| 00:02.59 | *** join/#brlcad n_reed_ (~molto_cre@66-118-151-70.static.sagonet.net) |
| 00:17.50 | *** join/#brlcad ankesh11 (sid8015@gateway/web/irccloud.com/x-fqimfiazyimhtbhr) |
| 00:28.41 | Notify | 03BRL-CAD:starseeker * 63945 brlcad/trunk/src/libbrep/shape_recognition.h: Comment tweaks |
| 00:29.09 | *** join/#brlcad teepee_ (~teepee@gateway/tor-sasl/teepee) |
| 00:32.10 | *** join/#brlcad ankesh11 (sid8015@gateway/web/irccloud.com/x-utrqruwpjzrawrsp) |
| 00:57.18 | *** join/#brlcad ankesh11 (sid8015@gateway/web/irccloud.com/x-viahqqxfyydejhrv) |
| 00:57.55 | Notify | 03BRL-CAD:starseeker * 63946 brlcad/trunk/src/libbrep/shape_recognition.cpp: More more code to library |
| 01:04.13 | Notify | 03BRL-CAD:starseeker * 63947 brlcad/trunk/src/libbrep/shape_recognition.cpp: Need to dupliate surface here, not just cast. |
| 01:22.29 | andromeda-galaxy | brlcad, ``Erik, starseeker: I'm thinking about eventually committing the changes to mkbundle.c that I've been doing, but there's the issue of rt_raybundle_maker(), which I had to change to use an rt_shootrays()-style api... what is the best way to handle that while preserving the deprecation policy? Create a new and renamed version that has the new behavior and convert all calling sites to use it/mark the old depre |
| 01:32.45 | *** join/#brlcad mikolalysenko (sid34553@gateway/web/irccloud.com/x-zenoyncxwndwxebp) |
| 02:16.01 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 02:21.24 | *** join/#brlcad teepee_ (~teepee@gateway/tor-sasl/teepee) |
| 02:51.47 | *** join/#brlcad mikolalysenko (sid34553@gateway/web/irccloud.com/x-bqqkszmsjumedyho) |
| 03:15.14 | andromeda-galaxy | brlcad, maths22: for the 'PROJECT: Implement web applicatoin for testing commit ranges' series of tasks, what is the web server being used for brlcad.org/what mechanism is used for dynamic content? (being able to use websockets for the live updating would be really fun...) |
| 03:21.50 | *** join/#brlcad jfjhh (~jfjhh@99-90-212-118.lightspeed.irvnca.sbcglobal.net) |
| 03:23.36 | maths22 | andromeda-galaxy: can you link to that? |
| 03:24.00 | andromeda-galaxy | maths22: |
| 03:24.03 | andromeda-galaxy | http://www.google-melange.com/gci/task/view/google/gci2014/5275647798673408 |
| 03:24.03 | gcibot | [[ PROJECT: Implement web application for testing commit ranges || BRL-CAD || Open || Mihai Neacsu, Dishank ]] |
| 03:24.08 | andromeda-galaxy | that's the first one in the series |
| 03:29.20 | andromeda-galaxy | maths22: my thinking (in the best case) is be 1) some kind of POST handler on the server to stick files in a queue directory, 2) a cron job to execute them in series (or something started from the POST handler, i.e. CGI script) and 3) a websockets server that can be notified of status updates and send push notifications to the connected clients... maybe 3) should incorporate 2) in a long-running process that uses i |
| 03:29.43 | andromeda-galaxy | maths22: if that cut off, let me know and I'll resend the end of it (unfortunately, irssi doesn't tell me when message length limits are exceeded) |
| 03:48.02 | *** join/#brlcad teepee_ (~teepee@gateway/tor-sasl/teepee) |
| 03:51.54 | starseeker | maths22: was talking with brlcad - is there any way, when we have a successful build of trunk with buildbot, to also run the "make dox" target and copy the results to the brlcad.org website? |
| 03:53.03 | *** join/#brlcad darshpreets (~darshpree@202.164.53.117) |
| 04:01.12 | *** part/#brlcad jfjhh (~jfjhh@99-90-212-118.lightspeed.irvnca.sbcglobal.net) |
| 04:01.28 | *** join/#brlcad jfjhh (~jfjhh@99-90-212-118.lightspeed.irvnca.sbcglobal.net) |
| 04:05.43 | Notify | 03BRL-CAD:starseeker * 63948 brlcad/trunk/db/faa/CMakeLists.txt: Add README with info about report with images |
| 04:11.06 | *** join/#brlcad YashM (~YashM@117.198.16.236) |
| 04:11.24 | *** part/#brlcad jfjhh (~jfjhh@99-90-212-118.lightspeed.irvnca.sbcglobal.net) |
| 04:14.03 | maths22 | andromeda-galaxy: I suggest that you look into at/batch (http://linux.about.com/library/cmd/blcmdl1_batch.htm) |
| 04:14.48 | maths22 | starseeker: I think that would be doable... |
| 04:15.05 | andromeda-galaxy | maths22: those might be useful, but I'm not sure how well they would mesh |
| 04:15.09 | andromeda-galaxy | with live status reports... |
| 04:15.28 | andromeda-galaxy | if it would work to write a process that acts as a websockets server, |
| 04:15.35 | maths22 | What sort of status do you expect? |
| 04:15.42 | andromeda-galaxy | maths22: well, reading some of the later task descriptions, |
| 04:15.59 | andromeda-galaxy | mostly just things like how far through the commit range + script return results for each commit |
| 04:16.45 | andromeda-galaxy | so my idea had been to run a websockets server process that also picked up on things added to a queue directory by a dumb server and ran them, sending status update messages to all connected clients |
| 04:16.50 | maths22 | I would think that we would write a wrapper script that then writes status to a log file |
| 04:17.17 | andromeda-galaxy | but how does that help with client live updating? (ideally, not polling)? |
| 04:17.53 | maths22 | You would want the server to be doing something equivalent to tail -f |
| 04:18.10 | maths22 | You don't want the script to die just because the users' connection died |
| 04:18.30 | andromeda-galaxy | hmm? I'm talking about running a daemon that runs a websockets (not HTTP) server |
| 04:18.39 | andromeda-galaxy | with websockets, you get notifications when clients connect/disconnect |
| 04:18.44 | andromeda-galaxy | and can push-update messages |
| 04:19.10 | maths22 | In my opinion, the websockets server should not be responsible for managing the jobs |
| 04:19.27 | maths22 | It should only be responsible for conveying status updates about jobs that are running "in the background" |
| 04:19.41 | andromeda-galaxy | maths22: that makes sense, I just thought either way would work... |
| 04:19.51 | maths22 | starseeker: does make dox generate any kind of archive? |
| 04:20.00 | andromeda-galaxy | is the HTTP server Apache with some set of the *CGI modules enabled? |
| 04:20.15 | maths22 | andromeda-galaxy: in your mind, how would it work? |
| 04:20.17 | andromeda-galaxy | (sorry to askyou about it, but I don't have a shell account and I can't find the information anywhere on the website) |
| 04:20.23 | andromeda-galaxy | maths22: in this case, something like: |
| 04:20.31 | maths22 | andromeda-galaxy: yes |
| 04:20.43 | maths22 | What language would you be looking for? |
| 04:20.55 | andromeda-galaxy | user submites request -> server drops file in queue dir -> jobs are processed in sequential order (background websockets server sends updates as necessary) |
| 04:21.19 | *** join/#brlcad mikolalysenko (sid34553@gateway/web/irccloud.com/x-btgnqznutnqmwvpw) |
| 04:21.21 | andromeda-galaxy | I was thinking about using Haskell, but if the Haskell RTS is too heavyweight I can try something else |
| 04:21.40 | maths22 | That is what I am describing- use "at" to start the job in the queue |
| 04:22.02 | andromeda-galaxy | right, that makes sense |
| 04:22.14 | maths22 | I think we were agreeing the whole time :) |
| 04:22.32 | andromeda-galaxy | probably :) is using haskell for the websockets server too depency-heavy? |
| 04:23.26 | andromeda-galaxy | maths22: so just to confirm, the workflow should look something like: |
| 04:23.32 | andromeda-galaxy | 1. status reporting daemon is started on server (maybe at boot) |
| 04:23.39 | maths22 | andromeda-galaxy: in my opinion, no |
| 04:24.06 | andromeda-galaxy | 2. user submits request, server creates any necessary files and schedules the job with at |
| 04:24.09 | maths22 | As in-haskell would be fine |
| 04:24.25 | andromeda-galaxy | 3. daemon uses *notify apis to get nice notifications about status and sends results |
| 04:24.28 | andromeda-galaxy | oh, good to know |
| 04:24.43 | *** join/#brlcad darshpreets (~darshpree@202.164.53.117) |
| 04:24.54 | andromeda-galaxy | I enjoy writing things in haskell... last thing --- which of the *many* Apache CGI extensions is brlcad.org currently using? |
| 04:25.57 | maths22 | Let me check.. |
| 04:26.02 | maths22 | Why do you want to know? |
| 04:26.13 | andromeda-galaxy | maths22: thanks! just so that when writing the part that handles the user submitting the form |
| 04:26.33 | andromeda-galaxy | which cgi api to deal with, since, e.g. fastcgi/wcgi sometimes need slightly different code |
| 04:26.48 | andromeda-galaxy | one other thing that I forgot about, sorry --- the server is fbsd, right? so no inotify |
| 04:26.52 | maths22 | Yes |
| 04:27.31 | maths22 | fcgi/scgi |
| 04:27.40 | andromeda-galaxy | okay, then, I'll probably use http://hackage.haskell.org/package/fsnotify for the websockets server |
| 04:28.02 | maths22 | If you are using haskell though, you probablly don't want to use cgi |
| 04:28.30 | maths22 | You probably want something like http://snapframework.com/ |
| 04:28.34 | andromeda-galaxy | maths22: true, though there is https://hackage.haskell.org/package/fastcgi... is there a better system for integrating it into the current server (running Yesod under url rewriting does seem a little to heavy)... |
| 04:28.47 | maths22 | Not really too heavy |
| 04:28.56 | maths22 | I do that for some other things |
| 04:29.01 | maths22 | apache mod_proxy |
| 04:29.03 | andromeda-galaxy | yeah, snap is kind of interesting, I've been doing stuff with Yesod before, but caan look into snap |
| 04:29.11 | maths22 | I don't care between yesod and snap |
| 04:29.22 | maths22 | I think my cs professor mentioned snap; that was all |
| 04:30.02 | maths22 | Also, if you have any questions about haskell stuff, I can pass them along to my professor (his specialty is haskell) |
| 04:30.18 | andromeda-galaxy | in that case, sounds good! so, you think that it would work well to use yesod/snap to server the main interface pages & respond to post requests (with batch for scheduling) and a separate daemon for the websockets? |
| 04:30.30 | andromeda-galaxy | (just checking that I have everything straight)... |
| 04:30.46 | andromeda-galaxy | thanks! just curious, who is it? I might of heard of him... |
| 04:30.52 | andromeda-galaxy | *have heard |
| 04:33.01 | andromeda-galaxy | maths22: well, I'll start working on the frontend pages for the first task for now... |
| 04:33.09 | maths22 | stuart kurtz |
| 04:33.36 | maths22 | see something like https://hackage.haskell.org/package/websockets-snap |
| 04:33.45 | maths22 | They really should probably be one daemon |
| 04:35.00 | andromeda-galaxy | mahts22: true, I was still thinking of using two because of the original cgi idea, they should be easy enough tto combine... |
| 04:35.07 | andromeda-galaxy | do you still think that the job running should be separate? |
| 04:35.40 | maths22 | Yes |
| 04:36.14 | maths22 | That way, the jobs don't get messed up even if the haskell web interface goes down |
| 04:36.39 | andromeda-galaxy | makes sense... I'll get started then; thanks for discussing it with me / explaining the server infrastructure :) |
| 04:39.19 | andromeda-galaxy | maths22: oh, and what about authentication/authorization/running jobs in a container? there isn't anything specific mentioned in the task pages, but not having anything seems like a really bad idea |
| 04:39.51 | maths22 | I agree... |
| 04:40.41 | maths22 | I would use chroot and a dedicated user with limited permissions |
| 04:41.38 | andromeda-galaxy | maths22: sounds good! just have the web server process chroot & drop permissions right at the beginning? |
| 04:42.03 | maths22 | The webserver doesn't need to worry about it |
| 04:42.17 | maths22 | That is really a job for the wrapper script that runs the user supplied script |
| 04:42.36 | andromeda-galaxy | true, well have that script do it then... |
| 04:44.00 | maths22 | Once GCI ends, I can pitch in and work on stuff that doesn't get done |
| 04:44.04 | Notify | 02GCI:gjeet * 4948190365220864 : Task Assigned - This task has been assigned to NGenius. You have 100 hours to complete this task, good luck! |
| 04:44.16 | maths22 | Also, do you know if the doxygen output is all in one file? |
| 04:44.36 | andromeda-galaxy | maths22: for the html output, at least, I don't think it is, but there might be an archive version... |
| 04:45.03 | andromeda-galaxy | (for now, I'll assume a preconfigured chroot environment with, e.g., bash in it for running the scripts in) |
| 04:45.15 | Notify | 02GCI:gjeet * 4958060736937984 : Task Assigned - This task has been assigned to Martin. You have 100 hours to complete this task, good luck! |
| 04:45.25 | Notify | 02GCI:gjeet * 6118055788675072 : Task Assigned - This task has been assigned to Zenon. You have 100 hours to complete this task, good luck! |
| 04:48.00 | *** join/#brlcad raptor_ (~raptor@194.28.75.209) |
| 04:49.07 | starseeker | maths22: it creates a directory of files in doc/doxygen_output |
| 04:50.27 | maths22 | OK |
| 04:51.30 | starseeker | or rather, doc/doxygen_output/html |
| 04:51.40 | starseeker | index.xhtml is the "starting" toplevel file |
| 04:52.13 | starseeker | it's about 970 megs of output |
| 04:53.24 | starseeker | quite a lot of files |
| 04:53.53 | starseeker | > 33000 |
| 04:54.32 | starseeker | needs a (reasonably) new doxygen and the graphviz tools installed |
| 04:55.30 | Notify | 02GCI:brlcad * 5050736735944704 : kerning - Raptor, is there a reason the kerning (spacing between letters) is inconsistent? |
| 04:56.48 | starseeker | not 100% sure, but it may be important to remove the old directory before re-generating the output |
| 04:58.16 | *** join/#brlcad MarcTannous (4f733b36@gateway/web/cgi-irc/kiwiirc.com/ip.79.115.59.54) |
| 04:59.21 | Notify | 02GCI:ralpt3748752_qejwhe * 5886097942904832 : Task Claimed - I would like to work on this task. |
| 04:59.46 | Notify | 02GCI:mihaineacsu29 * 5886097942904832 : Task Assigned - This task has been assigned to Raptor. You have 100 hours to complete this task, good luck! |
| 05:00.52 | Notify | 02GCI:ralpt3748752_qejwhe * 5886097942904832 : Ready for review - The work on this task is ready to be reviewed. |
| 05:02.06 | Notify | 02GCI:brlcad * 5867661997113344 : indeed - Rohit, ah using Sharan's is fine. It looked like the black link was smaller than the red link, but I see now it's simply because the red link is shaking... |
| 05:02.46 | Notify | 02GCI:ralpt3748752_qejwhe * 5886097942904832 : About this work - Calculations for finding V and A, B did my script |
| 05:07.16 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 05:17.47 | maths22 | Only ~136 meg compressed into a tgz |
| 05:22.49 | Notify | 02GCI:gjeet * 5222964656078848 : Nice and Different. - Hey this is a nice animation, I would like to know how are you making it. Like what software packages are used for such animation. Also please... |
| 05:22.50 | Notify | 02GCI:gjeet * 5222964656078848 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements needed to... |
| 05:23.40 | Notify | 02GCI:brlcad * 5340178071683072 : regions within regions - A "region" in BRL-CAD is supposed to indicate a single part, eg, a single piece of metal. That means each link and pin will likely be... |
| 05:23.41 | Notify | 02GCI:brlcad * 5340178071683072 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements needed to... |
| 05:25.17 | Notify | 02GCI:brlcad * 6466077978525696 : Task Closed - Congratulations, this task has been completed successfully. |
| 05:25.32 | Notify | 02GCI:deep10 * 5222964656078848 : Software - I used powerpoint to animate the bits, then rendered it as an mp4. I uploaded that mp4 and got a gif. There were no original files as such since it was... |
| 05:27.01 | Notify | 02GCI:deep10 * 5222964656078848 : Uploaded - I have uploaded the powerpoint and mp4 video. Please take a look. |
| 05:31.45 | Notify | 02GCI:brlcad * 5222964656078848 : what is BRL - So Ko0lGuY, this is pretty fun, interesting, and unexpected. However, I'm wondering why you stopped at "BRL-"? Our trademark is "BRL-CAD" and... |
| 05:32.14 | andromeda-galaxy | brlcad: are you going to create another one for dealing with the rest of rt? |
| 05:39.32 | Notify | 02GCI:deep10 * 5222964656078848 Animate our logo! #11 https://www.google-melange.com/gci/task/view/google/gci2014/5222964656078848: Edits Thank you for your suggestion I will add CAD in there. I... |
| 05:39.34 | gcibot | [[ Animate our logo! #11 || BRL-CAD || NeedsWork || Popescu Andrei, Ishwerdas ]] |
| 05:43.17 | Notify | 02GCI:brlcad * 5886097942904832 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements needed to... |
| 05:46.17 | Notify | 02GCI:ralpt3748752_qejwhe * 5886097942904832 : None - O), what is the underlying? |
| 05:48.18 | Notify | 02GCI:brlcad * 5886097942904832 : no regions? - Raptor, this is creative and nice, but it looks like you didn't define any regions in your model. This means while your model does define some... |
| 05:50.02 | Notify | 02GCI:ralpt3748752_qejwhe * 5886097942904832 : None - Okay, I will do |
| 05:53.43 | Notify | 02GCI:zehata * 6118055788675072 : Ready for review - The work on this task is ready to be reviewed. |
| 05:54.13 | Notify | 02GCI:brlcad * 5886097942904832 : your other model - the "BRL-CAD" was so close! ... are you going to go back to work on that one? |
| 05:57.23 | Notify | 02GCI:brlcad * 6118055788675072 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements needed to... |
| 05:57.27 | brlcad | andromeda-galaxy: yes, but first ... one more pattern! |
| 05:57.33 | Notify | 02GCI:ralpt3748752_qejwhe * 5886097942904832 : None - Yes going, but it's better to upload more ready))) |
| 05:57.37 | brlcad | and this can be two tasks as it's a bit more tricky |
| 05:58.02 | brlcad | I'm going to just describe it here and will get the tasks up in the morning |
| 05:59.02 | brlcad | basically, a pattern of rays based on some input geometry |
| 05:59.20 | andromeda-galaxy | what kind of geometry? |
| 05:59.24 | brlcad | any |
| 05:59.48 | brlcad | so the function will take one of the object containers like a struct rt_db_internal * |
| 06:00.14 | andromeda-galaxy | but where should the rays start/end? |
| 06:00.54 | brlcad | you don't define where a ray ends .. only start and direction ;) |
| 06:01.39 | brlcad | so the caller specifies the geometry and a direction vector |
| 06:02.23 | brlcad | you shoot rays at the geometry in the direction of that direciton vector through the entire grid that spans it's bounding box, effectively sampling it fully |
| 06:02.32 | brlcad | for each ray that hits, those are the rays you keep |
| 06:03.04 | brlcad | so if I shot through an "H" for example, I'll end up with a grid of rays that look like an H |
| 06:03.07 | *** join/#brlcad YashM (~YashM@117.198.16.236) |
| 06:03.34 | andromeda-galaxy | so it basically uses a gen_rect() style bundle of rays but throws out all the ones that aren't useful? |
| 06:03.43 | andromeda-galaxy | (also, is mkbundle.c now in a state to commit?) |
| 06:03.59 | brlcad | technically, any of the patterns that have been done to date could apply to this "object template" generator, but it's a fair bit more complex from an API perspective |
| 06:04.18 | brlcad | yeah, basically |
| 06:05.04 | andromeda-galaxy | so it should infer the "start point" based on direction & bounding box? |
| 06:05.11 | brlcad | the pattern is useful for projecting geometry onto other geometry |
| 06:05.24 | andromeda-galaxy | (i.e. a point outside the bounding box such that the centroid of the geometry is in <direction>) |
| 06:05.38 | andromeda-galaxy | that makes sense... sounds like a very interesting (if computationally intensive) pattern... |
| 06:06.39 | raptor_ | brlcad, i'm upload bicycle_chain |
| 06:07.17 | brlcad | andromeda-galaxy: yeah, it's a real-world example that has a lot of potential uses |
| 06:07.51 | andromeda-galaxy | brlcad: that makes sense... did you see my question about the start point used for the current patterns? |
| 06:08.24 | andromeda-galaxy | also, would it make sense for it to support, i.e., both perspective & orthogonal grid generators? also, there should probably be at least one gridsize parameter... |
| 06:08.28 | brlcad | like say I wanted to drop a pencil on a table, I could sample the pencil shape and then shoot those rays at the table to see where/when it will hit (imagine there are many objects on the table that it might hit too) |
| 06:09.24 | brlcad | there needs to be some sort of grid density parameter, not sure about whether the start point must be specified or can be inferred |
| 06:09.40 | brlcad | unlike the other patterns, simply specifying the geometry and a direction is sufficient |
| 06:09.48 | brlcad | at least for orthogonal rays |
| 06:09.57 | brlcad | and I think only orthogonal make sense |
| 06:10.18 | brlcad | perspective would be .. exploding geometry of some sort |
| 06:11.00 | Notify | 02GCI:casanova_9898 * 5791674965950464 : Task Claimed - I would like to work on this task. |
| 06:11.08 | andromeda-galaxy | brlcad: indeed... gen_rect() probably makes the most sense then... gridsize would need to be specified, but start could definitely be inferred from the geometry |
| 06:12.09 | andromeda-galaxy | brlcad: do you think it would be okay to always center the grid on the centroid of the geometry? |
| 06:12.56 | andromeda-galaxy | in any case, finding the extent of the grid to sample could be somewhat, annoying depending on direction: if it's pointing at a face of the bounding box it's easy, but otherwise it gets a little harder... |
| 06:13.02 | andromeda-galaxy | (to determine maximum possible extents) |
| 06:13.36 | andromeda-galaxy | brlcad: are there existing functions for doing that (basically, taking the geometry and a direction and giving the largest plane orthogonal to that direction in the bounding box)? |
| 06:15.27 | brlcad | a grid spacing or density would work too |
| 06:15.47 | brlcad | should match what the other one is doing |
| 06:15.54 | brlcad | (the gen_rect routine) |
| 06:16.09 | andromeda-galaxy | yeah, sorry I meant to say grid density (the equivealent of gen_rect's da/db) |
| 06:16.35 | brlcad | and while this is shooting a regular grid, I don't quite see it as being the same notion .. it's like the gridded cylinder routine |
| 06:16.52 | brlcad | which is consequently the pattern you'd get by specifying a sphere as the object, nice test case ;0 |
| 06:17.53 | brlcad | andromeda-galaxy: look at what nirt does, as it has a back-out option to make sure you're always shooting a ray from behind the geometry you're shooting at |
| 06:18.30 | brlcad | src/nirt follow logic for do_backout |
| 06:18.31 | andromeda-galaxy | brlcad: the problem that I was thinkinga bout wasn't only getting out of the bounding box |
| 06:18.41 | andromeda-galaxy | but also finding the size of the grid to shoot |
| 06:18.56 | andromeda-galaxy | if it's pointed directly at a face of the bounding box, it's easy :just the width/height of that face |
| 06:19.12 | andromeda-galaxy | brlcad: but if it's at some weird angle relative to the bb, I'm not quite sure how to calculate that... |
| 06:19.24 | brlcad | it's exactly what rt does |
| 06:19.54 | andromeda-galaxy | you mean the auto view stuff? good point... is that in any libraries, or is it in the rt application itself? |
| 06:20.36 | Notify | 02GCI:zehata * 6118055788675072 : Ready for review - The work on this task is ready to be reviewed. |
| 06:21.05 | *** join/#brlcad YashM (~YashM@117.198.16.236) |
| 06:21.26 | andromeda-galaxy | brlcad: if it's currently custom to rt, that might be another candidate for something to factor out into librt... |
| 06:24.42 | brlcad | andromeda-galaxy: both |
| 06:24.56 | brlcad | it's in the application, but there's also the autoview command that you can check out in src/libged/autoview.c |
| 06:26.29 | brlcad | you'll notice it's quite simple .. 2x the extents |
| 06:26.52 | andromeda-galaxy | brlcad: that's one approach... I was curious if there was a more optimal versioni anywhere |
| 06:26.56 | brlcad | that's gauranteed to always show the entire model, no matter the angle |
| 06:27.07 | andromeda-galaxy | i.e. that's always big enough, but it's (almost always) too big |
| 06:27.36 | andromeda-galaxy | it seems like it should be possible to compute the exact sizes, but I'm not sure how... well, if that turns out to be too difficult, I'll just waste a few rays on sampling extra space.. |
| 06:27.47 | brlcad | there's not currently a better version ... that's be a great function/task/feature too ;) |
| 06:28.26 | brlcad | if you work on it and make a function, let me know before friday so I can get the tasks all uploaded for you to claim |
| 06:29.35 | andromeda-galaxy | brlcad: sure, thanks! I'll work on it some tomorrow and see if it can be done easily |
| 06:30.22 | andromeda-galaxy | brlcad: I'll start by doing the new pattern generator with 2x extents & submit that, then see about refining that view... it would certainly be an interesting algorithmic problem |
| 06:31.26 | brlcad | excellent! |
| 06:32.33 | andromeda-galaxy | brlcad: I probably should go for the night soon, I'll start working on all that in the morning... can you ping me when the new bundling tasks are up so that I know what name to look for? |
| 06:32.43 | *** join/#brlcad dhs (~dhs@203.116.130.2) |
| 06:33.12 | dhs | Hello, I am trying to run mged to make my image black and white, but it is returning an error. |
| 06:34.30 | Notify | 02GCI:harman052 * 5791674965950464 : Task Assigned - This task has been assigned to Danny. You have 100 hours to complete this task, good luck! |
| 06:34.48 | brlcad | dhs: hello |
| 06:34.54 | brlcad | dhs: more details? |
| 06:35.15 | brlcad | andromeda-galaxy: yep, will do |
| 06:35.23 | andromeda-galaxy | brlcad: thanks! |
| 06:35.39 | dhs | I am trying to run the png-bw command in mged |
| 06:35.43 | dhs | Or should I do it in terminal? |
| 06:38.15 | brlcad | dhs: terminal |
| 06:38.37 | brlcad | some commands are mged commands, many others are unix command-line commands |
| 06:38.46 | dhs | OK thank you! |
| 06:38.56 | brlcad | sure np |
| 06:39.18 | brlcad | hang around on IRC and many others here will eventually help you if you wait for an answer ;) |
| 06:39.28 | brlcad | for me, have to run off for a bit -- cheers! |
| 06:39.41 | dhs | Have fun, Mr brlcad! |
| 06:48.34 | *** join/#brlcad andromeda_galaxy (~andromeda@108-225-17-54.lightspeed.sntcca.sbcglobal.net) |
| 06:49.22 | *** join/#brlcad andromedagalaxy (~andromeda@108-225-17-54.lightspeed.sntcca.sbcglobal.net) |
| 06:54.21 | Notify | 02GCI:deep10 * 5222964656078848 : 2nd Version - The 2nd Version is up and includes the "CAD". I do apologize for not putting it in the first place. As for uses, I was thinking maybe on a website or... |
| 06:56.11 | Notify | 02GCI:noobboy69 * 5218710037987328 : Ready for review - The work on this task is ready to be reviewed. |
| 06:58.41 | Notify | 02GCI:deep10 * 5222964656078848 : Powerpoint did not save :( - It had just come to my attention the powerpoint did not save. I will recreate the animation and upload it. |
| 07:08.28 | Notify | 02GCI:casanova_9898 * 5791674965950464 : Ready for review - The work on this task is ready to be reviewed. |
| 07:32.53 | *** join/#brlcad andrei_ (c35a6e7d@gateway/web/freenode/ip.195.90.110.125) |
| 07:35.45 | andromedagalaxy | brlcad: one other thing about the rt work that I just remembered: as I mention end,the perspective tests are still failing, but the differences in pixel coordinates appear to be rounding error... What should we do? |
| 07:43.47 | Notify | 02GCI:csh12231223 * 5802485667069952 : Task Claimed - I would like to work on this task. |
| 07:45.01 | dhs | exit |
| 07:50.37 | Notify | 02GCI:popescuandrei * 6118055788675072 : None - Zenon, You've done a good job designing this splash screen, it looks good and it's a clean, simple design! The one thing I would address is if you... |
| 07:51.01 | andrei_ | are these two too similar ? https://www.google-melange.com/gci/task/view/google/gci2014/5791674965950464 and https://www.google-melange.com/gci/task/view/google/gci2014/5218710037987328 ? |
| 07:51.02 | gcibot | [[ Beginner: Install and run BRL-CAD! #57 || BRL-CAD || NeedsReview (4 days 2 hrs 43 min) || Gauravjeet Singh, Ch3ck ]] |
| 07:51.02 | gcibot | [[ Beginner: Install and run BRL-CAD! #15 || BRL-CAD || NeedsReview (2 days 4 hrs 45 min) || Isaac Kamga, Mihai Neacsu ]] |
| 07:54.12 | Notify | 02GCI:izak * 5218710037987328 : Task Closed - Congratulations, this task has been completed successfully. |
| 07:54.54 | andrei_ | ... |
| 08:00.57 | Notify | 02GCI:popescuandrei * 5222964656078848 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements... |
| 08:03.32 | Notify | 02GCI:popescuandrei * 4596346644332544 : Task Closed - Congratulations, this task has been completed successfully. |
| 08:03.52 | Notify | 02GCI:popescuandrei * 5791674965950464 : Task Closed - Congratulations, this task has been completed successfully. |
| 08:12.03 | Notify | 02GCI:zehata * 6118055788675072 : Ready for review - The work on this task is ready to be reviewed. |
| 08:16.11 | Notify | 02GCI:zehata * 6118055788675072 : Review - Dear Mentors, I have tried my best to make the necessary changes. I have set the ppi to 600 and zipped the files as BRL-CAD Splash Screen v2.1. If the... |
| 08:17.51 | Notify | 02GCI:popescuandrei * 6118055788675072 : Task Closed - Congratulations, this task has been completed successfully. |
| 08:19.11 | Notify | 02GCI:popescuandrei * 5222964656078848 : None - Sure thing, I was just trying to give you a starting point, not a forced guideline. Feel free to do as you please, the more creative you are, the... |
| 08:20.11 | Notify | 02GCI:csh12231223 * 5802485667069952 : Ready for review - The work on this task is ready to be reviewed. |
| 08:23.24 | Notify | 02GCI:martizor56 * 4958060736937984 : None - I can't seem to submit my design as a psd format. Would it be okay if i just submit my design as a png format? |
| 08:23.30 | *** join/#brlcad YashM (~YashM@117.198.16.236) |
| 08:24.19 | Notify | 02GCI:martizor56 * 4958060736937984 : Ready for review - The work on this task is ready to be reviewed. |
| 08:25.19 | *** join/#brlcad merzo (~merzo@214-60-132-95.pool.ukrtel.net) |
| 08:29.46 | Notify | 02GCI:popescuandrei * 4958060736937984 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements... |
| 08:30.26 | Notify | 02GCI:popescuandrei * 5802485667069952 : Task Closed - Congratulations, this task has been completed successfully. |
| 08:31.11 | Notify | 02GCI:martizor56 * 4958060736937984 : None - i already put submitted a reference documented of all websites i used to get my images |
| 08:38.04 | *** join/#brlcad YashM (~YashM@117.198.16.236) |
| 08:40.56 | Notify | 02GCI:deep10 * 5222964656078848 : Ready for review - The work on this task is ready to be reviewed. |
| 08:42.06 | Notify | 02GCI:popescuandrei * 5222964656078848 : Task Closed - Congratulations, this task has been completed successfully. |
| 08:43.46 | Notify | 02GCI:deep10 * 5222964656078848 : Thank you - Thank you for your guidance, suggestions and constant feedback. Deep B (Ko0lGuY) |
| 08:56.17 | *** join/#brlcad luca79 (~luca@host73-111-dynamic.4-87-r.retail.telecomitalia.it) |
| 09:41.09 | *** join/#brlcad n_reed (~molto_cre@66-118-151-70.static.sagonet.net) |
| 09:42.35 | *** join/#brlcad Notify (~notify@66-118-151-70.static.sagonet.net) |
| 09:43.00 | Notify | 02GCI:martizor56 * 4958060736937984 : None - i put submitted new reference document just letting you know :) |
| 09:58.58 | Notify | 02GCI:hideyoshi1 * 4639899508539392 : Claim Removed - The claim on this task has been removed, someone else can claim it now. |
| 10:02.13 | Notify | 02GCI:hideyoshi1 * 4956557716488192 : Task Claimed - I would like to work on this task. |
| 10:02.28 | Notify | 02GCI:hideyoshi1 * 4956557716488192 : Claim Removed - The claim on this task has been removed, someone else can claim it now. |
| 10:11.25 | *** join/#brlcad merzo (~merzo@92.60.189.225) |
| 10:14.28 | Notify | 02GCI:thevk * 5340178071683072 : Ready for review - The work on this task is ready to be reviewed. |
| 10:20.39 | Notify | 02GCI:popescuandrei * 4958060736937984 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements... |
| 10:28.23 | *** join/#brlcad YashM (~YashM@117.198.16.236) |
| 11:10.05 | *** part/#brlcad andrei_ (c35a6e7d@gateway/web/freenode/ip.195.90.110.125) |
| 11:10.10 | *** join/#brlcad andrei_ (c35a6e7d@gateway/web/freenode/ip.195.90.110.125) |
| 11:11.24 | *** join/#brlcad YashM (~YashM@117.198.16.236) |
| 11:23.02 | *** join/#brlcad MarcTannous (4f733b36@gateway/web/cgi-irc/kiwiirc.com/ip.79.115.59.54) |
| 11:31.37 | MarcTannous | good afternnon guys |
| 11:31.51 | raptor_ | good) |
| 11:33.43 | andrei_ | hello |
| 11:48.02 | Notify | 02GCI:rohit_agarwal * 5867661997113344 : Ready for review - The work on this task is ready to be reviewed. |
| 11:50.29 | Notify | 02GCI:krishnab2014 * 4632441281052672 : Claim Removed - The claim on this task has been removed, someone else can claim it now. |
| 11:54.08 | Notify | 02GCI:rohit_agarwal * 5867661997113344 : None - I made the links as stable as I can. Sorry I cannot keep the image size smaller than this. The quality is better now. Hope you like it. And one more... |
| 11:55.28 | *** join/#brlcad d_rossberg (~rossberg@66-118-151-70.static.sagonet.net) |
| 12:51.20 | *** join/#brlcad darshpreets (~darshpree@202.164.53.117) |
| 13:27.20 | Notify | 02GCI:martizor56 * 4958060736937984 : Ready for review - The work on this task is ready to be reviewed. |
| 13:30.05 | *** join/#brlcad darshpreets (~darshpree@202.164.53.117) |
| 13:30.17 | Notify | 02GCI:martizor56 * 4958060736937984 : None - i managed to submit my psd file. It was just that before the psd file was too big,it exceeded the 35mb limit. |
| 13:35.52 | ignacio | Good morning... |
| 13:47.52 | MarcTannous | morning ignacio |
| 13:48.24 | *** join/#brlcad shardulc (~shardul@122.169.74.211) |
| 13:51.41 | shardulc | brlcad: are you there? |
| 14:01.16 | andrei_ | shardulc: is there a particular reason why you need brlcad? |
| 14:01.49 | shardulc | andrei_: yes, I wrote a script to detect dead code and he's the only one who knows about it (I think) |
| 14:02.05 | shardulc | he said he'll create a task for me to submit it |
| 14:02.25 | YashM | shadulc, if I am not wrong, you've been waiting for over a week? |
| 14:02.45 | shardulc | YashM: well yeah, I did other tasks |
| 14:03.20 | andrei_ | shardulc: I could create a "dummy task" just so you get credit for it, but I'm not sure if that's the best approach |
| 14:03.40 | shardulc | andrei_: ? |
| 14:03.59 | andrei_ | I don't know the details, so I could create a task with no description that you could claim, submit your file |
| 14:04.02 | andrei_ | then close it |
| 14:04.33 | shardulc | yes, but I'm doing another task currently |
| 14:04.40 | andrei_ | well, after |
| 14:04.52 | andrei_ | hm, I think we ll do the following thing |
| 14:05.24 | maths22 | andrei_: you actually can't exactly |
| 14:05.36 | maths22 | brlcad will have to publish it after you create it |
| 14:05.49 | andrei_ | maths22: not really, I can create it directly on melange |
| 14:06.01 | andrei_ | just like mihaineacsu did with some other task |
| 14:06.18 | andrei_ | shardulc: here's my idea. you keep working on other tasks, and if brlcad doesn't answer until the end of the week, I(or other mentors) will create that dummy task for you to claim, then close it ? |
| 14:07.01 | shardulc | as soon as my current task is reviewed, I'll claim this one |
| 14:07.04 | maths22 | andrei_: maybe I'm wrong, but I thought he has to check the "publish" box in essence for any task that we create, even if we do it in melange |
| 14:07.17 | shardulc | I'll tell you when I'm done on this channel |
| 14:07.33 | andrei_ | maths22: if you go to dashboard, you have a create task link/file |
| 14:08.03 | andrei_ | those tasks appear instantly on Melange |
| 14:08.37 | shardulc | andrei_: is that fine? |
| 14:08.38 | mihaineacsu | I did create a task directly on Melange, I wasn't aware at that time |
| 14:09.12 | andrei_ | shardulc: that depends when you finish your task. if you finish it today, you can claim another |
| 14:09.38 | mihaineacsu | the task should be first created on your list on brlcad.org (@maths22) |
| 14:09.45 | shardulc | it will probably be reviewed 14-15 hours from now (timezone problems!) |
| 14:10.10 | andrei_ | that depends what task it is, I've reviewed all, except modeling ones until now |
| 14:10.23 | shardulc | no, it's an apertium task :) |
| 14:10.28 | andrei_ | oh |
| 14:12.12 | MarcTannous | so much activity on the channel today :D |
| 14:13.19 | andrei_ | Marc: you've taken a break? I haven't seen you asking anyone's opinion on your tasks in a while? |
| 14:13.51 | MarcTannous | trying to catch up with real life, should submit something today though |
| 14:14.10 | MarcTannous | so I |
| 14:14.22 | MarcTannous | will ask about people's opinion pretty soon, no worries |
| 14:14.43 | andrei_ | how many tasks you've got? |
| 14:14.52 | MarcTannous | 40 something I think |
| 14:14.52 | andrei_ | my retarded proxy is blocking the ranking website for some reason |
| 14:15.46 | MarcTannous | 43 |
| 14:24.22 | andrei_ | that's a lot :) |
| 14:33.00 | Notify | 02GCI:tratigen * 4975346856230912 : Task Claimed - I would like to work on this task. |
| 14:36.27 | Notify | 02GCI:popescuandrei * 4975346856230912 : Task Assigned - This task has been assigned to Joel Teo. You have 100 hours to complete this task, good luck! |
| 14:37.32 | Notify | 02GCI:tratigen * 4975346856230912 : Claim Removed - The claim on this task has been removed, someone else can claim it now. |
| 14:37.47 | MarcTannous | has anyone seen http://vincentgarreau.com/particles.js/? This is amazing |
| 14:42.50 | Notify | 02GCI:tratigen * 4956557716488192 : Task Claimed - I would like to work on this task. |
| 14:43.55 | Notify | 02GCI:popescuandrei * 4956557716488192 : Task Assigned - This task has been assigned to Joel Teo. You have 100 hours to complete this task, good luck! |
| 14:44.01 | andrei_ | if he unclaims it again... |
| 14:52.15 | Notify | 02GCI:tratigen * 4956557716488192 : Claim Removed - The claim on this task has been removed, someone else can claim it now. |
| 14:54.50 | andrei_ | right, let him take a break now |
| 14:57.50 | Notify | 02GCI:shivam691999 * 4596409726664704 : Ready for review - The work on this task is ready to be reviewed. |
| 15:08.17 | Notify | 02GCI:rossberg * 5340178071683072 : It has still the "regions within regions" issue - For example link.r has the regions linkers.r and linkers3.r inside. You should read Sean's comment again. ... |
| 15:08.23 | Notify | 02GCI:rossberg * 5340178071683072 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements needed to... |
| 15:10.55 | *** join/#brlcad teepee_ (~teepee@gateway/tor-sasl/teepee) |
| 15:16.12 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 15:18.49 | Notify | 02GCI:rossberg * 5886097942904832 : More regions - A region is a building part of the over all geometry. In this sense cil2 isn't one but 6 regions/parts. Same for the other regions. As the... |
| 15:18.54 | Notify | 02GCI:rossberg * 5886097942904832 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements needed to... |
| 15:25.21 | Notify | 02GCI:rossberg * 5867661997113344 : Looks nice - but I'm still convicted that the red link should cover the black one (red link in foreground, black link in background). |
| 15:25.31 | Notify | 02GCI:rossberg * 5867661997113344 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements needed to... |
| 15:29.07 | *** join/#brlcad andrei___ (c35a6e7d@gateway/web/freenode/ip.195.90.110.125) |
| 15:42.25 | *** part/#brlcad shardulc (~shardul@122.169.74.211) |
| 15:44.54 | Notify | 02GCI:rossberg * 5252586642866176 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements needed to... |
| 15:46.59 | *** join/#brlcad teepee (~teepee@gateway/tor-sasl/teepee) |
| 15:47.08 | raptor_ | I added more regions |
| 15:49.40 | Notify | 02GCI:rossberg * 5886097942904832 : Almost - The regions are OK and the grouping of the regions is OK too - but you should set the color in the regions, not in the higher groups. |
| 15:49.50 | Notify | 02GCI:rossberg * 5886097942904832 : Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this task to satisfy the requirements needed to... |
| 15:52.09 | *** join/#brlcad andrei_ (c35a6e7d@gateway/web/freenode/ip.195.90.110.125) |
| 15:54.16 | raptor_ | ooh) |
| 15:55.41 | raptor_ | 3d printer or machine for color is not needed?) |
| 16:04.46 | raptor_ | i'm added color |
| 16:04.52 | Notify | 02GCI:ralpt3748752_qejwhe * 5886097942904832 : Ready for review - The work on this task is ready to be reviewed. |
| 16:05.47 | Notify | 02GCI:rossberg * 5340178071683072 : Hey! - You have 15 links which means you need 15 regions for them. Then you have 10 pins which require another 10 regions. Together you should come up with 25... |
| 16:12.16 | Notify | 02GCI:rossberg * 5886097942904832 : Even closer - The top level group containing all relevant elements is missing. And, when you are on it, you can remove the region_id attribute from cil1, cil2,... |
| 16:17.02 | Notify | 02GCI:rossberg * 5867661997113344 : Task Closed - Congratulations, this task has been completed successfully. |
| 16:24.19 | raptor_ | im add |
| 16:25.50 | d_rossberg | raptor_: the region_id attribute is still in the cil1, cil2, ... groups |
| 16:26.17 | raptor_ | hm |
| 16:26.32 | d_rossberg | value is -1 |
| 16:27.08 | raptor_ | How to remove it? |
| 16:28.18 | d_rossberg | attr rm ... |
| 16:29.44 | raptor_ | attr rm cil1 region_id ? |
| 16:32.38 | d_rossberg | attr rm cil1 region_id |
| 16:33.51 | raptor_ | i did it |
| 16:34.14 | raptor_ | all groups |
| 16:34.24 | *** join/#brlcad YashM (~YashM@117.198.16.236) |
| 16:34.34 | raptor_ | thanks!!! |
| 16:34.50 | Notify | 02GCI:rossberg * 5886097942904832 : Task Closed - Congratulations, this task has been completed successfully. |
| 16:35.05 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 16:35.40 | Notify | 02GCI:ralpt3748752_qejwhe * 5798062521843712 : Task Claimed - I would like to work on this task. |
| 16:39.05 | *** join/#brlcad gaganjyot (~gaganjyot@118.102.186.30) |
| 16:43.13 | Notify | 03BRL-CAD:starseeker * 63950 (brlcad/trunk/src/libbrep/shape_recognition.cpp brlcad/trunk/src/libbrep/shape_recognition.h): Start working on how to identify primitive candidates from subsets of the subbrep data. |
| 16:58.03 | Notify | 02GCI:ch3ck * 5798062521843712 : Task Assigned - This task has been assigned to Raptor. You have 100 hours to complete this task, good luck! |
| 16:58.19 | *** join/#brlcad nmz787_i (~nmccorkx@134.134.137.71) |
| 17:12.43 | *** join/#brlcad darshpreets (~darshpree@202.164.53.117) |
| 17:15.03 | Notify | 02GCI:ralpt3748752_qejwhe * 5798062521843712 : Ready for review - The work on this task is ready to be reviewed. |
| 17:37.59 | *** join/#brlcad darshpreets (~darshpree@202.164.53.117) |
| 17:40.55 | Notify | 02GCI:Melange * 4596409726664704 : No more Work can be submitted - Melange has detected that the deadline has passed and no more work can be submitted. The submitted work should be reviewed. |
| 17:58.54 | Notify | 03BRL-CAD:starseeker * 63951 brlcad/trunk/src/libbrep/shape_recognition.cpp: More work on subset identification - need to think about this, whether a face can be used more than once in this context... |
| 18:07.52 | MarcTannous | umm |
| 18:07.54 | MarcTannous | any mentors around? |
| 18:08.19 | MarcTannous | Yet another task copied that got accepted |
| 18:08.40 | MarcTannous | http://www.google-melange.com/gci/task/view/google/gci2014/5227763090849792 and http://www.google-melange.com/gci/task/view/google/gci2014/6189423246966784 |
| 18:08.41 | gcibot | [[ Design a BRL-CAD Sticker #8 || BRL-CAD || Closed || Daniel_R, Gauravjeet Singh ]] |
| 18:08.42 | gcibot | [[ Design a BRL-CAD Sticker #4 || BRL-CAD || Closed || Sean, Harmanpreet ]] |
| 18:08.43 | MarcTannous | are dentical |
| 18:08.46 | MarcTannous | identical* |
| 18:09.50 | MarcTannous | actually, one of them is the same design with the triangle being flipped vertically.. |
| 18:23.58 | MarcTannous | _andrei : Just submitted a task after the long hiatus :P |
| 18:24.00 | Notify | 02GCI:tannousmarc * 6103408255696896 : Ready for review - The work on this task is ready to be reviewed. |
| 18:36.55 | brlcad | pretty awesome, genetic algorithm hardware solution: http://www.damninteresting.com/on-the-origin-of-circuits/ |
| 18:38.22 | *** join/#brlcad darshpreets (~darshpree@202.164.53.117) |
| 18:45.58 | brlcad | andromedagalaxy: per your question about rounding error, the source of the error has to be identified before a determination can be made about what to do about it |
| 18:46.37 | andromeda-galaxy | brlcad: ahh... I spent a couple of hours during the last task looking through it, and it looks like the error just comes from the fact that |
| 18:46.58 | andromeda-galaxy | gen_rect() uses a slightly different method to generate the values (and there are a couple of VSCALEs to make things neater) |
| 18:47.06 | Notify | 03BRL-CAD:carlmoore * 63952 brlcad/trunk/src/util/pixcount.c: edit some comments; implement h for help |
| 18:51.44 | andromeda-galaxy | brlcad: I tried a couple of other possibilities in gen_rect(), but none of them improved it by much... It might just be that sending the data throguh a general purpose api can't be done without some amount of fp rounding error |
| 19:03.32 | *** join/#brlcad gaganjyot (~gaganjyot@118.102.186.30) |
| 19:14.31 | *** join/#brlcad raptor_ (~raptor@194.28.75.209) |
| 19:22.36 | *** join/#brlcad gaganjyot (~gaganjyot@118.102.186.30) |
| 19:25.06 | *** join/#brlcad LordOfBikes (~armin@dslb-088-064-033-009.088.064.pools.vodafone-ip.de) |
| 19:27.51 | starseeker | brlcad: that's pretty weird (genetic chip algorithm) |
| 19:29.11 | starseeker | in a way, it didn't really solve the problem though - it was so specific to that particular hardware/environment that it couldn't be copied to other FPGA hardware and work reliably |
| 19:30.18 | starseeker | wouldn't be surprised if it didn't work reliably as a function of time/tempature/humidity/etc. either... |
| 19:30.57 | starseeker | still, I suppose it has potential applications |
| 19:37.45 | *** join/#brlcad teepee_ (~teepee@gateway/tor-sasl/teepee) |
| 19:46.20 | Notify | 02GCI:brlcad * 5338260117127168 : follow-on - https://www.google-melange.com/gci/task/view/google/gci2014/6455158829481984 |
| 19:46.21 | gcibot | [[ Create script to detect dead code || BRL-CAD || Open || Sean ]] |
| 19:47.08 | andromeda-galaxy | starseeker: indeed... possible the largest problem is that it's impossible to make an ASIC from, and so rather difficult to scale up... |
| 19:48.06 | Notify | 02GCI:brlcad * 5338260117127168 : also - Also, all the remaining tasks to eliminate dead code are no longer marked as beginner tasks. Let me know if you need more! :) |
| 19:50.33 | *** join/#brlcad gcibot (ignacio@unaffiliated/ignaciouy/bot/gcibot) |
| 19:50.35 | brlcad | MarcTannous: the performance of that particles site is quickly pretty abysmal, but yeah, awesome work |
| 19:52.27 | brlcad | andromeda-galaxy: there's nearly always a source of rounding error that can be accommodated, but it can be difficult if you've never dug that deep before |
| 19:53.16 | brlcad | starseeker: but it did work .. if you want it to work on multiple chips, you factor in multiple chips into the GA fitness |
| 19:53.28 | andromeda-galaxy | brlcad: makes sense... I'll try looking into it a little bit more after this next pattern... |
| 19:53.57 | brlcad | GAs are *very* specific in that they will strictly converge to optimal solutions, which it most definitely did there |
| 19:58.34 | brlcad | andromeda-galaxy: https://www.google-melange.com/gci/task/view/google/gci2014/5880145319559168 |
| 19:58.34 | gcibot | [[ Investigate and fix source of rounding error || BRL-CAD || Open || Sean ]] |
| 19:58.58 | brlcad | but seriously, do the object ray generator task first ;) |
| 19:59.19 | andromeda-galaxy | brlcad: did you put that one up? I was planning on it, but I've checked a couple of times this mroning |
| 19:59.19 | brlcad | MarcTannous: I have a set of icon tasks that I'd love for you to try if you're interested, before GCI is over |
| 19:59.23 | andromeda-galaxy | (remembering to reload) |
| 19:59.29 | andromeda-galaxy | and it didn't show up |
| 19:59.39 | brlcad | I just put it up now |
| 19:59.52 | MarcTannous | brlcad: Sure thing. Is there any way I can get a follow up task of "mergin all the web galleries in one web page"? Because I'm pretty much done with these |
| 19:59.59 | andromeda-galaxy | brlcad: by the way, on the rounding error one --- I have it printing *all* ray values, and it's still not making much sense... definitely not a much of a cumulative effect |
| 20:00.01 | brlcad | MarcTannous: yep! |
| 20:00.07 | MarcTannous | br |
| 20:00.27 | MarcTannous | brlcad: Also, saw that the "think.invent.create" I came up with for Archer stuck with people |
| 20:00.38 | MarcTannous | saw it on loads on designs |
| 20:01.02 | andromeda-galaxy | brlcad: what did you call the other ray one? |
| 20:04.07 | brlcad | MarcTannous: https://www.google-melange.com/gci/task/view/google/gci2014/5778240039813120 |
| 20:04.07 | gcibot | [[ Create a gallery of all galleries || BRL-CAD || Open || Sean ]] |
| 20:04.45 | MarcTannous | brlcad: if you've got 1min, check this one out http://www.google-melange.com/gci/task/view/google/gci2014/6103408255696896 so that I can claim the one you just made |
| 20:04.46 | gcibot | [[ Create a web gallery of GCI sticker designs #3 || BRL-CAD || NeedsReview (2 days 17 hrs 35 min) || Ishwerdas, Ch3ck ]] |
| 20:04.48 | MarcTannous | thanks btwe |
| 20:04.50 | MarcTannous | btw* |
| 20:04.57 | andromeda-galaxy | brlcad: I can't find the one for making the object bundles, what didyou call it? |
| 20:07.46 | brlcad | MarcTannous: also: https://www.google-melange.com/gci/task/view/google/gci2014/5843395230564352 |
| 20:07.47 | gcibot | [[ Create a gallery of all 3D models #3 || BRL-CAD || Open || Sean ]] |
| 20:09.13 | MarcTannous | hmmm |
| 20:09.17 | MarcTannous | that one will take a while |
| 20:10.06 | MarcTannous | but I'll do it and then do the follow-up task |
| 20:10.53 | andromeda-galaxy | brlcad: I've tried reloading the page a couple more times, it's still not showing up... (object ray generator task |
| 20:10.56 | andromeda-galaxy | ) |
| 20:15.17 | brlcad | andromeda-galaxy: https://www.google-melange.com/gci/task/view/google/gci2014/5250537473703936 |
| 20:15.18 | gcibot | [[ Implement a single function to generate rays based on an object stencil || BRL-CAD || Open || Sean ]] |
| 20:15.43 | andromeda-galaxy | brlcad: thanks! I don't know why it took so long to show up |
| 20:15.45 | brlcad | MarcTannous: how long is "a while"? :) |
| 20:16.09 | MarcTannous | brlcad: if I can claim the 3D task tonight, I'll probably get it done by tomorrow |
| 20:16.15 | andromeda-galaxy | brlcad: oh, you want the user to specify the plane,not just a direction? |
| 20:16.26 | brlcad | andromeda-galaxy: let me know if the description isn't clear -- I tried to keep it simple |
| 20:17.01 | Notify | 02GCI:o7p9bxbnyj * 5250537473703936 : Task Claimed - I would like to work on this task. |
| 20:17.04 | andromeda-galaxy | brlcad: one thing, is for vertical/horizontal density the user would have to input one of vertical or horizontal |
| 20:17.13 | brlcad | but it could be even simpler still if you get stuck, like having to clamp rays to the input plane ... they could be behind it if that's easier, we could fix that later |
| 20:17.28 | brlcad | either separate or a single density, don't think it matters |
| 20:17.52 | brlcad | single density is probably fine, but double might make the api more consistent with the rect pattern .. your call |
| 20:18.04 | andromeda-galaxy | makes sense... I thought we were discussing using something like the nirt backing out to not need a plane and only need a direction? |
| 20:18.45 | andromeda-galaxy | (just thinking about it mentioning a plane instead ofjust a direction) |
| 20:18.54 | andromeda-galaxy | brlcad: also, why would you want to sample from inside the object? |
| 20:19.06 | andromeda-galaxy | The output is a set of rays parallel to the input ray and at or in front of the plane defined by that ray (so you could shoot from inside an object). |
| 20:29.45 | Notify | 03BRL-CAD:carlmoore * 63953 brlcad/trunk/doc/docbook/system/man1/en/pixfade.xml: fill out pixfade.xml, and stop its underscore of > |
| 20:34.10 | *** join/#brlcad merzo (~merzo@43-25-133-95.pool.ukrtel.net) |
| 20:42.25 | Notify | 03BRL-CAD:carlmoore * 63954 brlcad/trunk/src/util/pixfade.c: remove [-h] from Usage, and touch up comments |
| 20:46.34 | Notify | 03BRL-CAD:starseeker * 63955 brlcad/trunk/src/libbrep/shape_recognition.cpp: Start separating out faces into candidate primitive subsets. Need to be more selective when evaluating candidate faces - right now the only check is on type of surface, which isn't adequate. |
| 20:50.22 | *** join/#brlcad Ko0lGuY (uid61616@gateway/web/irccloud.com/x-vopfhzeitrkiixid) |
| 20:56.22 | *** join/#brlcad teepee_ (~teepee@gateway/tor-sasl/teepee) |
| 21:09.28 | Notify | 02GCI:brlcad * 5907391937773568 : follow-on - There are now a slew of additional icon tasks added! Just search on the title or tags for icons and you'll see there are five other icon sets we would... |
| 21:09.43 | Notify | 02GCI:brlcad * 6380328235040768 : follow-on - Yash, There are now a slew of additional icon tasks added! Just search on the title or tags for icons and you'll see there are five other icon sets we... |
| 21:10.13 | Notify | 02GCI:brlcad * 6436162138800128 : follow-on - Aditya, There are now a slew of additional icon tasks added! Just search on the title or tags for icons and you'll see there are five other icon sets... |
| 21:10.33 | Notify | 02GCI:brlcad * 6143905720434688 : follow-on - There are now a slew of additional icon tasks added! Just search on the title or tags for icons and you'll see there are five other icon sets we would... |
| 21:12.25 | Notify | 02GCI:deepakkumarsharma * 5250537473703936 : Task Assigned - This task has been assigned to Andromeda Galaxy. You have 120 hours to complete this task, good luck! |
| 21:15.35 | brlcad | andromeda-galaxy: it's pretty trivial to project backed-out rays onto a plane later, that's why I said it doesn't matter if you do it one way or the other |
| 21:15.58 | andromeda-galaxy | brlcad: right, I think I was just confused by what you meant by in front of the plane |
| 21:16.17 | andromeda-galaxy | i.e. when first reading it I thought you meant closer to the object then the plane, which doesn't really fit either scenario |
| 21:16.25 | brlcad | the silhoette of an object is potentially quite different from a given cut plane |
| 21:17.06 | andromeda-galaxy | brlcad: not sure I understand... |
| 21:17.34 | brlcad | imagine a chair being my object, looking at it from a corner ... fully backed out, I might see four legs, but from an interior plane, there may be other 3, 2, or 1 leg |
| 21:17.44 | brlcad | don't worry about it :) |
| 21:17.51 | brlcad | do whatever you can get working :) |
| 21:18.14 | brlcad | shooting the object and getting hit points, so you know which rays to keep is enough for one task anyways |
| 21:18.34 | andromeda-galaxy | brlcad: sounds good :) I'll try to figure out nirts backing-out right now |
| 21:18.43 | andromeda-galaxy | more globals... |
| 21:19.02 | andromeda-galaxy | wait, why isn't the bounding sphere stuff in librt? |
| 21:19.32 | andromeda-galaxy | brlcad: it looks like nirts backing-out uses the nirt bsphere functionality which isn't in librt and so can't be accessed by the gen functions... |
| 21:20.23 | andromeda-galaxy | brlcad: I see what you mean about the difference between the silhouette and the cut plane, projecting the rays back onto the cut plane shouldn't be hard |
| 21:24.06 | raptor_ | brlcad: that is a lot of tasks at the end ))) |
| 21:25.00 | raptor_ | eyes diverge |
| 21:28.15 | brlcad | raptor_: there was a huge dump at 20%, another at 40, and another at 75% ... there's only been a few more added here at 90% |
| 21:28.42 | brlcad | the 75% was late, supposed to be nearer to 65% |
| 21:29.05 | brlcad | oh, and 0% ;) the initial dump |
| 21:29.43 | brlcad | andromeda-galaxy: "why isn't the bounding sphere stuff in librt?" ... pretty much the same answer anywhere in our code |
| 21:30.10 | brlcad | either it's only used in one place, or nobody has gotten around to refactoring it into an api function |
| 21:30.44 | raptor_ | :) |
| 21:31.01 | andromeda-galaxy | brlcad: okay, that makes sense! |
| 21:31.51 | andromeda-galaxy | brlcad: just to clear up a terminology thing quickly: when something refers to the front of a model/scene it means the side closest to the observer, right? so for example the ... in front of the plane ... in the task description refers to moving in the opposite direction as towards the object? |
| 21:32.16 | andromeda-galaxy | also, should the function take an rt_i structure, or is there some other container of geometry data it should use? |
| 21:32.30 | raptor_ | good night guys! |
| 21:33.59 | andromeda-galaxy | brlcad: if I pastebin you the current state of mkbundle (the other changes to it prevoius to adding this function) can you tell me if it's ready to be committed? the main thing that I'm worried about is the set of chagnes to rt_raybundle_maker() |
| 21:35.45 | brlcad | andromeda-galaxy: a plane is defined by a point and a vector ... progressing "forward" means going in the direction of the vector |
| 21:36.19 | brlcad | imaging a ground plane ... the vector would point up (0,0,1) and in front of the ground is the air, behind the ground plane is .. dirt |
| 21:36.56 | andromeda-galaxy | hmm, that makes sense... I was just confused by the usage of "in front nof the plane" in the task description, because I had thought that we were discussing making the rays start either on the plane or behind it... |
| 21:37.49 | brlcad | well, the function is creating rays that will start behind it ... it has to in order to ensure a hit |
| 21:38.09 | brlcad | but the *result* from the function is a set of rays that start on the ground plane |
| 21:38.24 | andromeda-galaxy | here's the sentence that was confusing me: |
| 21:38.31 | andromeda-galaxy | The output is a set of rays parallel to the input ray and at or in front of the plane defined by that ray |
| 21:38.33 | brlcad | i'm not in a position to evaluate mkbundle at the moment with any sort of justice |
| 21:39.01 | *** join/#brlcad merzo (~merzo@43-25-133-95.pool.ukrtel.net) |
| 21:39.12 | brlcad | yes, I realize what was confusing you ... just not sure if you're still confused :) |
| 21:39.18 | maths22 | brlcad: starseeker said you were discussing doing something with the automatic builds and doxygen |
| 21:39.26 | maths22 | What is your idea? |
| 21:39.32 | andromeda-galaxy | brlcad: I don't think I'm confused anymore, thanks |
| 21:40.08 | andromeda-galaxy | brlcad: on mkbundle, I'm a little worried aobut the rt_raybundle_maker changes, since the entail a reasonably large change at call sites (people now have to use rt_shootrays() instead of rt_shootray_bundle()) |
| 21:40.10 | brlcad | maths22: basically, needing/wanting to tie in doxygen output into our nightly builds ... ideas on the best way to do that? |
| 21:40.49 | brlcad | andromeda-galaxy: the easy answer would be to leave that function alone and create a different one named consistent with the other new functions |
| 21:40.57 | brlcad | (though I think they all need to be renamed) |
| 21:41.23 | andromeda-galaxy | what do you think they should be renamed to? |
| 21:41.54 | maths22 | brlcad: I think we should have a special target-called "nightly" or something like that, that will run the applicable steps |
| 21:42.19 | maths22 | I.E. normal make all + make dox |
| 21:42.30 | maths22 | And anything else we want to add |
| 21:42.31 | brlcad | hmm, interesting |
| 21:42.42 | brlcad | I was thinking that this could be all buildbot |
| 21:43.10 | brlcad | since it's already doing (or should be doing) checkouts and builds |
| 21:43.19 | maths22 | That makes sense. |
| 21:43.22 | maths22 | I will look into it |
| 21:43.26 | brlcad | awesome |
| 21:43.48 | brlcad | I figure it's already got some testing logic that we wouldn't want to be part of the build system |
| 21:44.02 | brlcad | (or maybe we do) |
| 21:44.28 | andromeda-galaxy | brlcad: I can do the making-a-new-function-to-replace-raybundle-maker thing in aa little bit, but what do you think that the rest of the functions should be renamed to? |
| 21:44.42 | brlcad | maybe the buildbot scripts that it uses should live in the repo, so it uses them when it does a checkout |
| 21:44.59 | brlcad | andromeda-galaxy: not worried about the name, just noting that they suck |
| 21:45.23 | brlcad | can sort that out later ... thinking that through is subjective and will take some time, but very untechnical |
| 21:45.35 | andromeda-galaxy | brlcad: sounds good.. |
| 21:46.15 | Notify | 03BRL-CAD:starseeker * 63957 (brlcad/trunk/src/libbrep/shape_recognition.cpp brlcad/trunk/src/libbrep/shape_recognition.h): Start setting up to collect more information on which to base an include/no-include decision. |
| 22:51.56 | *** join/#brlcad circ-user-0UMVh (~circuser-@122.163.203.189) |
| 22:51.56 | circ-user-0UMVh | Hi |
| 22:52.22 | circ-user-0UMVh | Hi |
| 22:59.31 | Notify | 02GCI:Melange * 5002699137351680 : Task Reopened - Melange has detected that the final deadline has passed and it has reopened the task. |
| 23:02.04 | *** join/#brlcad teepee_ (~teepee@gateway/tor-sasl/teepee) |
| 23:05.02 | Notify | 03BRL-CAD:starseeker * 63959 (brlcad/trunk/src/libbrep/shape_recognition.cpp brlcad/trunk/src/libbrep/shape_recognition.h): Will need to think about tolerances at some point - this is a test of a shape in NIST model 2, which doesn't satisfy 0.01 on the tolerances for cylinders but does satisfy a looser tolerance. |
| 23:09.56 | andromeda-galaxy | brlcad: what about having a parameter to control whether or not the function should back out? then for simple objects where that is desired, it happens, and for complex objects where it might be intentional to look a the object from within the object, that's possible too... |
| 23:16.47 | Notify | 02GCI:maitreyav * 4635701832318976 : Ready for review - The work on this task is ready to be reviewed. |
| 23:32.04 | Notify | 02GCI:vs37nx * 5860136425881600 : Ready for review - The work on this task is ready to be reviewed. |
| 23:41.41 | andromeda-galaxy | brlcad: currently, any function wanting to override a_hit()/a_miss() for rt_shootrays() has lots of boilerplate to implement... what do you think of making the default a_hit() and a_miss() functions available publicly as, e.g., rt_shootrays_default_bundle_hit/rt_shootrays_default_bundle_miss? Then people implementing new a_hit/a_miss functions for use in an rt_shootrays() call could just call those functions and be |
| 23:42.00 | *** join/#brlcad brlcad (~sean@66-118-151-70.static.sagonet.net) |