| 00:17.15 | *** join/#brlcad ibot_ (i=ibot@rikers.org) | |
| 00:17.15 | *** topic/#brlcad is BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || for ken: http://www.mediawiki.org/wiki/Help%3ATemplates#Using_parameters_in_templates | |
| 00:33.14 | *** join/#brlcad SuperTaz (n=taz@adsl-69-210-240-223.dsl.chcgil.ameritech.net) [NETSPLIT VICTIM] | |
| 00:33.14 | *** join/#brlcad dtidrow_work (n=dtidrow@host169.objectsciences.com) [NETSPLIT VICTIM] | |
| 00:33.14 | *** join/#brlcad joevalleyfield (n=joevalle@bz.bzflag.bz) [NETSPLIT VICTIM] | |
| 00:48.29 | *** join/#brlcad dtidrow (n=dtidrow@c-69-255-182-248.hsd1.va.comcast.net) | |
| 02:03.40 | poolio | brlcad: Aha! |
| 02:04.16 | poolio | brlcad: when the point the ray is being shot from is inside the model, it includes the point outside the bounding box |
| 02:04.26 | poolio | I'm guessying you knew that though |
| 03:02.57 | *** join/#brlcad poolio (n=poolio@c-69-251-3-107.hsd1.md.comcast.net) | |
| 04:31.09 | brlcad | yes, the starting segment can be behind your starting hit point if you're inside something |
| 04:31.24 | brlcad | applications have to manually trim or back out accordingly |
| 04:31.51 | brlcad | and yes, quite intentional behavior :) |
| 04:47.20 | poolio | brlcad: so would it be acceptable to just truncate the negative inhit distance? |
| 04:49.21 | brlcad | I'm not sure I understand why you'd ever be shooting from inside the geometry |
| 04:50.02 | brlcad | otherwise, you sort of need that, gets to what I mentioned earlier about having your fitness use the union of their two bounding boxes |
| 04:50.42 | brlcad | that's a case where your sample bounding box is bigger than your input's so okay for that case I could see you might be inside, but that's where the union of the box becomes important for the fitness |
| 04:51.33 | poolio | well the issue is I have a fixed bounding box to make crossover work |
| 04:52.24 | poolio | And I'm shooting from inside the geometry because I have a fixed bounding box |
| 04:54.24 | poolio | an alternative would just be to say crop the bounding box if it was greater than the size of the source's bounding box |
| 04:54.45 | poolio | and grow other bounding boxes if they are too small |
| 04:55.08 | poolio | Actually I wouldn't need to do that |
| 04:55.59 | poolio | Just remember the grid spacing from the source, and a corner of the bounding box from the current object |
| 04:57.41 | poolio | Crap, but the issue with doing the ratio of say a ray's depth to its bounding box is that it means it messed up crossover in the z direction. if you cross two fit individuals the child could be terribly unfit because the depth of the objects could be so varied |
| 04:58.32 | poolio | so it looks like the best solution is to use a_ray_length and use the fixed coordinates for the rays for every single model |
| 05:03.53 | brlcad | you could reshoot the grid based on the union of both the genotype and the input, but that'd be horribly inefficient -- you should be able to shoot exactly the same rays you initially shot for the input as well as additional rays as needed if the model was bigger |
| 05:04.25 | brlcad | if that's too much, then just process your segments yourself clamping the starting and ending to be withing the original box |
| 05:05.26 | brlcad | I don't recall if a_ray_length will trim the segment or if it is like the starting segment such that it just guarantees there will not be a new segment encountered past that length (i.e. an entry) |
| 05:05.40 | poolio | ok |
| 05:06.27 | poolio | So I'm just going to clamp the input to 0, and the maximum to the mdl_max[Z] of the input |
| 05:06.42 | brlcad | probably easiest for now and should be fine |
| 05:07.02 | brlcad | could even try to guarantee that your sample geometry will not go outside those bounds either |
| 05:07.29 | brlcad | becomes really problematic once you add subtractions and unions, but it should be good enough for now |
| 05:08.55 | poolio | what do you mean trying not to guarantee that sample gemoetry will not go outside those bounds? |
| 05:09.45 | brlcad | well, why is your geometry outside the input bounding box? |
| 05:10.20 | poolio | well I see what you mean, but just cause it's outside the bounding box doesn't mean we don't want it |
| 05:10.29 | brlcad | exactly |
| 05:10.35 | brlcad | but you could clamp it if you wanted |
| 05:10.37 | poolio | As you said with subtractions and intersections, that shape could become vital |
| 05:10.38 | brlcad | at least for unions |
| 05:10.42 | poolio | true |
| 05:10.57 | poolio | but I mean it might be that our model is shapes like a hemisphere |
| 05:11.13 | poolio | in which case we want to allow the model to be outside of the bounding box |
| 05:11.27 | poolio | however, as I think I still havent done, shapes which don't intersect the bounding box are pointless |
| 05:12.00 | brlcad | to properly do subtractions and intersections, though, the genotype's bounding box is going to have to be accounted for when you shoot rays (by shooting additional rays) just for it to work |
| 05:12.05 | poolio | but ideally the GA would be able to magically sift those out :) |
| 05:12.25 | brlcad | yes, that's important .. all primtives must be enclosed by or intersect the input bounding box |
| 05:12.26 | poolio | wait, why would I need to shoot additional rays? |
| 05:13.31 | brlcad | well, if you did a cross-over with some shape that was bigger, and that intersected something else that was bigger, and the operation was an intersection -- there could be a whole chunk of material outside the bounds that you're shooting (that should otherwise lower the fitness because it's wrong) |
| 05:13.50 | brlcad | you could end up missing that geometry that was outside entirely, even though as primtivies, they actually overlap the input box |
| 05:15.43 | poolio | well i don't necessarily think that chunk should lower the fitness |
| 05:16.11 | poolio | and it actually would because of the tree depth |
| 05:16.18 | brlcad | you can probably go on the assumption that you're limiting your stuff to basic quadric surfaces so you won't encounter that case |
| 05:16.41 | brlcad | the depth could just be 0 or 1 still |
| 05:16.44 | poolio | well even with that case I feel like the GA will still be able to figure it out with the superflous material |
| 05:33.38 | brlcad | http://bzflag.bz/tmp/think_outside_the_box.png |
| 07:17.13 | *** join/#brlcad Laniakea (n=clock@zux221-122-143.adsl.green.ch) | |
| 07:19.29 | *** join/#brlcad tessier (n=treed@kernel-panic/sex-machines) | |
| 07:19.31 | tessier | Wow |
| 07:19.45 | tessier | I can't believe there is an actual brlcad community now! |
| 07:31.56 | *** join/#brlcad Laniakea_ (n=clock@zux221-122-143.adsl.green.ch) | |
| 08:22.49 | *** join/#brlcad Elperion (n=Bary@p54877C2E.dip.t-dialin.net) | |
| 09:39.17 | *** join/#brlcad elite01 (n=elite01@dslc-082-082-092-094.pools.arcor-ip.net) | |
| 12:52.08 | *** join/#brlcad dtidrow (n=dtidrow@c-69-255-182-248.hsd1.va.comcast.net) | |
| 12:53.26 | *** join/#brlcad SuperTaz (n=taz@adsl-69-210-240-223.dsl.chcgil.ameritech.net) [NETSPLIT VICTIM] | |
| 12:53.26 | *** join/#brlcad dtidrow_work (n=dtidrow@host169.objectsciences.com) [NETSPLIT VICTIM] | |
| 12:53.26 | *** join/#brlcad joevalleyfield (n=joevalle@bz.bzflag.bz) [NETSPLIT VICTIM] | |
| 12:54.09 | *** join/#brlcad jack- (i=jack@83.137.193.141) [NETSPLIT VICTIM] | |
| 12:54.09 | *** join/#brlcad Maloeran (n=maloeran@glvortex.net) [NETSPLIT VICTIM] | |
| 13:35.46 | *** join/#brlcad poolio (n=poolio@c-69-251-3-107.hsd1.md.comcast.net) | |
| 14:48.56 | LinuxMafia | hi |
| 14:49.12 | LinuxMafia | i am new in brlcad |
| 14:50.40 | *** join/#brlcad Laniakea (n=clock@zux221-122-143.adsl.green.ch) | |
| 14:50.49 | LinuxMafia | is there a way when i press 0 , make also the raytrace to move not only the shapes? |
| 14:50.59 | LinuxMafia | sorry press z* |
| 15:39.54 | poolio | brlcad: should I increase fitness for matching missed rays? |
| 16:10.17 | ``Erik | ��/blah |
| 16:15.06 | poolio | helo |
| 16:16.32 | Maloeran | http://www.rayforce.net/rf041.png - Weee... I'm not too sure how well that will compare with others at Siggraph |
| 16:18.53 | poolio | wait, you can render it at 14 fps? |
| 16:19.05 | Maloeran | Sure |
| 16:19.15 | Maloeran | ( With 3 moving lights ) |
| 16:19.50 | poolio | wow |
| 16:20.02 | poolio | you might even be able to run crysis ;) |
| 16:20.31 | brlcad | tessier: howdy |
| 16:21.03 | brlcad | LinuxMafia: what do you mean? If you're in an edit mode, then pressing z will edit the model, in a view mode and it just rotates the camera/view |
| 16:21.39 | LinuxMafia | brlcad, when i raytrace |
| 16:22.02 | LinuxMafia | i want the raytarced obj to rotate instead wireframe |
| 16:22.12 | brlcad | poolio: yes! .. that's why I just said "matching" earlier -- matching means matching both segments and matching air/misses correctly |
| 16:23.58 | poolio | also there were some bugs in my logic in the fitness routine, works better now, but the results aren't much better |
| 16:24.01 | LinuxMafia | brlcad, when i press z only wireframe rotates |
| 16:26.29 | brlcad | LinuxMafia: yes, because generally it takes a while to raytrace a given frame |
| 16:26.41 | brlcad | so it only does it on demand as you request it |
| 16:27.32 | poolio | eek, i'm at 75C :\ |
| 16:27.57 | LinuxMafia | brlcad, so can i rotate the raytraced obj ? (other than rotate wireframe and raytrace again) |
| 16:28.15 | brlcad | LinuxMafia: I'd gather you simply want "shaded displays", not necessarily caring whether it's ray-traced or not |
| 16:29.02 | LinuxMafia | brlcad, i just want to whatch it rotates beacause it will be so cool |
| 16:29.37 | brlcad | yes, I know |
| 16:29.46 | brlcad | you want shaded displays |
| 16:30.18 | brlcad | mged unfortunately doesn't do shaded displays, but that is part of the new modeling interface being worked on |
| 16:32.37 | LinuxMafia | brlcad, wow that would be cool , thanks alot |
| 16:41.27 | poolio | brlcad: ok maybe not...fitness function is still giving negative numbers |
| 16:42.15 | LinuxMafia | is there any undo key? |
| 16:43.23 | poolio | nope |
| 16:48.50 | LinuxMafia | thanks so if i did something wrong |
| 16:48.58 | LinuxMafia | no way back |
| 17:05.50 | LinuxMafia | how can i switch to edit mode and viw mode |
| 17:05.51 | LinuxMafia | ? |
| 17:09.03 | LinuxMafia | any one? |
| 17:09.27 | LinuxMafia | brlcad, are you here ? |
| 17:16.18 | *** join/#brlcad ibot (i=ibot@rikers.org) | |
| 17:16.18 | *** topic/#brlcad is BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || for ken: http://www.mediawiki.org/wiki/Help%3ATemplates#Using_parameters_in_templates | |
| 17:20.54 | *** join/#brlcad dtidrow (n=dtidrow@host131.objectsciences.com) | |
| 17:26.19 | *** join/#brlcad MinuteElectron (n=MinuteEl@silentflame/member/minuteelectron) | |
| 17:28.12 | LinuxMafia | no matter i found it |
| 17:28.59 | poolio | LinuxMafia: have you looked through the manuals on the website? |
| 17:29.32 | LinuxMafia | poolio, this is what i am reading , but i dont have patiant |
| 17:29.46 | LinuxMafia | poolio, i am reading the tutorial |
| 17:30.07 | brlcad | heh, not having patience is going to be difficult |
| 17:30.20 | brlcad | there is a LOT to take in and learn before you can be productive |
| 17:30.29 | LinuxMafia | poolio, one problem i have shift+alt+mouse does not work for me |
| 17:30.32 | brlcad | the tutorials are big, and they only scratch the surface |
| 17:31.43 | LinuxMafia | poolio, for example when i click shift+alt+middle mouse drag it is as same as shift + left mouse |
| 17:32.16 | LinuxMafia | it looks like Alt key not be seen |
| 17:33.43 | poolio | brlcad: I think my program is losing it's mind (or maybe I am) |
| 17:35.34 | LinuxMafia | wow i found a solution for that |
| 17:35.40 | LinuxMafia | :D |
| 18:40.16 | poolio | brlcad: I'm tempted to do some major code restructuring. The results just aren't making sense and I feel like it might be because of a buffer overflow I'm missing somewhere |
| 18:51.57 | poolio | brlcad: if I don't have it workin by monday would you have the time to look through it and hopefully make it work? |
| 18:52.32 | poolio | or if not I'll probably move on and implement mutation and the other things we talked about as I don't see myself going anywhere with this. struggled the whole week and pretty much got nowhere |
| 18:55.14 | *** join/#brlcad b0ef (n=b0ef@062016141081.customer.alfanett.no) | |
| 18:59.27 | MinuteElectron | Darn, I am going away again on Sunday. |
| 18:59.49 | MinuteElectron | Won't be back at a copmuter for a week, then might not have full access to one then. |
| 18:59.59 | MinuteElectron | After that week I have to go to Ireland. |
| 19:00.19 | MinuteElectron | For two weeks. |
| 19:00.36 | MinuteElectron | brlcad: How is the LDAP going? DOne any work with it yet on bzflag (just seeing if you got it going) |
| 19:03.24 | *** join/#brlcad b0ef (n=b0ef@062016141081.customer.alfanett.no) | |
| 19:26.01 | brlcad | MinuteElectron: it's coming along slowly, there've been a couple other interruptions (both here and for bz) |
| 19:26.09 | brlcad | MinuteElectron: can you enable my drupal access :) |
| 19:28.38 | ``Erik | blehhhhhhhh |
| 19:29.56 | ``Erik | brlcad: what's the deal with the tclcad bezier patch not being built anymore? |
| 19:36.57 | brlcad | it's gone |
| 19:37.00 | brlcad | no longer needed |
| 19:37.01 | brlcad | poof |
| 19:37.09 | brlcad | no longer used |
| 19:37.10 | ``Erik | between 7.10.0 and 7.10.2 ? |
| 19:37.14 | brlcad | yep |
| 19:37.37 | ``Erik | I was banging my head yesterday trying to figure out why my cvs copy built np using system tcl but the port didn't |
| 19:37.38 | ``Erik | heh |
| 19:37.59 | ``Erik | <-- will introduce system tcl/tk on the fbsd port with 7.10.2 |
| 19:38.07 | brlcad | that was the reason |
| 19:38.30 | brlcad | barring some itcl/itk LIBRARY path problems, we should be good for using system tcl/tk libs now |
| 19:38.57 | ``Erik | it seems to work on fbsd with head |
| 19:39.08 | brlcad | you tried system incr? |
| 19:39.13 | ``Erik | itcl/itk in the fbsd ports is still restricted to 8.4 |
| 19:39.16 | ``Erik | so I haven't attempted |
| 19:39.20 | brlcad | k |
| 19:39.31 | ``Erik | I might create 85 versions in the next few days |
| 19:39.55 | brlcad | incrtcl's versioning is diff |
| 19:40.56 | brlcad | ah, i see what you meant |
| 19:41.19 | brlcad | yeah, they should work with 4 or 5, I don't recall having to update incrtcl for 8.5 |
| 19:42.01 | ``Erik | hum |
| 19:42.15 | ``Erik | after 7.10.2 I may bug you about appropriate testing against older incr's then |
| 19:42.17 | brlcad | if I did, it was pretty minor |
| 19:42.37 | ``Erik | but then it'd require both tk 84 and 85 on the system, due to how ports are set up :/ |
| 19:43.29 | brlcad | 8.5 final shouldn't be too far off |
| 19:43.34 | ``Erik | bleh, I'm out sick today, i shouldn't be worrying about shit like that |
| 19:43.34 | ``Erik | O.o |
| 19:57.10 | MinuteElectron | brlcad: You made an account on the wiki yet? |
| 19:57.44 | brlcad | yes |
| 19:57.52 | MinuteElectron | Cool. |
| 19:58.00 | MinuteElectron | Hmm, when? |
| 19:59.15 | brlcad | that one was just today, the druapl one was a week or so ago |
| 19:59.29 | MinuteElectron | Ahh, |
| 19:59.31 | MinuteElectron | I see. |
| 20:03.30 | MinuteElectron | brlcad: Hmm, not sure quite what you mean by 'enable'. |
| 20:04.56 | brlcad | admin access to make mods |
| 20:05.02 | MinuteElectron | Ahh, done then. |
| 20:05.19 | MinuteElectron | brlcad: You now have top-level permissions on both the wiki and drupal. |
| 20:05.44 | MinuteElectron | Any idea who 'yogesh kumar dhiman' is? |
| 20:05.48 | brlcad | cool |
| 20:05.49 | brlcad | heh, no |
| 20:05.53 | brlcad | sounds like a spammer |
| 20:06.11 | MinuteElectron | probably |
| 20:06.43 | dtidrow_work | or somebody from india? |
| 20:07.11 | MinuteElectron | possibly |
| 20:07.20 | dtidrow_work | or maybe some other asian country... |
| 20:12.19 | brlcad | http://recaptcha.net/ would be good to have on the site |
| 20:12.58 | MinuteElectron | Indeed. |
| 20:13.39 | brlcad | it can, I installed it on a diff. site just recently :) |
| 20:14.23 | MinuteElectron | :) |
| 20:20.22 | MinuteElectron | There appears to be a problem with PHP\Apache, the include path is being registered as /usr/web/ftp.brlcad.org/w/ instead of my. |
| 20:24.20 | brlcad | my is the same as ftp |
| 20:24.53 | MinuteElectron | Ignore that, my bad. |
| 20:24.54 | MinuteElectron | lol |
| 20:25.08 | MinuteElectron | I mucked up the files. |
| 20:25.11 | MinuteElectron | Fixed now. |
| 20:25.37 | MinuteElectron | Wheee: http://my.brlcad.org/w/index.php?title=Special:Userlogin&type=signup |
| 20:26.36 | brlcad | cool |
| 20:57.46 | *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1177871705.dsl.bell.ca) | |
| 20:58.53 | IriX64 | http://www3.sympatico.ca/mario.dulisse2/screen.png <--- progress :) |
| 20:59.59 | IriX64 | vmware is very good |
| 21:00.07 | MinuteElectron | Is it free? |
| 21:00.13 | IriX64 | for 30 days |
| 21:00.19 | MinuteElectron | bah |
| 21:00.23 | IriX64 | heh |
| 21:00.25 | MinuteElectron | casts his eyes away |
| 21:00.41 | IriX64 | i may buy it |
| 21:00.53 | MinuteElectron | How much does it cost? |
| 21:02.02 | IriX64 | 169.00 |
| 21:02.08 | IriX64 | us |
| 21:02.12 | MinuteElectron | holy crap |
| 21:02.26 | IriX64 | thats not much |
| 21:03.20 | MinuteElectron | Perhpas it is good software. |
| 21:03.34 | IriX64 | i deal honestly if the software performs as expected i pay |
| 21:04.20 | IriX64 | registered winrar even though it continues to work past experation date :) |
| 21:04.44 | MinuteElectron | 7-zip? |
| 21:04.49 | IriX64 | and mirc and stuff |
| 21:04.59 | IriX64 | no dont have 7-zip |
| 21:05.05 | IriX64 | tried it though |
| 21:05.24 | MinuteElectron | FLOSSFTWFM |
| 21:05.31 | IriX64 | ? |
| 21:06.05 | poolio | IriX64: vmware has a free "vmware player" |
| 21:06.15 | IriX64 | whats that |
| 21:06.17 | poolio | if you created a disk image while using the trial software, you can continue using it using the free vmware player |
| 21:06.20 | MinuteElectron | Free\Libre Open Source Softeare For The Way For Me. |
| 21:06.30 | poolio | it's kind of a stripped down version that they distribute so people can use the vmx images |
| 21:06.39 | IriX64 | ahh i see |
| 21:07.04 | IriX64 | i bet you need the full version to install though right? |
| 21:07.13 | poolio | http://www.vmware.com/products/player/ |
| 21:07.14 | poolio | no |
| 21:07.21 | poolio | I mean if you are running fine right now, you can use vmware player |
| 21:07.23 | poolio | (i think) |
| 21:07.34 | IriX64 | im still installing |
| 21:08.04 | poolio | Like I know some software I've used (numenta) distributes a linux image that you can run freely using vmware player |
| 21:09.22 | MinuteElectron | IriX64: You could be our vmware distributor. |
| 21:09.27 | MinuteElectron | :P{ |
| 21:09.29 | MinuteElectron | *:P |
| 21:09.34 | IriX64 | i see just reading about it, I can't distribute what im installing though man its IBMs warp 4 |
| 21:09.53 | IriX64 | MinuteElectron :) |
| 21:10.09 | poolio | It wouldn't be that hard |
| 21:10.10 | IriX64 | legal copy by the way |
| 21:10.30 | poolio | Just setup vmware (you'd have to have it) and install something like CentOS, install BRL-CAD, and distribute the image. |
| 21:11.12 | IriX64 | for windows users you mean? whats wrong with distributing a handfull of dll'd and brlcad? |
| 21:11.19 | poolio | http://www.numenta.com/for-developers/software/nuvm.php |
| 21:11.24 | IriX64 | dlls too |
| 21:11.26 | poolio | IriX64: There isn't anything wrong as long as it works |
| 21:11.36 | IriX64 | it works man |
| 21:11.45 | IriX64 | just taking a break from it |
| 21:11.55 | poolio | :) |
| 21:16.03 | IriX64 | heh /update.png lemme know what you're thinking :) |
| 21:17.29 | IriX64 | this on workstation 6 |
| 21:40.21 | MinuteElectron | brlcad: reCAPTCHA on Drupal aint working now - I will fixor it tomorrow hopefully. But then I am gone again for another week or three (one or the other). |
| 21:40.25 | MinuteElectron | Goodnight. |
| 22:01.25 | *** join/#brlcad cad91 (n=18836d32@bz.bzflag.bz) | |
| 22:10.49 | *** join/#brlcad ``Erik (i=erik@c-69-250-155-85.hsd1.md.comcast.net) | |
| 22:38.06 | *** join/#brlcad yukonbob (n=bch@whthyt224-180.northwestel.net) | |
| 23:34.56 | *** join/#brlcad tessier_ (n=treed@kernel-panic/sex-machines) | |
| 23:44.22 | *** join/#brlcad Laniakea (i=clock@217-162-231-34.dclient.hispeed.ch) | |
| 23:49.44 | *** join/#brlcad Twingy (n=justin@74.92.144.217) | |
