcarp | <PROTECTED> |
| 18:00.37 | mcarp | <PROTECTED> |
| 18:00.39 | mcarp | bn_mat_inv: singular matrix |
| 18:00.58 | mcarp | quick fix suggestions? |
| 18:01.27 | brlcad | short of recompiling with the latest, not really |
| 18:01.37 | mcarp | k, not a problem i can live without seeing it |
| 18:01.44 | brlcad | that will require a debugging session |
| 18:01.46 | mcarp | shame tho |
| 18:01.53 | brlcad | good to know, but quite unexpected |
| 18:02.05 | mcarp | hey, if it can be broken i can break it |
| 18:02.08 | brlcad | the windows release hasn't been around for very long, that is the very first release |
| 18:03.15 | mcarp | i had a machine that could be available for bsd or linux but that hdd was ancient and has gone *poof* |
| 18:03.37 | mcarp | well ancient for maxtor *shrug* |
| 18:04.34 | mcarp | not to mention i'd hate to give up my dual display |
| 18:05.10 | mcarp | max + photoshop you need every available square inch of real estate you can get |
| 18:06.24 | mcarp | i dunno how ppl without dual displays live without them when editing uvw maps in max |
| 18:06.31 | mcarp | ok i do know cuz i used to |
| 18:07.26 | *** join/#brlcad IriX64 (n=IriX64@bas3-sudbury98-1168054347.dsl.bell.ca) |
| 18:30.09 | mcarp | drawing bl_main_power_line.r or bl_bl.s2 causes the raytracer to exit |
| 18:30.20 | mcarp | all other objects drawn, successful raytrace |
| 18:31.12 | mcarp | bbiaf |
| 18:43.26 | ``Erik | *yawn* |
| 18:55.08 | Maloeran | Eheh. Erik, mcarp is the cathedral & galleon modeller |
| 18:55.57 | Maloeran | Galleon which is quite a test for any raytracer acceleration structure |
| 18:56.44 | mcarp | back |
| 18:58.31 | ``Erik | cool, hi |
| 18:58.52 | mcarp | hi |
| 18:59.24 | ``Erik | <-- is Erik, not a modeller |
| 18:59.46 | mcarp | okeydoke |
| 18:59.53 | Maloeran | Erik, how am I supposed to compile 32 bits executables on amd64 with your autoconf/automake stuff? It tries to link against /usr/lib/libSDL.so which is 64 bits |
| 19:00.05 | Maloeran | Rather than letting ld find the proper library itself |
| 19:00.13 | ``Erik | I'm a model :D *pose* |
| 19:00.23 | brlcad | mcarp: thanks for the details, it's scheduled to get looked into |
| 19:00.24 | ``Erik | mal: that'd be an issue with sdl-config itself... :/ |
| 19:00.32 | mcarp | yw |
| 19:01.20 | Maloeran | No hint of solution then? Besides quickly putting together a working Makefile or even a bash script |
| 19:01.58 | ``Erik | um, I'm sure there is one... |
| 19:02.29 | Maloeran | If I just use -lSDL, ld will find the proper library by itself |
| 19:02.39 | ``Erik | ermmmmmmm |
| 19:02.50 | Maloeran | But the automake stuff tries to force a particular version by -L |
| 19:02.51 | ``Erik | you link with @SDL_LIBS@ ? |
| 19:03.18 | ``Erik | um, I'm kinda wondering if that's libtool doing it, not automake... |
| 19:03.18 | Maloeran | Yes |
| 19:03.50 | CIA-9 | BRL-CAD: 03brlcad * 10brlcad/BUGS: rendering toyjeep.g on Windows causes a crash on 'all', as well as bl_main_power_line.r or bl_bl.s2. bn_mat_inv() singular matrix error. bug found and reported by mcarp, thanks |
| 19:04.11 | ``Erik | try to echo ${SDL_LIBS} in configure.ac... or better yet, read what was printed on the screen to link, heh, should be two very very long lines, one that calls libtool with flags, the next being what libtool tries to execute |
| 19:04.38 | brlcad | ah, sdl-config doesn't know/care about 64bit iirc |
| 19:05.19 | Maloeran | gcc -std=gnu99 -I./RT -I/usr/include/SDL -D_REENTRANT -m32 -o .libs/rftest main.o env.o crudemesh.o rtirender.o rtirenderflat.o -Wl,-rpath -Wl,/usr/lib -L/usr/lib /usr/lib/libSDL.so /usr/lib/gcc/x86_64-pc-linux-gnu/3.4.5/libstdc++.so -L/usr/x86_64-pc-linux-gnu/bin -L/usr/x86_64-pc-linux-gnu/lib -L/usr/lib/gcc/x86_64-pc-linux-gnu/3.4.5 -L/usr/lib/gcc/x86_64-pc-linux-gnu/3.4.5/../../../../x86_64-pc-linux-gnu/lib/../lib -L/usr/lib/gcc/x86_64-pc-linux-gnu/3. |
| 19:05.23 | Maloeran | <PROTECTED> |
| 19:05.27 | mcarp | CIA-9: yessir |
| 19:06.20 | Maloeran | CIA-9 is a bot, but I'm sure he appreciates the acknowledgement of its existence :) |
| 19:06.25 | ``Erik | the line before it is what you care about, mal |
| 19:06.29 | mcarp | ah :/ |
| 19:06.39 | Maloeran | mkdir .libs |
| 19:06.46 | Maloeran | <PROTECTED> |
| 19:06.46 | mcarp | you might amend that to 'raytrace exit' doesnt crash brl-cad |
| 19:06.49 | ``Erik | should be like ../../libtool --mode=compile |
| 19:07.06 | ``Erik | ok, so it's libtool screwing the pooch |
| 19:07.12 | ``Erik | not automake or autoconf, hrm |
| 19:08.44 | ``Erik | damn my ribs hurt :/ |
| 19:09.14 | Maloeran | Too much boat rowing with Sean? |
| 19:09.20 | brlcad | heh |
| 19:10.08 | brlcad | you could add an --enable-64bit option of some sort and when enabled/detected do something like `sdl-config --libs | sed 's/lib/lib64/g'` |
| 19:10.38 | Maloeran | Compiling 64 bits is quite fine, it doesn't manage to build on 32 bits |
| 19:10.46 | Maloeran | Unless I compile on another box anyway |
| 19:11.03 | Maloeran | I maintain that we probably would have less problems with one Makefile per platform, it's simple and it works, but.. *sigh* |
| 19:11.06 | brlcad | what does sdl-config --libs output? |
| 19:11.24 | Maloeran | -L/usr/lib -Wl,-rpath,/usr/lib -lSDL -lpthread |
| 19:11.41 | Maloeran | It should be /usr/lib32/libSDL.so for 32 bits libraries |
| 19:12.28 | brlcad | ah, well similar 'trick' should do the trick |
| 19:12.59 | brlcad | ``Erik: cad crash? |
| 19:13.17 | ``Erik | huh? |
| 19:13.23 | Maloeran | Pricy BWM crash? |
| 19:13.27 | ``Erik | yeah :( |
| 19:13.50 | brlcad | just lost connectivity |
| 19:13.56 | ``Erik | http://math.missouristate.edu/~erik/m3/20060925/ (my camera sucks, I need a new one) |
| 19:14.17 | brlcad | system, not the software |
| 19:14.44 | Maloeran | Ahh! |
| 19:14.47 | Maloeran | What happened? |
| 19:15.05 | ``Erik | caught some grass, slid, hit a tree, rolled |
| 19:15.20 | Maloeran | This is horrible. |
| 19:15.26 | ``Erik | no shit? |
| 19:15.46 | brlcad | eek |
| 19:16.12 | Maloeran | I really wasn't aware of that. I hope you aren't too damaged?... |
| 19:16.45 | ``Erik | um, some bruises, stretched ligamints in my rib cage, banged up my elbow, and a minor concussion |
| 19:16.56 | brlcad | that sucks |
| 19:17.07 | ``Erik | yeah, need another car :/ |
| 19:17.07 | brlcad | when was that? |
| 19:17.10 | Maloeran | Ouch, ouch.. |
| 19:17.13 | ``Erik | saturday afternoon |
| 19:17.18 | brlcad | damn |
| 19:17.47 | Maloeran | Were you speeding at the time? |
| 19:18.41 | ``Erik | didn't think I was, but the cops seem to think I was a tiny bit over the limit |
| 19:19.13 | Maloeran | Could that be a problem for the car insurance? |
| 19:19.32 | ``Erik | so far, no? *shrug* I'm crossing my fingers and hoping I'll have a check on friday |
| 19:20.42 | Maloeran | I'm really sorry to hear this... |
| 19:21.08 | Maloeran | At least you came out of it mostly intact, it seems |
| 19:23.11 | Maloeran | Did you lose consciousness or just felt dizzy?.. |
| 19:23.20 | ``Erik | neither |
| 19:23.35 | ``Erik | rolled it, got out through the sunroof and was walking around |
| 19:23.59 | Maloeran | Ah, so it was quite minor, right |
| 19:24.46 | Maloeran | ( the concussion that is ) |
| 19:37.12 | dtidrow_work | banged your head good, in other words? |
| 19:37.14 | Maloeran | Were air bags helpful? I have often wondered how effective these cushions were |
| 19:37.14 | dtidrow_work | well, I suspect they don't do much in a rollover |
| 19:37.15 | Maloeran | They could prevent strain on the upper back, as the torso and head are being thrown around |
| 19:39.55 | ``Erik | I d'no, I hit a tree and rolled it and came out pretty decent, so I guess between the seat belt and airbag that deployed *shrug* |
| 19:40.32 | Maloeran | Yes, it really could have been worse |
| 19:42.43 | Maloeran | I wish you a good recovery, please take care of yourself Erik |
| 19:43.22 | ``Erik | heh, I'm trying to not get dead :) I think I came out better than you did with your bike accident |
| 19:43.29 | ``Erik | those bikes are horrible death traps ;) *duck* |
| 19:44.34 | Maloeran | :) Yes that's probably true, bicycles only get dangerous when there are motorized vehicules around |
| 19:44.57 | ``Erik | I d'no, I've heard about people taking tumbles on trails and stuff |
| 19:45.24 | mcarp | plus those derned helmets can muss yer hair |
| 19:45.45 | ``Erik | meh, programmers hair is always mussed up |
| 19:46.15 | mcarp | well sure but...thats beside the point of the inconvenience |
| 19:46.31 | Maloeran | I suppose so. I once rammed fairly hard in a tree to avoid a child on roller blade who fell just in front of me, but that car door sure was worse |
| 19:46.44 | mcarp | s/programmers/all computer nerds/ |
| 19:47.19 | ``Erik | damn nonprogrammer computer nerds, trying to get into the programmer geek clique... :D |
| 19:47.45 | mcarp | <---15 years x86 assembly booya |
| 19:47.59 | brlcad | heh |
| 19:47.59 | ``Erik | um, ok, so you're a masochist, too |
| 19:48.00 | ``Erik | :) |
| 19:48.06 | mcarp | correct |
| 19:48.33 | mcarp | add to that win32 usage :~ |
| 19:49.08 | ``Erik | I had an x86 15 yrs ago, but still used my commodores heavily |
| 19:49.13 | mcarp | better watch it, is that a mactel? |
| 19:49.17 | ``Erik | no |
| 19:49.20 | mcarp | lol |
| 19:49.29 | ``Erik | on a dual g5 right now, have a g4 powerbook and g3 ibook at home |
| 19:49.51 | mcarp | to each his own |
| 19:50.07 | ``Erik | but I tend to do crunch work on freebsd boxes with many opteron chips pushing |
| 19:50.38 | mcarp | hey, they're all computers i can live with it |
| 19:50.51 | Maloeran | The g4 powerbook was somewhat poor, half the ray-tracing crunch of a Sempron 1.8ghz laptop |
| 19:51.14 | mcarp | bummer |
| 19:51.27 | ``Erik | if you blow the pipeline and stall left and right, yeah... *cough* *cough* |
| 19:51.29 | mcarp | im using a sempron 2200 on a 1.5Ghz mb |
| 19:51.48 | Maloeran | Laptop or desktop? |
| 19:51.53 | mcarp | desk |
| 19:51.56 | mcarp | (asus) |
| 19:52.05 | Maloeran | Mmhm. Let's get you an upgrade soon :) |
| 19:52.35 | mcarp | doubt it will matter much considering the non-graphicly intense requirements of brl-cad |
| 19:52.36 | ``Erik | heh, pheer my 850mhz tbirt at home :) |
| 19:52.51 | ``Erik | erm, mged and 'rt' are pretty graphically intense |
| 19:52.56 | dtidrow_work | heh |
| 19:53.26 | mcarp | maybe when you get into 100s of 1000s of objects |
| 19:53.37 | ``Erik | small geometries hehehe |
| 19:53.46 | mcarp | im doing ok on 2million faces in max |
| 19:54.01 | mcarp | lags until cached, but otherwise not bad |
| 19:56.20 | Maloeran | Your galleon takes 200mb of RAM to ray-trace too, somewhat heavy |
| 19:57.04 | mcarp | its a good thing the capt of the roter lowe is not alive to hear you call his frigate a gallean or *slash slash* |
| 19:57.47 | Maloeran | :) Right right, frigate |
| 19:57.48 | mcarp | now of course spanish galleons were about that size |
| 19:59.14 | mcarp | Maloeran: build me a perspective viewport flat shader |
| 19:59.26 | ``Erik | http://www.pacific-tall-ships.com/Roterlo_files/Roterlo_1.htm pheer |
| 19:59.49 | mcarp | aye |
| 20:00.17 | mcarp | the site i got my drawings from misspelled roter-lowe to 'rotter-lowe' |
| 20:00.37 | mcarp | i was unable to google for images until i had nearly completed the model in its entirity :/ |
| 20:00.38 | ``Erik | http://www.historicships.com/TALLSHIPS/Mamoli/RoterLoweMV19.htm |
| 20:00.44 | mcarp | this was...a slight inconvenience |
| 20:01.19 | mcarp | yeah that last image was the one i used for flag reference |
| 20:01.43 | Maloeran | Quite cice. I still prefer a 3d model to a wood model |
| 20:01.47 | Maloeran | nice, even |
| 20:01.58 | mcarp | wood are quite satisfying to build tho |
| 20:02.35 | mcarp | it should be completely obvious that i made up my own (non-period) color scheme since i didnt have any images |
| 20:03.13 | mcarp | however i would have done a few things differently had i had access to google images at the time |
| 20:03.30 | mcarp | and i totally ignore that in the drawins 'roter-lowe' was spelled correctly :? |
| 20:03.32 | mcarp | :/ |
| 20:04.16 | mcarp | note the tripple block/tackle loading on the fore and main stays |
| 20:04.20 | mcarp | mine are singles :/ |
| 20:05.31 | Maloeran | Oh... Well, still something that could easily be fixed if you desire so. It's a great model in any case |
| 20:05.44 | mcarp | perhaps |
| 20:06.20 | mcarp | i also found another interesting fact (language gap) roter lowe == "red lion" |
| 20:07.13 | Maloeran | That explains the lion figure |
| 20:07.16 | mcarp | so you see, history of a particular subject is good to have |
| 20:07.22 | mcarp | indeed and that it is red |
| 20:07.49 | mcarp | all i had were black and white fuzzy line drawings |
| 20:09.00 | Maloeran | Don't worry about that :), it can be interesting to fill the missing pieces as you see them |
| 20:09.18 | mcarp | i almost wonder if the web site owner misspelled rotter-lowe to keep us from finding additional info :P |
| 20:09.52 | mcarp | well as you know it began as a personal fulfilment project so yeah, it doesnt matter that mauch |
| 20:09.54 | mcarp | much |
| 20:10.33 | *** join/#brlcad clock_ (i=clock@84-72-92-36.dclient.hispeed.ch) |
| 20:10.34 | mcarp | at least one thing is for sure, bitmaps are easy to switch out heh |
| 20:12.18 | Maloeran | Eheh yes. I might try swapping a few textures once I get to worry about eye-candy |
| 20:12.33 | mcarp | Maloeran: build me a perspective viewport flat shader |
| 20:12.50 | Maloeran | In BRL-CAD? |
| 20:13.02 | mcarp | begin spam repeat 2,000,000 |
| 20:13.13 | mcarp | of course |
| 20:13.36 | Maloeran | It's a bit early for that, I'm supposed to integrate my stuff in there in January or so |
| 20:13.53 | Maloeran | Until then, I can post crude screenshots :p http://www.rayforce.net/galleon01.png |
| 20:14.20 | mcarp | that wont do me anygood building csg :P |
| 20:16.48 | mcarp | whew, you need some normal smoothing |
| 20:18.17 | mcarp | hey, havent you downloaded the release version yet? |
| 20:19.01 | Maloeran | I did, not converted it yet. It's a bit low priority, I'm supposed to code and document a bunch of code for last friday |
| 20:19.52 | mcarp | ah |
| 20:38.07 | ``Erik | screw smoothed normals, just add more triangles, once they hit the subpixel range, it's all good |
| 20:38.07 | Maloeran | Nah. Screw triangles, use curved facets |
| 20:38.07 | ``Erik | heh |
| 20:38.07 | Maloeran | Triangles are just a special case of curved facets |
| 20:38.07 | ``Erik | and curved facets are a special case of b-splines, yadda yadda |
| 20:38.08 | ``Erik | somethin' about hyperdimensional manifolds, blah |
| 20:38.08 | ``Erik | which, of course, are all just a coarse approximation of a particle render |
| 20:38.10 | mcarp | and hey, we're all just collections of harmonic bumps in the quantum strata |
| 20:38.11 | ``Erik | "Today, a young man on acid realised that all matter is merely energy condensed to a slow vibration. That we are all one consciousness experiencing itself subjectively. There is no such thing as death, life is only a dream and we're the imagination of ourselves. Here's Tom with the weather." |
| 20:38.11 | ``Erik | -billhicks |
| 20:38.12 | mcarp | gasp |
| 20:38.13 | mcarp | i suggested to my brother a few weeks ago that after his "death" his particular electro-chemical brain pattern might become manifest in some distant part of the universe |
| 20:38.14 | mcarp | there was no reply |
| 20:38.14 | Maloeran | Neat, I think that's a more likely occurance than the existence of some god-guided after-life |
| 20:38.14 | mcarp | indeed mal |
| 20:38.14 | mcarp | altho for those who feel they need to believe in such things, i suppose its a matter of syntax |
| 20:39.31 | mcarp | bbl |
| 20:43.34 | *** join/#brlcad IriX64_ (n=IriX64@bas3-sudbury98-1168057179.dsl.bell.ca) |
| 20:57.56 | *** join/#brlcad dtidrow (n=dtidrow@c-69-255-182-248.hsd1.va.comcast.net) |
| 21:05.59 | *** join/#brlcad IriX64 (n=IriX64@bas3-sudbury98-1168056657.dsl.bell.ca) |
| 21:43.57 | *** join/#brlcad IriX64_ (n=IriX64@bas3-sudbury98-1168057671.dsl.bell.ca) |
| 21:56.33 | *** join/#brlcad IriX64__ (n=IriX64@bas3-sudbury98-1168057409.dsl.bell.ca) |
| 23:59.38 | brlcad | gah, damn opengl |
