IRC log for #brlcad on 20141212

00:03.28 Notify 02GCI:mitsukimoon * 4949440150372352 Design a BRL-CAD splash screen #4: Ready for review - The work on this task is ready to be reviewed.
00:55.26 *** join/#brlcad ofnlut_ (~ofnlut_@c-68-48-58-20.hsd1.dc.comcast.net)
01:00.42 starseeker hah! http://brlcad.org/~starseeker/bunny_chull.png
01:01.30 andromeda-galaxy starseeker: awesome!
01:05.26 starseeker brlcad: here's what it did with a rotated cube: http://brlcad.org/~starseeker/rotated_cube_chull.png
01:06.13 starseeker performance isn't so hot (not sure the dragon model will finish - have to try it overnight) but at least it's working
01:07.30 andromeda-galaxy starseeker: asking on #cmake about some cmake/ctest argument-quoting woes
01:07.57 andromeda-galaxy starseeker: the chulls look great!
01:08.37 starseeker nods - yeah, the quoting thing kinda sucks
01:08.40 starseeker thanks!
01:09.06 starseeker andromeda-galaxy: now all we need is for someone to read the quickhull paper and update chull3d.c to use that approach ;-)
01:09.29 starseeker n_reed is right - need to profile it to see where time is spent
01:10.30 starseeker from 3155 lines in the original hull .c and .h files to 1222 in chull3d.c as it stands today
01:11.41 andromeda-galaxy wow, that's a nice reduction
01:12.14 andromeda-galaxy the problem with ctest is that it won't pass empty arguments through to the test --- if I say add_test(... tester_.. "" 5), it calls tester_... "5"
01:12.50 starseeker andromeda-galaxy: I think in that situation I've just added a specific option to the command that tells it to test empty
01:12.57 starseeker ./tester_yadda -empty
01:12.59 starseeker or some such
01:17.12 andromeda-galaxy yeah... for a couple of others that could only have one argument blanck, I set it up to interpret a missing argument as an empty string
01:17.35 andromeda-galaxy just doesn't always work if there are multiple different combinations of possibly empty arguments
01:17.48 andromeda-galaxy strcmp() it with empty or some such special value is a possibility, of course...
01:18.39 andromeda-galaxy If I figure out how to fix this, we should be down to only the bu_semaphore() tests failing on single core machines, as far as I know...
01:44.41 Notify 02GCI:Melange * 4949440150372352 Design a BRL-CAD splash screen #4: No more Work can be submitted - Melange has detected that the deadline has passed and no more work can be submitted. The submitted...
01:48.28 *** join/#brlcad ofnlut_ (~ofnlut_@c-68-48-58-20.hsd1.dc.comcast.net)
02:02.41 andromeda-galaxy starseeker: are you still here?
02:03.02 andromeda-galaxy I was talking to a couple of people in #cmake, it seems like there isn't erally any way to make CMake pass the empty string through
02:03.34 andromeda-galaxy what do you think of defining a BRLCAD_ADDTEST() macro that evaluates to using add_test() with a little sh wrapper script that converts arguments like $EMPTY into ""?
02:03.41 andromeda-galaxy I mean, arguments like EMPTY
02:05.43 Stragus I don't think a bash script would be acceptable for portable compilation
02:05.53 andromeda-galaxy /bin/sh?
02:06.03 Stragus Yes yes, what about on Windows? :)
02:06.49 andromeda-galaxy good point...
02:07.15 andromeda-galaxy well, we can't do it in CMake, and putting a thing to check for each argument in *every* test seems annoying
02:07.30 andromeda-galaxy I guess we could write a C prograam that goes through its arguments, fixes them, and then calls system()...
02:07.39 Stragus I have no good suggestion to offer, I can't stand CMake myself
02:07.41 andromeda-galaxy then have the BRLCAD_ADDTEST() macro call that...
02:08.29 andromeda-galaxy sStragus: agreed.. unfortunately, the rest of BRLCAD uses it...
02:09.51 Notify 02GCI:parthmiglani_21 * 6620713125412864 Write manual page documentation (for archer): Ready for review - The work on this task is ready to be reviewed.
02:10.18 ignacio Leaderboard updated! Please look it! http://kuckuck.treehouse.su:5000
02:13.29 andromeda-galaxy Stragus: now if we did that, where should the script go?
02:16.24 Stragus There's already a sh/ directory with bash scripts, I guess that would seem appropriate
02:16.51 Stragus But I don't think compilation should depend on bash just for that little "" issue
02:19.33 andromeda-galaxy true... I was saying, if I do write a little C90 program to do it
02:19.39 andromeda-galaxy I'm not really sure where that should go
02:19.59 andromeda-galaxy I was thinking of misc/CMake/BRLCAD_Tests.cmake for the CMake fie
02:20.01 andromeda-galaxy *file
02:24.43 starseeker andromeda-galaxy: what about just constructing your test app to have a -e option that takes a numerical argument (1, 2, 3, etc.) to simulate various empty argument combinations?
02:25.00 starseeker Stragus: if you can't stand CMake try autotools
02:26.17 andromeda-galaxy starseeker: it's just that then we would have to do that for *every* test
02:26.20 starseeker the argc/argv arguments to the test app don't have to literally express what you want to run
02:26.41 starseeker what do we do for libbu's vls?
02:26.55 andromeda-galaxy well, up until a couple of days ago, those were failing to
02:27.11 andromeda-galaxy then I set it up so that the tests interpreted a argc of one less than expected as meaning that expected was empty
02:27.33 andromeda-galaxy since this isn't an optimal solution. I was looking for something better to do today...
02:27.47 starseeker andromeda-galaxy: we control the test applications ourselves
02:27.57 starseeker so they can do whatever they need to do
02:28.25 starseeker is find with having something like a -t # in any test app we need to have it so we can simulate various "odd" conditions
02:28.28 andromeda-galaxy starseeker: I just don't like duplicating all of that...
02:28.37 starseeker how is it duplicating?
02:28.47 andromeda-galaxy well, right now each test app is a standalone C program
02:28.53 starseeker right...
02:29.02 andromeda-galaxy each one would need that functionality, and we don't currently have a libtest or some such to use
02:29.14 andromeda-galaxy that's why I was thinking of writing one C program to wrap them
02:29.30 andromeda-galaxy and the writing a BU_ADDTEST() macro that automatically wraps the tests in calls to that program
02:29.30 starseeker you just define a bu_getopt to grab the number, and then a switch statement for the possibiliteis
02:30.09 andromeda-galaxy right, just seems a shame to be duplicting ~5-10 lines of code per test
02:30.25 starseeker well, that's not too huge
02:30.31 starseeker small price to pay for robustness
02:31.20 starseeker suspects any other solution will be more complicated...
02:32.26 andromeda-galaxy what's wrong with using a small C program to wrap them (see above)?
02:32.41 starseeker it adds an extra layer
02:33.02 starseeker probably need to discuss with brlcad - my preference is to keep the test apps as dirt simple as possible
02:33.30 andromeda-galaxy It seems to me like the individual applications should be simpler with a common wrapper to execute them properly...
02:34.06 starseeker worries about portability for things like that... I've written small C programs before to do things in the configure process, and they're almost always portability headaches
02:34.23 andromeda-galaxy not something for configure, just something to run for each test
02:34.40 andromeda-galaxy that iterates through argv, copies to a new array with "EMPTY" replaced by "" or with the -t argument
02:34.57 andromeda-galaxy and then calls the real script with system() etc. portable C
02:35.10 starseeker see, it's that last step I'm wary of
02:35.54 starseeker andromeda-galaxy: do you have an example that displays the problem?
02:36.24 andromeda-galaxy well for one, bu_binstr_to_hexstr_test4
02:36.29 andromeda-galaxy that's why that one keeps failing
02:36.54 starseeker bitv-tests.cmake?
02:36.59 andromeda-galaxy indeed.
02:37.09 andromeda-galaxy that test calls the tester with "" as the first argument, which gets stripped out
02:38.38 andromeda-galaxy what about <stdlib.h> int system(const char *command)? the man page says CONFORMING TO: C89, C99, POSIX.1-2001
02:41.11 *** join/#brlcad YashM (~YashM@117.222.23.4)
02:41.26 starseeker andromeda-galaxy: there are several ways to go about this
02:41.36 starseeker andromeda-galaxy: can you test on Windows?
02:41.48 starseeker the regress tests don't work on Windows, but in principle these should
02:41.49 andromeda-galaxy I don't have a windows vm with me now unfortunately...
02:43.30 *** join/#brlcad YashM_ (~YashM@117.198.20.181)
02:44.33 andromeda-galaxy starseeker: I could grab an IE testing vm form https://www.modern.ie/en-us/virtualization-tools, I guess... getting brlcad set up on it would be kind of annoying though
02:44.53 starseeker andromeda-galaxy: let me try something once...
02:45.00 andromeda-galaxy starseeker: sure
02:45.48 Notify 02GCI:zehata * 5834136975572992 Design a web page that overviews the BRL-CAD contributors #2: Ready for review - The work on this task is ready to be reviewed.
02:51.33 *** join/#brlcad YashM (~YashM@117.222.18.69)
02:52.35 Notify 03BRL-CAD:starseeker * 63699 (brlcad/trunk/src/libbu/tests/bitv-tests.cmake brlcad/trunk/src/libbu/tests/bu_bitv.c): Add a special case for empty input testing of hexstr_to_binstr
02:52.41 starseeker andromeda-galaxy: what about that?
02:52.50 andromeda-galaxy let me look at it...
02:53.55 andromeda-galaxy that might work...
02:54.24 andromeda-galaxy I'd probably prefer to just check argc() in the test function
02:54.40 starseeker andromeda-galaxy: as you pointed out though, that's not a general solution
02:54.47 starseeker what if you wanted the other arg to be empty?
02:54.51 andromeda-galaxy but I know how to do a couple of workarounds in that way, what this was about was finding a more general way to make all the tests work automatically without including special cases
02:55.00 andromeda-galaxy right
02:55.49 starseeker is OK with special casing such things - it's hard to predict what the patterns will be when you're testing such a wide range of functions, if there even are such patterns - but that's just my opinion
02:56.00 starseeker andromeda-galaxy: feel free to make a stab at it
02:56.09 starseeker you've thought about it harder than I have :-)
02:57.56 andromeda-galaxy starseeker: do you know if system() works on windows?
02:58.03 andromeda-galaxy supposedly it's part of C*9
02:58.05 andromeda-galaxy C*(
02:58.07 andromeda-galaxy C89
02:58.38 starseeker I'm afraid I don't. Even if it nominally does, I'm worried that some antivirus programs might decide they don't like it...
02:59.05 starseeker no idea if that's a possibility...
02:59.07 andromeda-galaxy argh.. the program has to be able to spawn a child portable *somehow*
03:01.24 starseeker it might just work - I guess I always think back to editit.c, which is what we use to launch an external editor
03:01.43 starseeker complicated business, but maybe launching a test app will be simpler
03:02.39 starseeker andromeda-galaxy: brlcad's the guy to talk to for this sort of stuff - his knowledge is much deeper than mine
03:03.09 andromeda-galaxy starseeker: okay...
03:03.32 starseeker andromeda-galaxy: the objection to r63699 is verbosity?
03:04.01 andromeda-galaxy partly, and partly the fact that similar code could be required in potentially all of the tests
03:05.31 starseeker wishes we had more powerful option parsing - this is an ideal case for flags that can optionally accept an input or exist on their own
03:06.01 andromeda-galaxy Maybe the easiest thing to do is to have some libbu function that each test invokes at the beginning that creates a new argv with any args with a specific value replaced by ""
03:06.01 starseeker ./tester -i "str1" -f "format_str" vs. ./tester -i -f "format_str"
03:06.41 andromeda-galaxy then the verbosity is *much* lower, duplication is mostly avoided, we don't need a CMake macro, and we don't have to worry about cross-platform execution
03:06.44 andromeda-galaxy !
03:07.02 starseeker andromeda-galaxy: that might be the way to go
03:07.50 starseeker lord knows upgrading the option parsing is beyond gci scope...
03:08.08 Stragus Some library functions to preprocess/filter arguments would be a sane way to go
03:08.13 starseeker andromeda-galaxy: yeah, the more I think about it the more that makes sense
03:08.25 andromeda-galaxy starseeker, Stragus: great! I'll start working on that
03:08.55 andromeda-galaxy would it be alright to call fix_args(&argc, argv) and let it mutate the caller's argc and argv?
03:09.14 andromeda-galaxy or would it make more sense to do something like new_argc = fix_args(argc, argv, &new_argv)?
03:09.32 Stragus new_argv would be problematic
03:09.56 Stragus First choice really makes it easier for everybody
03:10.06 starseeker doesn't recall offhand what the issues might be with mucking with the input argv...
03:10.22 Notify 02GCI:icyjuban * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: None - Almost done, just one final question: what do you mean when you say to simplify the prose?
03:10.23 starseeker but yeah, from a function api standpoint I agree
03:10.27 andromeda-galaxy the only thing left then is that we'd need another function/macro to free the malloc()d argv at the end
03:10.31 starseeker can always duplicate the argv array
03:10.51 andromeda-galaxy and reset &argv to point to the duplicate?
03:11.09 Stragus By the C standard, argc and argv are writable
03:11.41 andromeda-galaxy then I'll probably just make it take (&argc, argv) and overwrite the pointers with pointers to "" on any necessary arguments
03:11.59 starseeker how about this - have bu_test_argv(argc, argv) return a char ** pointer to either the original argv (if no changes were made by the function) or a pointer to the new malloced argv copy
03:12.19 andromeda-galaxy starseeker: would work... you need some way to modify argc as well...
03:13.01 starseeker have the function create a null terminated argv array, so you can iterate over it looking for the null?
03:13.06 starseeker can get acount that way
03:13.20 andromeda-galaxy would work that way as well, it's just a little more work for the caller
03:14.11 andromeda-galaxy or int new_argc; char **new_argv; new_argv= bu_test_argv(argc, argv, &new_argc)
03:14.21 starseeker so char **working_argv = bu_test_argv(argc, argv); if (working_argv != argv) { iterate to get argc; } else { working_argc = argc }
03:14.37 Stragus This bu_test_argv() could have other uses such as a special syntax to say "read parameters from that file", or whatever other generic argument operations desired in all programs
03:15.14 andromeda-galaxy Stragus: true
03:15.27 andromeda-galaxy starseeker: would work, but again, why make the caller have to do more than necessary?
03:15.53 starseeker if we go with your way (which I agree is probably simpler for the caller) then how about this?
03:16.57 starseeker int parse_result = bu_test_argv(argc, argv, &ac, &av); with the return code assigned to parse result containing information about what happened
03:17.13 andromeda-galaxy int ac, char ** av?
03:17.23 andromeda-galaxy so it sets av to be a pointer to the new argv and ac to be the new argc?
03:17.34 starseeker 0 = no change, 1 = successful parse with one or more changes from the original, -1 if error
03:17.49 andromeda-galaxy starseeker: right!
03:18.09 Stragus If the call modifies argc and argv, then you just need one extra line of code, people can keep using same ol' good argc/argv
03:18.09 starseeker right - av will either be the new argv array (return code 1), the original argv pointer (return code 0), or NULL (error)
03:18.25 andromeda-galaxy I'll work on that in a little bit...
03:18.29 starseeker Stragus: you're sure those are modifiable?
03:18.32 Stragus Yes
03:18.36 andromeda-galaxy http://port70.net/~nsz/c/c89/c89-draft.html
03:18.42 Stragus So says the C standard
03:19.00 andromeda-galaxy * The parameters argc and argv and the strings pointed to by the argv array shall be modifiable by the program, and retain their last-stored values between program startup and program termination.
03:19.01 starseeker andromeda-galaxy: OK, try that then - same parse result codes, just no extra parameters
03:19.08 andromeda-galaxy starseeker: sure, will do
03:19.24 starseeker and the caller will know whether they need to free argv based on the return code
03:19.40 andromeda-galaxy sounds good! I think that I need to go for a little while now, I'll work on that when I get back...
03:19.54 Notify 02GCI:Melange * 4938019295461376 Design a new website favicon! #10: No more Work can be submitted - Melange has detected that the deadline has passed and no more work can be submitted. The submitted...
03:20.07 starseeker or actually I suppose it doesn't really matter, since this isn't something intended for other libraries but only for programs...
03:20.24 starseeker andromeda-galaxy: sounds good
03:21.07 starseeker still, would be nice to know via return code for debugging purposes if bu_test_argv worked or not...
03:21.17 Stragus How can it fail? :)
03:21.26 starseeker someone might feed it garbage
03:21.36 starseeker argv array that doesn't match its argc
03:21.44 Stragus Then it's going to segfault
03:22.07 starseeker let me guess - the standard doesn't specify null termination on an argv array?
03:22.47 Stragus Ah yes, looks like argv[argc] must be zero
03:23.06 starseeker so then there should be a way to sanity check
03:31.41 Notify 02GCI:pgambs * 4639899508539392 Write manual page documentation (for g-jack): Task Claimed - I would like to work on this task.
03:33.51 Notify 02GCI:pgambs * 4639899508539392 Write manual page documentation (for g-jack): Claim Removed - The claim on this task has been removed, someone else can claim it now.
03:56.22 *** join/#brlcad chick_ (~capslock@41.205.22.41)
04:06.53 ofnlut_ can anyone help me out with the cmake dev-C++ task?
04:10.58 *** join/#brlcad ofnlut__ (~ofnlut_@c-68-48-58-20.hsd1.dc.comcast.net)
04:14.39 *** join/#brlcad Michael_Huang (~mike@2601:9:740:44:85f3:d68:3131:e7bd)
04:17.59 *** join/#brlcad ofnlut_ (~ofnlut_@c-68-48-58-20.hsd1.dc.comcast.net)
04:38.31 andromeda-galaxy starseeker, Stragus: is this something to put in argv.c/file.h?
04:45.04 *** join/#brlcad YashM (~YashM@59.88.27.35)
04:51.23 andromeda-galaxy starseeker, Stragus: I just realized that if it's checking for #EMPTY and replacing that with "", then it doesn't really need to change argc, so the type, at least for now, is int bu_test_argv(int argc, char **argv) (it updates argv in-place)
04:57.30 MarcTannous did Sean say his opinion regarding the bu seemaphore testS?
04:57.36 MarcTannous and good morning everyone
05:01.06 Notify 02GCI:brlcad * 5907948270256128 BRL-CAD Promotional Video #2 http://www.google-melange.com/gci/task/view/google/gci2014/5907948270256128: have the right idea Ekamjot, you are definitely on the...
05:01.11 Notify 02GCI:brlcad * 5907948270256128 BRL-CAD Promotional Video #2: 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...
05:02.12 Stragus andromeda-galaxy, I think the point of allowing an update to argc was to make it future-proof
05:02.34 Stragus For example if, one day, you wanted to add extra parameters to read input from a file, or about anything else
05:04.55 MarcTannous brlcad: please tell me if this one layout is okay http://www.google-melange.com/gci/task/view/google/gci2014/5256853575434240 so I can make the others ( like you said yesterday )
05:06.48 andromeda-galaxy Stragus: ah, okay... No harm in allowing it but not using it, I guess.
05:11.44 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
05:18.21 *** join/#brlcad ofnlut_ (~ofnlut_@c-68-48-58-20.hsd1.dc.comcast.net)
05:19.45 *** join/#brlcad ofnlut_ (~ofnlut_@c-68-48-58-20.hsd1.dc.comcast.net)
05:19.53 andromeda-galaxy Stragus: in that manner, we might as well make it take char ***argv as well, in case we need to e.g. skip a few arguments or totally reassign argv...
05:19.59 andromeda-galaxy does takingthe address of argv work?
05:22.11 Stragus You'll have to take the address of argv if you want to realloc and add arguments... Just replacing or removing arguments could use the existing array
05:23.40 Notify 02GCI:reduceqq * 6366650240598016 Animate our logo! #8: Ready for review - The work on this task is ready to be reviewed.
05:23.41 Notify 02GCI:brlcad * 5256853575434240 Create a logo press kit #5: looking good - This is looking really good, much better and more flexibly useful. As for your model, is there a specific reason you...
05:23.50 Notify 02GCI:brlcad * 5256853575434240 Create a logo press kit #5: 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...
05:26.06 Notify 02GCI:adityagulati * 6082457623330816 Model BRL-CAD logo in BRL-CAD #3: Ready for review - The work on this task is ready to be reviewed.
05:27.41 Notify 02GCI:deepakkumarsharma * 4956557716488192 Fix at least 8 spelling mistakes in at least 8 different files #5: Task Assigned - This task has been assigned to spawnflames. You have 100 hours to...
05:29.11 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
05:30.23 andromeda-galaxy Stragus: right... I couldn't figure out according to thes tandard if that is allowed
05:30.26 andromeda-galaxy do you know?
05:30.52 Stragus argc and argv are writable, yes
05:31.04 *** join/#brlcad adityagulati (dce32db2@gateway/web/cgi-irc/kiwiirc.com/ip.220.227.45.178)
05:31.14 adityagulati hi
05:31.26 andromeda-galaxy I know that they're writable, but does the spec say that they have to be addressable?
05:32.25 Stragus I believe argv[i] is valid by the standard, and argv[i][j] might be undefined
05:32.41 andromeda-galaxy I meat, does &argv have to be defined?
05:33.00 andromeda-galaxy *meant
05:33.06 Stragus Of course, it's a variable like any other
05:33.47 Notify 02GCI:brlcad * 4959933829218304 Create a logo press kit #3: Task Closed - Congratulations, this task has been completed successfully.
05:35.18 andromeda-galaxy Stragus: okay, good! I just found out that [] binds more tightly than *
05:35.30 andromeda-galaxy i.e. (*argv)[*argc] and *argv[*argc] mean subtly different things
05:35.39 andromeda-galaxy (different enough to segfault all the tests!)
05:37.03 Notify 02GCI:brlcad * 4959933829218304 Create a logo press kit #3: some tweaks - Hi hackerboy, there were tweaks that I normally would have insisted on, but you clearly already put a lot of time and...
05:37.53 Notify 02GCI:brlcad * 5818109499801600 Design a BRL-CAD Sticker #14: Task Closed - Congratulations, this task has been completed successfully.
05:39.37 Stragus :) Yes, operator precedence is a tricky thing, also when you want to declare an array of const function pointers and such
05:39.44 Notify 02GCI:spawnflames * 4956557716488192 Fix at least 8 spelling mistakes in at least 8 different files #5: Using SVN - Do you need to have an SVN Client to do this task?
05:41.30 Notify 02GCI:brlcad * 4954632228962304 Fix any failing unit test #7: Task Closed - Congratulations, this task has been completed successfully.
05:43.39 Notify 02GCI:brlcad * 6356614713966592 Design a BRL-CAD Sticker #11: Task Closed - Congratulations, this task has been completed successfully.
05:44.05 Notify 02GCI:tannousmarc * 5256853575434240 Create a logo press kit #5: Lowered down the pegs size - Lowered down the pegs size to fit a requirement I saw on some other task, somebody pointing out an...
05:45.00 MarcTannous by the way, for anyone that wants to work on fixing unit tests : https://docs.google.com/spreadsheets/d/1DlCGUQqUMsIh1Hnn7TljnAUcWYmypKQycPk_VzBT-w4/edit?usp=sharing
05:46.59 Notify 02GCI:o7p9bxbnyj * 5814267383119872 Eliminate any library global variable #3: Task Claimed - I would like to work on this task.
05:47.44 andromeda-galaxy Stragus: indeed :-)
05:48.54 Notify 02GCI: * : -
05:49.09 MarcTannous I think notify just failed
05:49.46 *** join/#brlcad andrei_ (c35a6e7d@gateway/web/freenode/ip.195.90.110.125)
05:49.58 andrei_ hello
05:50.02 MarcTannous morning
05:50.26 YashM hi
05:51.39 MarcTannous andromeda-galaxy: are you going to submit a patch with the if/else for bu_semaphore?
05:51.50 MarcTannous andromeda-galaxy: or should I?
05:53.13 *** join/#brlcad YashM (~YashM@59.88.27.35)
05:54.16 andromeda-galaxy MarcTannous: I'll do it, if you don't mind...
05:54.39 Notify 02GCI:popescuandrei * 5898744557993984 Create a logo press kit: None - Hey, monimont! Seems you have completed the task requirements, well done! Thanks for your effort and hopefully we ll see...
05:54.45 andrei_ what's with the bu_semaphore :o
05:54.45 Notify 02GCI:popescuandrei * 5898744557993984 Create a logo press kit: Task Closed - Congratulations, this task has been completed successfully.
05:54.53 andromeda-galaxy MarcTannous: rather, I'll talk to brlcad and do it if he agrees that's a good way to do it...
05:55.02 MarcTannous he didn't give a reply last night?
05:55.08 andromeda-galaxy I don't believe so...
05:55.31 MarcTannous andrei_: we created a list of tasks failing on 32bits/64bits because there were a few that failed just because of the system difference
05:55.50 MarcTannous andrei_: bu_semaphore in particular standed out because it always stopped the make test for a few minutes, as it had 1000 reps
05:55.51 andromeda-galaxy andrei_: two of the semaphor tests fail on platforms with <3 cpus
05:56.14 andromeda-galaxy it's easy to see looking at the test output log, bu_parallel() errors out
05:56.28 MarcTannous andrei_: we looked into the source code and the issue was the VM being single-cored, the easiest fix would just be to count the test as a success if the actualcores are lower than expectedcores
05:56.43 andrei_ that's weird but if bu_semaphore behaves similar to an OS semaphore
05:57.00 Notify 02GCI:brlcad * 5898744557993984 Create a logo press kit: wrong logo version - Monimont, You used the wrong version of the logo. Note that the task description says to look at any of the completed...
05:57.10 andromeda-galaxy andrei_: there are two tests run for each P-value, one for sequential modification and one in parallel
05:57.32 andromeda-galaxy the parallel one uses bu_parallel to do the parallel modifications, but bu_parallel bombs when you ask it to be more concurrent than ncpus
05:57.45 andrei_ brlcad: sorry about monimont's task, I didn't look carefully at the logo
05:58.04 andromeda-galaxy oooh! aactually, there's a better way to fix it...
05:58.25 Notify 02GCI:brlcad * 4949440150372352 Design a BRL-CAD splash screen #4: Task Closed - Congratulations, this task has been completed successfully.
05:58.26 andromeda-galaxy maybe...
05:58.40 Notify 02GCI:tannousmarc * 5256853575434240 Create a logo press kit #5: Ready for review - The work on this task is ready to be reviewed.
05:58.54 andromeda-galaxy actually, that doesn't work, it requires that the test be run with bu built in debug mode
05:59.05 Notify 02GCI:brlcad * 6620713125412864 Write manual page documentation (for archer): Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this...
06:01.06 Notify 02GCI:ronaldoh373 * 6648047639461888 Design a BRL-CAD Sticker #10: None - Files I have uploaded now- "brl cad sticker.psd" "brl cad sticker copy.jpg"
06:01.11 Notify 02GCI:ronaldoh373 * 6648047639461888 Design a BRL-CAD Sticker #10: Ready for review - The work on this task is ready to be reviewed.
06:03.36 Notify 02GCI:brlcad * 6620713125412864 Write manual page documentation (for archer): no idea - Parth, I have no idea what you submitted, but it has absolutely nothing to do with BRL-CAD. It's clearly...
06:03.41 andromeda-galaxy or maybe it does... testing now
06:06.39 andromeda-galaxy MarcTannous: I may have figured out a better way to make everything work without ever disabling tests, if it works do you mind if I submit it?
06:06.50 MarcTannous go ahead
06:07.57 Notify 02GCI:mukulkamboj * 6353662997692416 Design a BRL-CAD splash screen #3: Ready for review - The work on this task is ready to be reviewed.
06:08.09 andromeda-galaxy MarcTannous: great, I will.
06:09.02 Notify 02GCI:juskirat2000 * 6268869404524544 Design a prototype CAD GUI layout #9: Task Claimed - I would like to work on this task.
06:09.48 Notify 02GCI:mukulkamboj * 6353662997692416 Design a BRL-CAD splash screen #3: sorry - I'm very very sorry for this and i am confident enough to say that you will mark this task as complete. thank you
06:10.43 Notify 02GCI:brlcad * 5834136975572992 Design a web page that overviews the BRL-CAD contributors #2: blank page in safari - Zenon, this is looking good, but results in a blank page in Safari (rather, it...
06:10.44 Notify 02GCI:brlcad * 5834136975572992 Design a web page that overviews the BRL-CAD contributors #2: Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s)...
06:11.58 Notify 02GCI:brlcad * 4938019295461376 Design a new website favicon! #10: Deadline extended - The deadline of the task has been extended with 2 days and 2 hours.
06:14.37 Notify 02GCI:brlcad * 4938019295461376 Design a new website favicon! #10: unable? - Unable to render a cube? That's like the very first tutorial and we provide a virtual machine where BRL-CAD is already...
06:14.48 Notify 02GCI:brlcad * 4938019295461376 Design a new website favicon! #10: 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...
06:15.18 Notify 02GCI:brlcad * 5814267383119872 Eliminate any library global variable #3: Task Assigned - This task has been assigned to Andromeda Galaxy. You have 100 hours to complete this task, good luck!
06:15.19 Notify 02GCI:brlcad * 6268869404524544 Design a prototype CAD GUI layout #9: Task Assigned - This task has been assigned to JASKIRAT SINGH. You have 100 hours to complete this task, good luck!
06:16.38 Notify 02GCI:o7p9bxbnyj * 5814267383119872 Eliminate any library global variable #3: Ready for review - The work on this task is ready to be reviewed.
06:20.19 Notify 02GCI:brlcad * 6366650240598016 Animate our logo! #8: Task Closed - Congratulations, this task has been completed successfully.
06:20.24 Notify 02GCI:popescuandrei * 6082457623330816 Model BRL-CAD logo in BRL-CAD #3: None - Can you please upload the .g as well? Thank you!
06:20.34 Notify 02GCI:popescuandrei * 6082457623330816 Model BRL-CAD logo in BRL-CAD #3: 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...
06:23.21 Notify 02GCI:brlcad * 6082457623330816 Model BRL-CAD logo in BRL-CAD #3: hairline - Aditya, Is the hairline fracture on the logo's upper/center red link intentional? It's hard to tell and it makes the...
06:23.51 Notify 02GCI:adityagulati * 6082457623330816 Model BRL-CAD logo in BRL-CAD #3: None - Sir, I have uploaded The .g file By the name ,"logo1.g". It Was There From The Start. Regards, Aditya
06:24.01 Notify 02GCI:adityagulati * 6082457623330816 Model BRL-CAD logo in BRL-CAD #3: Ready for review - The work on this task is ready to be reviewed.
06:25.42 Notify 02GCI:popescuandrei * 5256853575434240 Create a logo press kit #5: Task Closed - Congratulations, this task has been completed successfully.
06:27.37 Notify 02GCI:adityagulati * 6082457623330816 Model BRL-CAD logo in BRL-CAD #3: None - Sir, I Have Provided The Hairline Fracture To Give A Sense Of Depth, It Makes The Logo Look More Real. Regards, ...
06:28.47 Notify 02GCI:brlcad * 5256853575434240 Create a logo press kit #5: 64 - Looks like the 64 size seems to be clipped on the right side, but otherwise, these look great!
06:29.33 Notify 03BRL-CAD:peter-sa * 63700 (brlcad/trunk/src/libbu/CMakeLists.txt brlcad/trunk/src/libbu/date-time.c): Apply GCI task http://www.google-melange.com/gci/task/view/google/gci2014/4954632228962304 that makes bu_utctime() use the 64-bit time functions from https://github.com/schwern/y2038, allowing the bu_date-time tests to pass on architectures that use 32-bit time_t values.
06:34.40 Notify 02GCI:tannousmarc * 6449379632218112 Create an utility library (LIBBU) API unit test for bomb.c: Task Claimed - I would like to work on this task.
06:35.40 Notify 02GCI:popescuandrei * 6449379632218112 Create an utility library (LIBBU) API unit test for bomb.c: Task Assigned - This task has been assigned to Marc Tannous. You have 100 hours to complete this...
07:07.21 Notify 02GCI:amshenoy * 4959933829218304 Create a logo press kit #3: Thank You - I will keep your comments in mind for my future tasks... Thank You
07:10.52 Notify 02GCI:icyjuban * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: Ready for review - The work on this task is ready to be reviewed.
07:12.02 Notify 02GCI:juskirat2000 * 6268869404524544 Design a prototype CAD GUI layout #9: Ready for review - The work on this task is ready to be reviewed.
07:17.06 Notify 02GCI:adityagulati * 6268869404524544 Design a prototype CAD GUI layout #9: Sorry For Interrupting - Hi Jaskirat, Please Go View My Task That I Made For The Same(Or You Might Have Already Been...
07:49.59 Notify 02GCI:Melange * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: No more Work can be submitted - Melange has detected that the deadline has passed and no more work can be...
07:50.54 Notify 02GCI:mehak_s * 6359310477033472 Design a banner for our Facebook page (and/or other social media sites) #3: Task Claimed - I would like to work on this task.
08:02.09 YashM What does this error mean in Archer? db5_crack_disk_header() bad magic1 -- database has become corrupted
08:02.09 YashM <PROTECTED>
08:03.57 Stragus The structs store some magic integer as first member, a bad value means something else wrote over it
08:04.05 Stragus So, a bug
08:07.53 YashM Stragus, so what can be done?
08:08.19 YashM I get this error when raytracing
08:09.29 Stragus There are a few options. Have you used Valgrind before? (are you on Linux?)
08:09.56 YashM Nope windows
08:10.20 Notify 02GCI:arnav * 5230915127083008 Compile BRL-CAD with GCC trunk: Claim Removed - The claim on this task has been removed, someone else can claim it now.
08:10.36 YashM insure++?
08:10.56 YashM or dr memory or something
08:11.46 YashM Its not even my .g file so i dont know why i am getting this error and not the person who made it
08:11.46 Stragus Hum, I don't know what tools are available on Windows. You need a memory debugger of some sort
08:11.55 Stragus A standard debugger probably won't do any good
08:14.34 *** join/#brlcad mihaineacsu (~mihaineac@92.81.146.140)
08:16.40 YashM i should dl dr memory then
08:17.20 Notify 02GCI:deepakkumarsharma * 6359310477033472 Design a banner for our Facebook page (and/or other social media sites) #3: Task Assigned - This task has been assigned to Mehak Sharma. You have 100 hours...
08:22.20 Notify 02GCI:mihaineacsu29 * 6082457623330816 Model BRL-CAD logo in BRL-CAD #3: Task Closed - Congratulations, this task has been completed successfully.
08:24.41 Notify 02GCI:Melange * 5221860329390080 Design a new website favicon! #8: Task Reopened - Melange has detected that the final deadline has passed and it has reopened the task.
08:25.17 Notify 02GCI:adityagulati * 6643302640123904 Create a BRL-CAD style guide / cheat sheet #2: Task Claimed - I would like to work on this task.
08:25.50 adityagulati could someone tell me the process name for archer in windows task manager
08:26.48 YashM bwish.exe?
08:27.42 adityagulati Yeah it is there, Thanks
08:30.02 adityagulati My Archer Application has went 'not responding' in windows, is there any way to retrieve the file or make the application respond?
08:38.11 Notify 02GCI:arnav * 5765799415382016 Create a BRL-CAD presentation template: Task Claimed - I would like to work on this task.
08:42.28 *** join/#brlcad YashM (~YashM@59.95.37.209)
08:42.39 Notify 02GCI:mihaineacsu29 * 6082457623330816 Model BRL-CAD logo in BRL-CAD #3: None - Very nice work, Aditya! Congrats!
08:49.04 *** join/#brlcad YashM_ (~YashM@117.223.173.12)
08:50.22 Notify 02GCI:thevk * 5627897880510464 Model a toy tank in 3D using BRL-CAD #5: Time - Well, i spent a couple of hours reading the guide and something like 5 hrs to model the tank. I'm new to 3D modeling,...
08:50.43 Notify 02GCI:juskirat2000 * 6268869404524544 Design a prototype CAD GUI layout #9: None - The work on this task is ready to be reviewed.
08:54.21 Notify 02GCI:thevk * 5578661449170944 Model BRL-CAD logo in BRL-CAD #5: Task Claimed - I would like to work on this task.
08:56.57 YashM_ In archer, my object has stopped moving but when i raytrace it, it has moved
09:01.40 Notify 02GCI:harman052 * 5578661449170944 Model BRL-CAD logo in BRL-CAD #5: Task Assigned - This task has been assigned to Vladimir Kuznetsov. You have 100 hours to complete this task, good luck!
09:03.32 Notify 02GCI:mihaineacsu29 * 6268869404524544 Design a prototype CAD GUI layout #9: Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned to this...
09:04.32 *** join/#brlcad ries (~ries@D979C47E.cm-3-2d.dynamic.ziggo.nl)
09:04.54 Notify 02GCI:harman052 * 5765799415382016 Create a BRL-CAD presentation template: Task Assigned - This task has been assigned to Arnav. You have 100 hours to complete this task, good luck!
09:05.24 Notify 02GCI:harman052 * 6643302640123904 Create a BRL-CAD style guide / cheat sheet #2: Task Assigned - This task has been assigned to Aditya Gulati. You have 100 hours to complete this task, good luck!
09:08.09 Notify 02GCI:mihaineacsu29 * 6268869404524544 Design a prototype CAD GUI layout #9: Disappointed - Jaskirat, I believe this task is a great opportunity to come up with something original and creative....
09:11.05 Notify 02GCI:juskirat2000 * 6268869404524544 Design a prototype CAD GUI layout #9: Claim Removed - The claim on this task has been removed, someone else can claim it now.
09:14.35 Notify 02GCI:juskirat2000 * 5221860329390080 Design a new website favicon! #8: Task Claimed - I would like to work on this task.
09:16.56 Notify 02GCI:popescuandrei * 5221860329390080 Design a new website favicon! #8: Task Assigned - This task has been assigned to JASKIRAT SINGH. You have 100 hours to complete this task, good luck!
09:20.05 Notify 02GCI:popescuandrei * 5814267383119872 Eliminate any library global variable #3: Task Closed - Congratulations, this task has been completed successfully.
09:20.53 *** join/#brlcad merzo (~merzo@user-94-45-58-138-1.skif.com.ua)
09:24.44 Notify 02GCI:popescuandrei * 6648047639461888 Design a BRL-CAD Sticker #10: Task Closed - Congratulations, this task has been completed successfully.
09:26.14 Notify 02GCI:juskirat2000 * 5221860329390080 Design a new website favicon! #8: Ready for review - The work on this task is ready to be reviewed.
09:31.27 Notify 02GCI:jatish * 6178630262587392 Design a banner for our Facebook page (and/or other social media sites) #6: Claim Removed - The claim on this task has been removed, someone else can claim it now.
09:31.28 Notify 02GCI:jatish * 6178630262587392 Design a banner for our Facebook page (and/or other social media sites) #6: Task Claimed - I would like to work on this task.
09:37.05 Notify 02GCI:harman052 * 6178630262587392 Design a banner for our Facebook page (and/or other social media sites) #6: Task Assigned - This task has been assigned to jatish. You have 100 hours to complete...
09:37.15 Notify 02GCI:mihaineacsu29 * 5221860329390080 Design a new website favicon! #8 http://www.google-melange.com/gci/task/view/google/gci2014/5221860329390080: None Jaskirat, This is the second time...
09:37.54 Notify 02GCI:mihaineacsu29 * 5221860329390080 Design a new website favicon! #8: Task Reopened - This task has been Reopened.
09:38.54 *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
09:48.09 Notify 02GCI:mehak_s * 6359310477033472 Design a banner for our Facebook page (and/or other social media sites) #3: Ready for review - The work on this task is ready to be reviewed.
09:50.51 Notify 02GCI:adityagulati * 6643302640123904 Create a BRL-CAD style guide / cheat sheet #2: Ready for review - The work on this task is ready to be reviewed.
10:05.11 *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
10:06.54 Notify 02GCI:ishwerdas * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: None - Hey Olivia, Can you please upload the entire code in a zip / .tar.gz format or may be on github...
10:07.09 Notify 02GCI:ishwerdas * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: Deadline extended - The deadline of the task has been extended with 0 days and 20 hours.
10:07.19 Notify 02GCI:ishwerdas * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s)...
10:08.24 Notify 02GCI:mihaineacsu29 * 6643302640123904 Create a BRL-CAD style guide / cheat sheet #2: None - Fantastic work, Aditya! However try to center the file formats columns. Make sure the line dividing the...
10:08.34 Notify 02GCI:mihaineacsu29 * 6643302640123904 Create a BRL-CAD style guide / cheat sheet #2: Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned...
10:14.15 Notify 02GCI:adityagulati * 6643302640123904 Create a BRL-CAD style guide / cheat sheet #2: Ready for review - The work on this task is ready to be reviewed.
10:20.37 Notify 02GCI:ishwerdas * 6359310477033472 Design a banner for our Facebook page (and/or other social media sites) #3: Review for 01.jpg - This review is on 01.jpg. I appreciate your efforts but the text...
10:20.52 Notify 02GCI:ishwerdas * 6359310477033472 Design a banner for our Facebook page (and/or other social media sites) #3: Task Needs More Work - One of the mentors has sent this task back for more work. Talk...
10:22.42 Notify 02GCI:parthmiglani_21 * 6620713125412864 Write manual page documentation (for archer): Claim Removed - The claim on this task has been removed, someone else can claim it now.
10:25.43 Notify 02GCI:mihaineacsu29 * 6643302640123904 Create a BRL-CAD style guide / cheat sheet #2: Task Closed - Congratulations, this task has been completed successfully.
10:29.03 Notify 02GCI:ishwerdas * 6643302640123904 Create a BRL-CAD style guide / cheat sheet #2: None - I like your design, just a tip for your future tasks, try to have bit more contrast especially in headings....
10:29.54 *** join/#brlcad merzo (~merzo@user-94-45-58-138-1.skif.com.ua)
10:30.03 Notify 02GCI:rohit_agarwal * 5564446986469376 Create a BRL-CAD style guide / cheat sheet #3: Claim Removed - The claim on this task has been removed, someone else can claim it now.
10:30.18 Notify 02GCI:adityagulati * 5887275384700928 Model "BRL-CAD" in BRL-CAD: Task Claimed - I would like to work on this task.
10:31.03 Notify 02GCI:rohit_agarwal * 5564446986469376 Create a BRL-CAD style guide / cheat sheet #3: None - I have found another task that I would like to do first. I will soon claim this task again after...
10:31.13 Notify 02GCI:rohit_agarwal * 5221860329390080 Design a new website favicon! #8: Task Claimed - I would like to work on this task.
10:31.43 Notify 02GCI:ishwerdas * 6353662997692416 Design a BRL-CAD splash screen #3: 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...
10:31.53 Notify 02GCI:mihaineacsu29 * 5221860329390080 Design a new website favicon! #8: Task Assigned - This task has been assigned to Rohit Agarwal. You have 100 hours to complete this task, good luck!
10:34.49 Notify 02GCI:ishwerdas * 6353662997692416 Design a BRL-CAD splash screen #3 http://www.google-melange.com/gci/task/view/google/gci2014/6353662997692416: None I really appreciate your efforts but please...
10:35.54 Notify 02GCI:ishwerdas * 5887275384700928 Model "BRL-CAD" in BRL-CAD: Task Assigned - This task has been assigned to Aditya Gulati. You have 100 hours to complete this task, good luck!
10:42.51 Notify 02GCI:adityagulati * 5887275384700928 Model "BRL-CAD" in BRL-CAD: Ready for review - The work on this task is ready to be reviewed.
10:53.33 *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
10:55.53 *** join/#brlcad ries (~ries@D979C47E.cm-3-2d.dynamic.ziggo.nl)
11:17.31 *** join/#brlcad adityagulati (dce32db2@gateway/web/cgi-irc/kiwiirc.com/ip.220.227.45.178)
11:17.43 adityagulati hi
11:24.19 adityagulati could someone tell me, Whether in the following task do we have to make our own content
11:24.24 adityagulati https://www.google-melange.com/gci/task/view/google/gci2014/5863782450462720
11:24.39 adityagulati or copy it from somewhere
11:25.02 adityagulati By content I mean Text not Graphics
11:31.41 adityagulati ????
11:35.32 Notify 02GCI:rohit_agarwal * 5221860329390080 Design a new website favicon! #8: Ready for review - The work on this task is ready to be reviewed.
11:36.17 Notify 02GCI:rohit_agarwal * 5221860329390080 Design a new website favicon! #8: Files Uploaded - I have uploaded two files - Favicon 2.xcf/.gif
11:44.40 *** join/#brlcad Rohit (~Ravi@49.249.53.22)
11:48.52 *** join/#brlcad rohitagarwal (~Ravi@49.249.53.22)
11:49.55 rohitagarwal Is any mentor online
11:51.17 *** part/#brlcad rohitagarwal (~Ravi@49.249.53.22)
12:11.22 adityagulati could someone tell me, Whether in the following task do we have to make our own text or copy it from somewhere?
12:11.31 adityagulati https://www.google-melange.com/gci/task/view/google/gci2014/5863782450462720
12:29.08 Notify 02GCI:mukulkamboj * 6353662997692416 Design a BRL-CAD splash screen #3: Ready for review - The work on this task is ready to be reviewed.
12:42.23 *** join/#brlcad mandarj (~mandar@117.195.61.186)
12:43.15 mandarj I am writing the docbook xml for the g-step converter (GCI task at "https://www.google-melange.com/gci/task/view/google/gci2014/5001720170020864")
12:43.46 mandarj But I dont quite understand what g-step does and what to write in the manual
12:43.59 mandarj Could anybody please help me?
12:47.18 mandarj ?
12:53.09 mandarj ?
12:56.26 *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
13:01.46 Notify 02GCI:mjuvekar7 * 5001720170020864 Write manual page documentation (for g-step): None - Can you explain what the g-step command actually does (I don't quite understand)?
13:02.58 ``Erik starseeker: isn't ther cmake fu in place to prevent BRL-CAD from using tk86?
13:03.20 adityagulati could someone tell me, Whether in the following task do we have to make our own text or copy it from somewhere?
13:03.29 adityagulati https://www.google-melange.com/gci/task/view/google/gci2014/5863782450462720
13:03.37 ``Erik starseeker: I'm looking at https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=195885
13:05.01 *** join/#brlcad d_rossberg (~rossberg@66-118-151-70.static.sagonet.net)
13:06.19 ``Erik adityagulati: you can use parts from other sources as long as they're properly credited (preferably on page credit as well as a link to the original source), but I think most of the content will have to be original to make it concise and applicable...
13:10.53 mandarj starseeker: are you there?
13:16.15 *** join/#brlcad andrei_ (c35a6e7d@gateway/web/freenode/ip.195.90.110.125)
13:16.41 mandarj can anyone tell me what the g-step command does?
13:17.59 Notify 02GCI:ishwerdas * 5221860329390080 Design a new website favicon! #8: needs to be more innovative - I like your effort but we already had rotating brl-cad logo as a favicon submission. You may want...
13:18.09 Notify 02GCI:ishwerdas * 5221860329390080 Design a new website favicon! #8: 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...
13:19.04 Notify 02GCI:juskirat2000 * 5221204004700160 Design a prototype CAD GUI layout #3: Task Claimed - I would like to work on this task.
13:19.28 *** join/#brlcad mandarj_ (~mandar@117.195.33.84)
13:21.57 *** join/#brlcad MarcTannous (5679c4d5@gateway/web/freenode/ip.86.121.196.213)
13:22.04 MarcTannous hello
13:22.18 adityagulati hi
13:24.44 *** join/#brlcad gjeet (3b5b7397@gateway/web/cgi-irc/kiwiirc.com/ip.59.91.115.151)
13:26.10 Notify 02GCI:gjeet * 5221204004700160 Design a prototype CAD GUI layout #3: Task Assigned - This task has been assigned to JASKIRAT SINGH. You have 100 hours to complete this task, good luck!
13:31.56 *** join/#brlcad ries (~ries@D979C47E.cm-3-2d.dynamic.ziggo.nl)
13:34.16 Notify 02GCI:juskirat2000 * 5221204004700160 Design a prototype CAD GUI layout #3: Ready for review - The work on this task is ready to be reviewed.
13:35.44 andrei_ Marc: isn't this your work : http://www.google-melange.com/gci/task/view/google/gci2014/5221204004700160 ?
13:36.22 YashM_ andrei_ its the third time he submitted someone else's work
13:36.49 MarcTannous andrei_ : yep that's it
13:36.58 YashM_ check these: 5221860329390080 6268869404524544
13:37.09 MarcTannous he didn't change anything
13:37.13 MarcTannous ....
13:37.59 gjeet Let me have a look
13:38.07 YashM_ someone ban this guy
13:38.13 MarcTannous let me have a look?
13:38.42 MarcTannous Mate, this is theft
13:40.21 gjeet Marc: yeah I understand that, and proper action would be taken for sure. I was just checking out the work he submitted and comparing it with one you gave.
13:40.33 andrei_ It's his third attempt
13:40.38 andrei_ I ve removed his claim
13:40.43 Notify 02GCI:popescuandrei * 5221204004700160 Design a prototype CAD GUI layout #3: Task Reopened - This task has been Reopened.
13:40.56 MarcTannous gjeet : you'd have to compare it if he made any modification
13:41.02 MarcTannous he literally took my PSD and uploaded it as his
13:41.08 gjeet Oh !
13:41.13 andrei_ YashM: did you find any forges, or what?
13:41.27 MarcTannous yes, the ones yashM submitted are also forges
13:41.42 YashM_ yeah
13:41.46 mandarj_ can anyone please tell me what g-step does?
13:41.50 YashM_ those are the other two
13:42.02 andrei_ http://www.google-melange.com/gci/task/view/google/gci2014/5221860329390080
13:42.04 andrei_ this is a forge?
13:42.05 andrei_ to which?
13:42.27 andrei_ Melange does not have a ban capability, unfortunately
13:42.39 YashM_ Check Mihai's comment
13:42.43 YashM_ it says it is http://www.google-melange.com/gci/task/view/google/gci2014/5221860329390080
13:42.48 andrei_ but me, mihai and gjeet are aware of it
13:42.52 ignacio For GCI Users: The leaderboard now update every 10 minutes.
13:42.53 mandarj_ ?
13:43.08 *** join/#brlcad gcibot (~gcibot@unaffiliated/ignaciouy/bot/gcibot)
13:43.12 ignacio For GCI Users: The leaderboard now update every 10 minutes.
13:43.23 d_rossberg mandarj_: it should convert a BRL-CAD database into STEP format, an industry standard CAD format
13:43.38 andrei_ We ll start looking into forges and remove claims
13:43.38 Notify 02GCI:monimont * 6396924525543424 Compile BRL-CAD using Microsoft Visual Studio Express 2013: Task Claimed - I would like to work on this task.
13:44.35 MarcTannous andrei_ thank you
13:44.46 mandarj_ d_rossberg: a BRL-CAD database has a file extension of .g right?
13:44.50 adityagulati 5107273286287360 5170304011730944 , these two are also the same and both have been approved
13:45.49 MarcTannous mandarj_: yes
13:45.54 Notify 02GCI:juskirat2000 * 5221204004700160 Design a prototype CAD GUI layout #3: Task Claimed - I would like to work on this task.
13:46.29 Notify 02GCI:juskirat2000 * 5221204004700160 Design a prototype CAD GUI layout #3: None - sir why you unclaim this task
13:46.44 Notify 02GCI:gjeet * 5221204004700160 Design a prototype CAD GUI layout #3: Task Reopened - This task has been Reopened.
13:46.55 andrei_ adityagulati: I apologize. We ve been trying to get a response to you students as fast as possible, we didn't suspect this issue, but I can assureit will be looked into
13:47.05 mandarj_ MarcTannous: what are the options for the g-step command? Where can I find a list of them?
13:47.09 Notify 02GCI:rossberg * 6396924525543424 Compile BRL-CAD using Microsoft Visual Studio Express 2013: Task Assigned - This task has been assigned to monimont. You have 100 hours to complete this task, good...
13:47.20 andrei_ I ve got a meeting, but stay tuned for updates
13:48.03 mihaineacsu_away unfreaking believable
13:48.30 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
13:48.37 d_rossberg andarj_: type g-step without parameters or 'brlman g-step'
13:49.06 adityagulati 6362007112515584 5186134254551040, these are some forges two one is mine the other one has been copied from my task
13:49.14 d_rossberg (sorry: mandarj_, copy-n-paste error)
13:51.11 mihaineacsu adityagulati: on it
13:51.15 Notify 02GCI:gjeet * 5221204004700160 Design a prototype CAD GUI layout #3: None - Here's the thing, it's third time it's been reported that you submitted someone else's work, pretending that you did it....
13:52.36 Notify 02GCI:juskirat2000 * 5032082837340160 Design a prototype CAD GUI layout #4: Task Claimed - I would like to work on this task.
13:53.05 YashM_ Okay guys, we have a serious troller here
13:53.27 mihaineacsu mhm
13:56.21 Notify 02GCI:gjeet * 5032082837340160 Design a prototype CAD GUI layout #4: None - Hey Jaskirat kindly read my comment on your recent previous task.
13:57.16 Notify 02GCI:mjuvekar7 * 5001720170020864 Write manual page documentation (for g-step): Claim Removed - The claim on this task has been removed, someone else can claim it now.
13:57.31 Notify 02GCI:mjuvekar7 * 5924891446476800 Write manual page documentation (for g-voxel): Task Claimed - I would like to work on this task.
13:57.37 *** join/#brlcad merzo (~merzo@user-94-45-58-138-1.skif.com.ua)
13:58.31 Notify 02GCI:gjeet * 5924891446476800 Write manual page documentation (for g-voxel): Task Assigned - This task has been assigned to mandarj. You have 100 hours to complete this task, good luck!
13:59.39 mandarj_ whois
14:01.22 Notify 02GCI:mjuvekar7 * 5924891446476800 Write manual page documentation (for g-voxel): None - What does the g-voxel command do? I went over the .c file but I could not understand what the g-voxel...
14:01.27 Notify 02GCI:juskirat2000 * 5032082837340160 Design a prototype CAD GUI layout #4: None - what was it
14:01.43 mandarj_ I changed my task as the file I was supposed to make (g-step.xml) already exists. Please look into the matter.
14:03.02 Notify 02GCI:mjuvekar7 * 5001720170020864 Write manual page documentation (for g-step): File already exists! - I have removed my claim from this task as the file I was supposed to make (g-step.xml)...
14:03.03 Notify 02GCI:juskirat2000 * 5032082837340160 Design a prototype CAD GUI layout #4: Claim Removed - The claim on this task has been removed, someone else can claim it now.
14:03.23 d_rossberg mandarj_: maybe there is enough room to improve the g-step manual page
14:04.25 mandarj_ d_rossberg: the man page seems to be quite self-explanatory and seems to contain everything (I) can understand from the code
14:06.47 *** join/#brlcad winghouchan (~holoirc@genkt-056-125.t-mobile.co.uk)
14:07.36 *** part/#brlcad mandarj_ (~mandar@117.195.33.84)
14:07.42 *** join/#brlcad mandarj_ (~mandar@117.195.33.84)
14:08.01 mandarj_ gjeet: are you there?
14:08.12 gjeet yeah
14:08.41 mandarj_ I am doing the task at https://www.google-melange.com/gci/task/view/google/gci2014/5924891446476800
14:08.42 gcibot Write manual page documentation (for g-voxel)
14:08.42 gcibot Status: Claimed (4 hours, 3 minutes left)
14:08.42 gcibot Mentor(s): Harmanpreet, Ishwerdas
14:09.27 YashM_ anyone know if new tasks coming in?
14:09.38 mandarj_ I cannot understand I cannot understand what the g-voxel command does from the code. Could you please explain?
14:09.54 Notify 02GCI:icyjuban * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: None - Yes, the first link contains my latest work. And I'll submit a tar.gz format right now.
14:10.09 mandarj_ (I am asking you as you were the one who assigned the task to me)
14:11.14 Notify 02GCI:icyjuban * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: Ready for review - The work on this task is ready to be reviewed.
14:12.31 gjeet mandarj_: Well sorry to say but I don't know about these commands, I guess you need someone better than me for your query.
14:13.04 mandarj_ gjeet: ok
14:13.07 *** part/#brlcad mandarj_ (~mandar@117.195.33.84)
14:13.12 ``Erik mandarj_: it voxelizes the implicit geometry... part of the task is the ability to read the source and understand the meaning of the arguments. Is this a task...
14:13.34 ``Erik impatience. *sigh*
14:15.21 gjeet Erik: He raised same query on melange, so It would be better if you answer it there too https://www.google-melange.com/gci/task/view/google/gci2014/5924891446476800
14:15.22 gcibot Write manual page documentation (for g-voxel)
14:15.22 gcibot Status: Claimed (4 hours, 3 minutes left)
14:15.22 gcibot Mentor(s): Harmanpreet, Ishwerdas
14:16.57 ``Erik melange sucks, I might address it later, but I have no interest in dealing with the melange interface right now :) other stuff to do
14:17.19 d_rossberg i'll answer his question on melange
14:18.05 Notify 02GCI:shardulc * 6423086379302912 Write manual page documentation (for g-dot): Ready for review - The work on this task is ready to be reviewed.
14:23.33 *** join/#brlcad luca79 (~luca@net-37-116-120-48.cust.vodafonedsl.it)
14:31.59 Notify 02GCI:rossberg * 5924891446476800 Write manual page documentation (for g-voxel): None - g-voxel generates a list of voxels (fat points, you may see them as cubes) from a .g file and writes them in...
14:32.54 Notify 02GCI:ljudusika * 4797643658100736 Model "BRL-CAD" in BRL-CAD #2: How to render? - I have created this file (sg). But I can not to render with comand rt -A0.75 -s1024 -c {set ambSamples=256} I...
14:39.00 Notify 02GCI:mjuvekar7 * 5924891446476800 Write manual page documentation (for g-voxel): None - what options can it take?
14:47.44 YashM_ I made something in MGED then imported to Archer, when i right click object and then "Shaded", Archer closes without error
14:47.57 Notify 02GCI:mjuvekar7 * 5924891446476800 Write manual page documentation (for g-voxel): None - also, if possible, could you give an example?
14:50.24 ``Erik in the terminal that launched archer, without running anything else, what does "echo $?" say?
14:50.41 ``Erik (if it's lost, do "archer ; echo $?" and repeat the crash)
14:51.01 YashM_ terminal?
14:51.03 ``Erik the number might possibly be important
14:51.14 ``Erik um, yeah, how are you starting archer? what os, etc?
14:51.18 YashM_ windows
14:51.22 YashM_ i just open the application?
14:52.08 ``Erik ah, um, shoot... I guess post a bug report with enough data to reproduce it
14:52.35 YashM_ can i do it with CMD?
14:53.24 ``Erik I don't know if windows apps can capture a return code in the cmd (terminal) window... :/ sorry, I'm not a windows guy
14:53.33 YashM_ well
14:53.35 YashM_ heres the .g
14:53.35 YashM_ http://ge.tt/8cRW3n62/v/0?c
14:53.41 YashM_ maybe you can try it
14:53.45 ``Erik used a commodore64 until '96 to avoid dos/windows, then went to linux :)
14:53.54 YashM_ hehe
14:55.51 ``Erik hm, my install of BRL-CAD is failing, libpng was updated... I downloaded the .g, I'll recompile and take a peek, might take a while
14:56.07 YashM_ okay, no worries
14:56.37 ``Erik even cmake doesn't want to work *sigh*
14:56.48 ``Erik erik@precious ~/src/brlcad/build/gcc$ make install
14:56.48 ``Erik make: /usr/local/Cellar/cmake/3.0.1/bin/cmake: No such file or directory
14:57.01 YashM_ haha
14:57.07 ``Erik (mac using homebrew... running cmake directly now)
14:57.29 ``Erik I swear, at one time, I did have the ability to run BRL-CAD!
14:58.10 *** join/#brlcad wltan (~wltan@101.127.134.59)
14:58.34 maths22 YashM_: @echo off
14:58.34 maths22 my_nify_exe.exe
14:58.34 maths22 if errorlevel 1 (
14:58.34 maths22 <PROTECTED>
14:58.34 maths22 <PROTECTED>
14:58.36 maths22 )
14:59.06 maths22 save that as a .bat file, replace my_nifty_exe.exe with what you want to run
15:00.18 YashM_ replace with directory or just the exe
15:01.28 ``Erik the .exe file
15:01.45 ``Erik could be C:\path\to\mged.exe
15:05.44 ``Erik ok, libbu is all sorts of broken
15:05.59 YashM_ well archer application is itself a .bat
15:06.09 ``Erik on both mac and linux now...
15:06.46 ``Erik archer is actually a tcl program, there's a batch file to launch it on windows and shell scripts on everything-not-windows...
15:07.15 maths22 ``Erik: what sorts of broken?
15:07.33 YashM_ so does that echo work on .bat
15:07.39 YashM_ also, i guess it only returns a code?
15:07.51 Notify 02GCI:o7p9bxbnyj * 5238032693198848 Implement new API function to reduce duplication #4: Task Claimed - I would like to work on this task.
15:08.01 YashM_ i have cygwin, can i do something with it
15:08.37 ``Erik maths22: "dow_year_start" is unused, comment that out and ntohll redefinition errors
15:09.41 Notify 02GCI:maths22_m * 5238032693198848 Implement new API function to reduce duplication #4: Task Assigned - This task has been assigned to Andromeda Galaxy. You have 100 hours to complete this task,...
15:10.30 adityagulati Does Anyone Know What's Happening To The Ones Who Forged The Tasks
15:12.25 adityagulati ??
15:12.59 maths22 adityagulati: I don't think any of their forged tasks were accepted
15:13.16 maths22 We are still working out anything beyond that...
15:14.21 MarcTannous maths22: if you are ever in doubt about a task being forged, tell me its name and I'll check it out as well
15:14.30 maths22 MarcTannous: thanks
15:14.59 Notify 02GCI:ljudusika * 4797643658100736 Model "BRL-CAD" in BRL-CAD #2: Render - s.png - printscreened from Archer render window. How to save this render into file?
15:15.11 MarcTannous maths22: what is the purpose of bombs.c?
15:15.19 Notify 02GCI:o7p9bxbnyj * 5238032693198848 Implement new API function to reduce duplication #4: Ready for review - The work on this task is ready to be reviewed.
15:15.21 MarcTannous I understand what the functions do, just want to know its purpose in the source code
15:15.23 MarcTannous what is it used in?
15:16.29 maths22 MarcTannous: usefull tool: http://brlcad.org/xref/ident
15:16.36 maths22 For example, search bu_bomb
15:16.37 Notify 02GCI:ljudusika * 4797643658100736 Model "BRL-CAD" in BRL-CAD #2: Ready for review - The work on this task is ready to be reviewed.
15:19.21 MarcTannous http://brlcad.org/xref/ident?i=free
15:19.23 MarcTannous weird
15:19.26 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
15:19.32 MarcTannous free is being used in bomb.c but never referenced
15:19.48 Notify 03BRL-CAD:carlmoore * 63701 (brlcad/trunk/include/bn/chull.h brlcad/trunk/src/libbrep/boolean.cpp brlcad/trunk/src/libbu/y2038/time64.c): remove trailing blanks/tabs; fix spellings
15:19.53 MarcTannous [defined*
15:24.07 maths22 free is a standard library function: http://www.tutorialspoint.com/c_standard_library/c_function_free.htm
15:24.15 *** part/#brlcad maths22 (~maths22@66-118-151-70.static.sagonet.net)
15:24.22 *** join/#brlcad maths22 (~maths22@66-118-151-70.static.sagonet.net)
15:26.14 *** join/#brlcad Ljudusika (4e6fba96@gateway/web/freenode/ip.78.111.186.150)
15:26.50 MarcTannous oh thanks, have not worked with memory alloc before
15:27.08 MarcTannous the identifier tool is really really helpful, thanks a lot
15:30.17 *** join/#brlcad chick_ (~capslock@41.205.22.41)
15:36.04 Notify 02GCI:chiyabansal * 5564670123442176 Fix at least 8 spelling mistakes in at least 8 different files #3: None - Sir, I could not understand the task and request you to please explain it. I would be...
15:36.55 andromeda-galaxy maths22: I really like your buildbot... do you know when any ofthe buildslaves will be back up?
15:44.09 maths22 andromeda-galaxy: not sure
15:45.57 andromeda-galaxy maths22: oh well... it'll be pretty neat when the buildbot's working properly on multiple platforms, especially if we finish fixing all of the tests.
16:06.31 Notify 03BRL-CAD:starseeker * 63702 brlcad/trunk/src/libbn/tests/bn_chull3d.c: turn second test back on
16:06.40 MarcTannous maths22: is there any way to check if a function exitteD?#
16:06.59 maths22 ?
16:07.48 MarcTannous <PROTECTED>
16:07.56 MarcTannous is there any way to check if it did get to that point?
16:10.26 maths22 MarcTannous: I'm not sure. Hopefully someone else can help
16:10.35 MarcTannous ``Erik: any idea?
16:14.47 MarcTannous :(
16:15.16 mihaineacsu_away why don't you print something it runs exit?
16:15.39 mihaineacsu something before it run exit*
16:17.44 Notify 03BRL-CAD:starseeker * 63703 brlcad/trunk/src/libbn/chull3d.c: Try to remove most of the fixed constants for mallocing - not sure all of these are sensible defaults but nothing seems to break. Also put the random shuffling back in, but at least in initial testing it doesn't seem to help much. The Stanford dragon model is too much for this code as it stands, so we'll need to figure out how to fix it at some point. A
16:17.46 Notify gdb interruption after a few minutes seems to indicate we're in chull3d_search a lot, so perhaps there is something about that function or its callees that doesn't scale well.
16:18.13 MarcTannous I am doing a unit test, want to check if it reaches the exit point. I thought about printing something right before that, but is there any way to find out if something was printed?
16:18.22 MarcTannous by just calling the function
16:19.44 d_rossberg MarcTannous: a program returns an int; this int can be set in the program's code via exit(intvalue) or 'return intvalue' in main()
16:20.45 d_rossberg i.e. if a program returns 4 and there is exacly one exit(4) in the program code then you know where the program has stopped
16:21.14 MarcTannous ty
16:23.07 d_rossberg btw, in brl-cad there we have bu_bomb() and bu_exit() which are able to print additional information before the program dies
16:26.24 MarcTannous I am doing a unit test for bomb.c :D
16:29.52 d_rossberg did you know that you can catch bombs with BU_SETJUMP/BU_UNSETJUMP?
16:31.21 andromeda-galaxy d_rossberg: sorry to interrupt, but I'm having a little trouble starting Archer with a specific debug value, any idea why?
16:35.47 d_rossberg andromeda-galaxy: os, parameters? (maybe i've missed your post here)
16:37.25 andromeda-galaxy d_rossberg: archer seems to try to use anything I input as data
16:37.53 andromeda-galaxy i.e. I tried -x like it says in the header comment about RT_G_DEBUG, but that didn't work, neither did setting RT_G_DEBUG as an environment variable
16:43.45 d_rossberg which OS?
16:44.41 andromeda-galaxy liux
16:44.46 andromeda-galaxy *Linux
16:45.13 andromeda-galaxy (Arch, to be specific --- Gentoo takes too long to build things on my laptop)
16:49.13 d_rossberg the shell script which starts archer sets Archer::debug somewhere, maybe you can do similar; archer itself is a tcl program (src/tclscripts/archer) where again a debug variable is referenced
16:49.48 andromeda-galaxy I know, but what I actually need to debug
16:49.59 andromeda-galaxy is one of the C commands from libged that Archer calls
16:50.08 andromeda-galaxy that uses RT_G_DEBUG&DEBUG_TREEWALK
16:50.17 andromeda-galaxy so I'm trying to figure out how to set RT_G_DEBUG
16:50.47 andromeda-galaxy d_rossberg: in raytrace.h, it says to use the "-x" option, but Archer doesn't take that option and pass it through...
16:53.28 Notify 02GCI:jose99peter * 4639899508539392 Write manual page documentation (for g-jack): Task Claimed - I would like to work on this task.
16:53.58 Notify 02GCI:deepakkumarsharma * 4639899508539392 Write manual page documentation (for g-jack): Task Assigned - This task has been assigned to JOEPETER FRANCIS. You have 100 hours to complete this task,...
16:58.43 *** join/#brlcad andrei_ (~andrei@5-12-49-227.residential.rdsnet.ro)
16:59.35 d_rossberg the only idea i have is to run archer in gdb, set a breakpoint somwhere in librt before your function will be called an set RTG.debug in the debugger
17:01.10 Notify 02GCI:jose99peter * 4639899508539392 Write manual page documentation (for g-jack): Claim Removed - The claim on this task has been removed, someone else can claim it now.
17:05.16 Notify 02GCI:vaishnavisvp * 5564670123442176 Fix at least 8 spelling mistakes in at least 8 different files #3: Task Claimed - I would like to work on this task.
17:06.01 Notify 02GCI:Melange * 6353662997692416 Design a BRL-CAD splash screen #3: No more Work can be submitted - Melange has detected that the deadline has passed and no more work can be submitted. The submitted...
17:20.27 Notify 02GCI:rohit_agarwal * 5221860329390080 Design a new website favicon! #8: Ready for review - The work on this task is ready to be reviewed.
17:22.27 Notify 02GCI:rohit_agarwal * 5221860329390080 Design a new website favicon! #8: 2 Files Uploaded - I have uploaded a new favicon as you asked. My new files are - Favicon 3.xcf/.gif
17:25.03 Notify 02GCI:marioscappini * 6429879809605632 Design a new website landing page for BRL-CAD #5: Ready for review - The work on this task is ready to be reviewed.
17:25.49 *** join/#brlcad Vaishnavisvp (b641e9a2@gateway/web/freenode/ip.182.65.233.162)
17:26.08 YashM_ Hi Vaishnavisvp, you can leave your question here and someone answer
17:26.15 YashM_ someone will*
17:26.23 Notify 02GCI:keshashah * 5564670123442176 Fix at least 8 spelling mistakes in at least 8 different files #3: Task Assigned - This task has been assigned to Vaishnavi Prasad. You have 100 hours to complete...
17:26.38 Vaishnavisvp Thank you so much :D
17:26.59 Vaishnavisvp May I know where I can get the files?
17:28.25 YashM_ Vaishnavisvp, which task?
17:29.08 Vaishnavisvp I must fix at least 8 spelling mistakes in 8 different files.
17:30.21 YashM_ do you have SVN?
17:30.33 Notify 02GCI:adityagulati * 5887275384700928 Model "BRL-CAD" in BRL-CAD: Ready for review - The work on this task is ready to be reviewed.
17:31.06 Vaishnavisvp YashM_ : SVN? I'm really new to this, apologies.
17:31.45 *** join/#brlcad deepak (~chatzilla@117.199.104.249)
17:32.04 YashM_ Are you on Windows?
17:34.48 Vaishnavisvp YashM_ yes
17:35.06 YashM_ you need the BRL-CAD source files
17:35.17 YashM_ You can download TortoiseSVN
17:35.24 YashM_ and get the files on your PC
17:35.25 Vaishnavisvp YashM_ yes.
17:35.46 YashM_ then you need to do an SVN checkout
17:35.52 YashM_ a*
17:36.47 Vaishnavisvp YashM_ thank you very much! I shall try it now. :D
17:38.08 YashM_ Do check the BRL-CAD wiki: http://brlcad.org/wiki
17:38.17 YashM_ ~ask
17:38.17 infobot Questions in the channel should be specific, informative, complete, concise, and on-topic. Don't ask if you can ask a question first. Don't ask if a person is there; just ask what you intended to ask them. Better questions more frequently yield better answers. We are all here voluntarily or against our will.
17:41.08 ignacio for gcibot users, the time bug, has been fixed
17:46.56 Notify 02GCI:denisapirlea * 4639899508539392 Write manual page documentation (for g-jack): Task Claimed - I would like to work on this task.
17:47.11 Notify 02GCI:deepakkumarsharma * 4639899508539392 Write manual page documentation (for g-jack): Task Assigned - This task has been assigned to Denisa. You have 100 hours to complete this task, good luck!
17:48.31 Notify 02GCI:sharan1998 * 5821990405406720 Model a toy tank in 3D using BRL-CAD #4: Task Claimed - I would like to work on this task.
17:49.21 Notify 02GCI:mihaineacsu29 * 5821990405406720 Model a toy tank in 3D using BRL-CAD #4: Task Assigned - This task has been assigned to Sharan Narayan. You have 100 hours to complete this task, good luck!
17:50.03 MarcTannous for anyone coding under the Debian VM
17:50.12 MarcTannous do NOT drag files into gedit, as it WILL crash
17:50.19 MarcTannous I mean crash the whole thing
17:50.21 mihaineacsu :))
17:50.53 *** join/#brlcad Michael_Huang (~mike@ip-other.acalanes.k12.ca.us)
17:51.00 Notify 02GCI:sharan1998 * 5821990405406720 Model a toy tank in 3D using BRL-CAD #4: I would like to complete - Sorry I got no reply so didn't continue Also there were exams going on but now it's over ;) ...
17:57.01 Notify 02GCI:deepakkumarsharma * 6429879809605632 Design a new website landing page for BRL-CAD #5: Good but ... - Hello Marioscappini, You give a good try to it but there are certain things that needs...
17:57.07 Notify 02GCI:deepakkumarsharma * 6429879809605632 Design a new website landing page for BRL-CAD #5: Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s)...
18:03.06 *** join/#brlcad teepee (~teepee@gateway/tor-sasl/teepee)
18:03.27 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
18:03.49 Notify 03BRL-CAD:carlmoore * 63704 brlcad/trunk/src/util/pixblend.c: add h? for help; remove two 'else' usages because there is 'return' just ahead of them
18:04.26 Notify 02GCI:marioscappini * 6429879809605632 Design a new website landing page for BRL-CAD #5: None - Ok! i will improve it and send it again
18:04.47 andromeda-galaxy Any chance of a task to fix the bug that YashM_ reported?
18:11.45 MarcTannous what is the syntax for running a specific test?
18:17.05 *** join/#brlcad Michael_Huang (~mike@ip-other.acalanes.k12.ca.us)
18:21.14 *** part/#brlcad Michael_Huang (~mike@ip-other.acalanes.k12.ca.us)
18:26.18 andromeda-galaxy YashM_: I don't know for sure if htis is the case, but your crash may have been related to the fact that the skt_0 that asas is extruded from is empty
18:34.46 ``Erik MarcTannous: which test set? the "old" style tests would just be running the test executable
18:35.00 MarcTannous I am working on a unit test, want to test it separately
18:35.06 Notify 02GCI:marioscappini * 6429879809605632 Design a new website landing page for BRL-CAD #5: None - what do you mean by the menu items, do you mean shift it toward the logo of the menu?
18:35.11 MarcTannous just "make test executablename"?
18:35.29 ``Erik no, "src/libbu/test_bu" or whatever
18:35.36 ``Erik no make to it, just run the executable
18:36.12 ``Erik the cmake tests are different, though...
18:36.38 MarcTannous what do you mean by just run the executable
18:36.52 andrei_ ``Erik, can you explain me quickly how bu_parallel works
18:37.07 andrei_ in relation to fork() or winapi?
18:37.09 MarcTannous oh
18:37.11 MarcTannous found it
18:37.39 andrei_ or, how does BRL-CAD implement portable fork? Or it doesn't?
18:37.44 ``Erik bu_parallel shouldn't mess with fork, it should be all threads, posix model being almost natural
18:38.22 andrei_ as far as I remember, simulating fork on windows is difficult
18:38.23 MarcTannous ``Erik: if I try to run the executable nothing happens
18:38.45 ``Erik um, I d'no about a portable fork, but there should be something roughly equivalent to system() that is portable, so "rt" and such can run from mged
18:39.43 ``Erik MarcTannous: "nothing" is a strange word... did you check the return code? the unix approach (which most of BRL-CAD adheres to) is to print nothing on success
18:40.09 andrei_ ``Erik I need to figure out when the process I spawned ended, that's what I want it for
18:40.26 MarcTannous what do you mean by "run the applcation"? navigating to brlcad-build/src/libbu/tests/tester_bu_example and double click it?
18:40.49 ``Erik andrei_: unix style is wait() (I like wait4() myself)
18:41.54 ``Erik MarcTannous: that'll throw away the return code, can you run it from a cmd prompt doing the weird dos weirdness Herr Rossberg mentioned earlier?
18:41.57 mihaineacsu_away MarcTannous: if you get the test running, you can check the test's result from bash by running "echo $?"
18:42.58 andrei_ what Mihai said ^ is a very useful linux command
18:43.17 andrei_ I use it a lot
18:43.37 ``Erik $? is dandy when you're using anything but windows, I suspect MarcTannous is using windows...
18:43.47 MarcTannous I'm on a debian VM
18:44.02 ``Erik ah, "myprogram ; echo $?" boom
18:44.06 andrei_ ``Erik: :))
18:44.36 ``Erik little bit of bash fu, && and || can be used to do new actions based on the return value of a previous code...
18:45.01 ``Erik myprog && echo "Success!" || echo "Failure!"
18:45.40 andrei_ { foo; echo $?; } | awk '!/[^0-9]/ {exit($0)} {Â…}'
18:45.42 andrei_ ^^
18:46.47 andrei_ theoretically, this should feed the exit code to another prog
18:47.23 ``Erik myprog ; echo $? | newprog
18:47.34 ``Erik or myprog ; newprog < $?
18:47.36 ``Erik or ...
18:47.36 Notify 02GCI:denisapirlea * 4639899508539392 Write manual page documentation (for g-jack): Claim Removed - The claim on this task has been removed, someone else can claim it now.
18:47.58 andrei_ Marc is writing a unit test for a file that happens to contain bu_exit
18:48.20 andrei_ and you can't really test the return value of bu_exit since it .. exits the program
18:48.34 andrei_ my approach was to fork and to see the child's process ret val, but that's not portable
18:48.41 andrei_ how does BRL-CAD handle this sort of issues?
18:48.54 ``Erik is the test of bu_exit() or something that calls bu_exit()?
18:49.04 andrei_ of bu_exit()
18:49.18 andrei_ Marc, can you paste your work so far
18:49.19 andrei_ or so?
18:50.06 ``Erik I'd make a seperate executable for each bu_exit() call case and wrap them in the ctest macros in the build system... but that is a tricky thing to test
18:51.03 andrei_ I don t see too many options
18:51.08 MarcTannous <PROTECTED>
18:51.22 andrei_ you get a compiler error because it doesn't get to call return
18:51.37 andrei_ the program "wants" to end with bu_exit()
18:51.40 andrei_ hence, no return exists
18:52.51 ``Erik bu_exit() should call exit() which zomfg shoots the proc in the head, it's all over, there is no stack unwind, no return value, it's just over
18:52.55 MarcTannous how should I go about creating a separate process just for the purpose of it getting exitted?
18:53.08 MarcTannous ``Erik : ))
18:53.52 ``Erik MarcTannous: you can either have a fork which could be nonportable, or multiple calls to the program with expected exits...
18:54.29 MarcTannous not sure which one of these is the better/correct approach though
18:55.44 ``Erik fork() is used friendly in the code, without wrapping... so I'd assume it's a necessary func and your test can use it
18:55.52 ``Erik s/friendly/freely/
18:56.33 MarcTannous what about portability?
18:56.37 ``Erik src/libged/nirt.c src/libbu/backtrace.c src/rt/worker.c src/mged/plot.c
18:56.56 ``Erik given the use I'm seeing, I'd assume it's portable enough?
18:56.57 andrei_ MarcTannous: you can test bu_exit() only on linux
18:57.05 mihaineacsu_away yup
18:57.13 andrei_ life #ifdef linux_stuff
18:57.23 andrei_ defien the function and use fork
18:57.26 andrei_ otherwise, simply don't test
18:58.35 MarcTannous #ifdef __linux__
18:58.38 MarcTannous right?
18:58.41 ``Erik libbu might need a portable fork to deal with windows CreateProcess() goop
18:58.53 ``Erik MarcTannous: I'd say !__WIN32__
18:59.19 MarcTannous correct
18:59.21 maths22 ``Erik: The cmake logic should just only run the test if fork() exists
18:59.34 ``Erik mac, fbsd, solaris, aix, hpux, tru64, photon, beos, obsd, etc all deal with fork() right... windows is the slow kid
18:59.49 maths22 Otherwise it shouldn't be run as a pass or a fail
19:00.02 ``Erik yeh, cmake can test for fork()
19:00.35 mihaineacsu ``Erik: so I guess brlcad doesn't ever need to fork a process? (since it doesn't have a portable fork)
19:01.16 maths22 mihaineacsu: http://brlcad.org/xref/ident?i=fork
19:01.17 ``Erik mihaineacsu: it does... frequently, windows is just a special case :/
19:01.32 maths22 windows has fork() redefined to -1
19:01.33 andrei_ haha
19:01.37 maths22 http://brlcad.org/xref/source/include/config_win.h#L311
19:01.46 mihaineacsu lol
19:02.06 ``Erik yeah, src/libged/rt.c has the fu to fork() the rt proc, it's wrapped in a #ifndef _WIN32
19:02.09 mihaineacsu * faking it"
19:02.14 andrei_ hahaha
19:02.23 ``Erik src/libged/rt.c:177
19:02.50 ``Erik windows vs the rest of the world :)
19:02.59 mihaineacsu as always
19:03.36 ``Erik bu_fork() might be a good gci task, though
19:03.52 andrei_ just like the quarternions, perhaps
19:03.52 andrei_ :p
19:03.59 andrei_ quaternions*
19:04.25 ``Erik quaternions are awesome, I chased down a math professor to explain them to me back in '00 or os
19:04.28 ``Erik so
19:05.30 MarcTannous there's a task regarding unit tests for the quaternion library
19:05.35 MarcTannous nobody's touched it thisf ar
19:06.16 ``Erik in the US, quaternions are typically taught in college level linear algebra, so it doesn't surprise me that high school students are reluctant
19:06.43 andrei_ I'm surprised andromeda didn't touch them
19:06.49 andrei_ he seems passionate about math
19:07.20 MarcTannous in Romania, students in their final year of highschool, with a gold medal in international maths olympiad do not know quaternions
19:07.32 MarcTannous trust me, I've asked
19:09.57 andrei_ MarcTannous, it's not an untouchable topic, I'm sure it would take you less than two days to get a hold of them
19:10.13 ``Erik quaternions are a bit of an odd subject; you can learned the minimal applied capability and it's a bunch of black magic, you can learn the pure mathematical basis and it's a bunch of unapplicable black magic... if you can grok the mathematical side and understand how it's applied, suddenly it's the best stuff ever
19:11.00 MarcTannous how does a child process return its exit status?
19:11.16 MarcTannous googled around, found out the syntax for a child process is if ((pid = fork()) == 0) { commands here
19:11.16 ``Erik the lynchpin is that the entire 3d universe is a unit sphere in quaternion space
19:11.31 MarcTannous I only want to have it exit and then return the status it exitted with
19:11.35 ``Erik MarcTannous: fork() will return the child proc, then you have to call wait() to get the return results
19:12.05 ``Erik man 2 wait
19:15.04 MarcTannous so just pid_t wait(int *status);
19:15.06 MarcTannous ?
19:15.40 *** join/#brlcad MarcTannous_ (5679c4d5@gateway/web/freenode/ip.86.121.196.213)
19:15.59 MarcTannous_ oh damn, I just accidentally closed about 20 pages of documentation
19:16.10 ``Erik MarcTannous: pretty much... I usually use wait4()
19:16.10 MarcTannous_ and my IRC
19:16.21 MarcTannous_ what's the difference betwen wait and wait4?
19:16.27 ``Erik um, if it's chrome or firefox, history should have a "restore previous session"?
19:16.35 ``Erik wait4 gives you more info
19:19.36 andrei_ Marc, let me sketch an example
19:19.55 MarcTannous_ ty
19:20.38 MarcTannous_ having never worked with child processes before, it's strange that I can get the status it terminated with at a later time
19:20.51 andrei_ MarcTannous_ : http://www.amparo.net/ce155/fork-ex.html
19:20.56 andrei_ here s exactly what you want
19:21.16 Notify 02GCI:marioscappini * 6429879809605632 Design a new website landing page for BRL-CAD #5: Ready for review - The work on this task is ready to be reviewed.
19:21.41 ``Erik :) on linux, a terminated process that hasn't had the status collected yet is called a zombie process, I'm going to guess that you've seen zombies before? know ya know why :)
19:23.16 MarcTannous_ thank you guys, I now have a basic understanding of fork(), hope my unit test works
19:23.36 andrei_ it's called zombie because you can t interract with it
19:24.03 andrei_ like
19:24.05 andrei_ it's "dead"
19:24.09 andrei_ but it s still using resources
19:25.49 ``Erik yeah, terminated, but it hasn't had the return code read so it can't be completely released
19:27.21 andrei_ linux is quite full of nerdjokes
19:27.23 andrei_ in various ways
19:28.09 ``Erik bogomips.
19:28.14 andromeda-galaxy ``Erik, andrei_: just going over the backlog, I actually have been planning to look into some of the quaternion functions :-) the only reason that I haven't done them yet is that I usually use Mathematica for generataing the correct results, but I can't find a Mathematica function that is equivalent to quat_sblerp()
19:28.33 andromeda-galaxy also, I was reading that fork() is *really* hard to do right in Windows
19:29.07 ``Erik sblerp? I'm familiar with slerp, what's the b?
19:29.23 andrei_ mihaineacsu and I have done it as a homework assignment as part of a course
19:29.26 andromeda-galaxy sorry, sberp
19:29.32 andrei_ but yeah, it s not easy
19:29.33 andromeda-galaxy spherical bezier interpolation
19:30.00 andrei_ and quite pointless
19:30.08 ``Erik ah, gotcha, my background is game crap from 10-15 years ago, slerp (spherical linear interpolation) was the hot shit
19:30.30 andromeda-galaxy yeah... I can't find much documentation on sberp
19:30.39 andromeda-galaxy that's and slerp are probably the hardest ones to test
19:31.26 mihaineacsu <andromeda-galaxy>also, I was reading that fork() is *really* hard to do right in Windows - yes, it's sort of nerve racking
19:33.01 ``Erik we have a live example of a windows fork() in src/libged/rt.c:177... it's not that hard, just weird
19:33.12 andrei_ what annoys me the most on windows is the genericity of their api
19:33.19 andrei_ you got a function that does 10 things
19:33.27 andrei_ with a lot of parameters
19:33.34 andrei_ and most of the time a lot of them are null
19:33.36 andrei_ or 0
19:33.52 deepak Hey guys, I'm reading your chat regarding Quaternions. It seems quite interesting I have tried to explore to but couldn't make myself understand. Can you please briefly explain me what it is all about.
19:33.58 ``Erik loads an HWND into a trebuchet and flings it at andrei :D
19:34.20 ``Erik deepak: are you familiar with complex numbers?
19:34.23 andrei_ haha, a handle? :))
19:34.24 andromeda-galaxy oh well... ``Erik: I guess that what I was reading was about how Cygwin tries to emulate process apis
19:34.35 deepak Yes!
19:35.15 andromeda-galaxy quaternions are like complex numbers, excep that instead of just a+bi, they're a+bi+cj+dk (if I'm remembering corectly)
19:35.34 andromeda-galaxy they're a fun ngroup
19:36.10 andromeda-galaxy ``Erik, andrei_: do you know of any papers about BRL-CAD's sberp algorithm?
19:36.19 andrei_ I don't, for sure
19:37.08 ``Erik deepak: we can generalize 3d space as x,y,z. We can talk about imaginary numbers as ax+bi where i is the imaginary component... we could extend complex numbers into a complex 3d space by saying that the imaginary component is 3d, a+bi+cj+dk, where ijk is the imaginary 3d space
19:37.33 ``Erik andromeda-galaxy: cygwin tries to be unix, where we might just want a subset
19:37.56 andromeda-galaxy ``Erik: right.. duplicating the *exact* behavior of fork is really hard, but making something close enough isn't too bad?
19:38.09 ``Erik andromeda-galaxy: yes
19:39.00 andromeda-galaxy ``Erik: makes sense... I can't find *any* information about the expected behavior of spherical bezier interpolation! (apart from this one set of slides that is verey general and has the text positioning messed up: http://caig.cs.nctu.edu.tw/course/CA/Lecture/slerp.pdf)
19:39.10 ``Erik fork() and exec*() have a lot of implicit details that are very very unix specific, but the 99% case is fairly routine
19:39.13 deepak Can you please provide me with any real world example.
19:41.19 ``Erik a single point in quaternion space has enough information that we can translate it to a pair of vectors in 3d cartesian space; a directional vector and an 'up' vector... this is adequate to describe any orientation/position we could care about
19:42.54 ``Erik the sweet part is that a quaternion space point is discreet, there is only one. If you were dealing with 3d directions in another system like euler coordinates, as you get to the poles, there are an infinite number of representations.... so you get ugly behaviors like gimbal lock, or you don't know which way is a shortcut to the next position, so the point whips around in a crazy way
19:43.32 ``Erik with a quaternion, you move, you project to the unit sphere, translate to the cartesian space, and you're done... no gimbal lock, no whipping... it just works
19:44.31 ``Erik um, the plib guy, uh, steve baker I think? did a great article on this a long time ago
19:44.57 ``Erik http://www.sjbaker.org/steve/omniv/eulers_are_evil.html
19:46.31 andromeda-galaxy ``Erik: so do you know of any reasonably easy way to get expected values for slerp/sberp?O
19:47.19 andromeda-galaxy The rest of them are alright, but slerp/sberp don't seem to have many ``reference implementations'' so to speak
19:47.56 ``Erik andromeda-galaxy: nope! mathematica should give you decent values to expect... you could do 'em longhand on paper, you might be able to google a test suite... but I don't think there's any good easy cheat
19:48.39 andromeda-galaxy ``Erik: yeah, the problem is that Mathematica doesn't actually have very good qmath facilities
19:48.41 deepak Seems interesting. I want to explore it more and more. Can you just tell me where to start so that the concept get exactly clear for me.
19:49.04 andromeda-galaxy ``Erik: http://reference.wolfram.com/language/Quaternions/guide/QuaternionsPackage.html, no interpolation!
19:49.47 ``Erik andromeda-galaxy: maybe you can be the first to make a de-facto test suite!
19:50.00 andromeda-galaxy ``Erik: would be fun...
19:50.39 ``Erik <lame game company> so, uh, why should we hire you?" <andromeda> bitch, I wrote the book on quaternion slerp testing!
19:52.18 andromeda-galaxy ``Erik: I'll probably start with the other functions, then look into the slerp stuff, maybe do some calculations by hand
19:52.42 andromeda-galaxy the real problem will be with the sberp stuff, at least there's documentation about the slerp algorithm out there...
19:54.03 ``Erik I like to think I'm pretty goddamn smart, and I don't know squat about quaternion sberp... anything you do with that is going to be world class
19:54.31 andromeda-galaxy ``Eirk: thanks! :-)
19:54.48 andromeda-galaxy oh look, every google result about quaternion sberp is from libbn source code!
19:55.44 ``Erik so you'd be the world leader in SBERP with experince in the world leading platform for SBERP...
19:57.00 andromeda-galaxy now I'm really curious who wrote that function...
19:57.12 ``Erik svn log
19:57.34 ``Erik wonders if it was twinky
19:58.19 andromeda-galaxy nothing about interpolation in any of the log messages at least...
20:03.40 ``Erik one of the previous developers of BRL-CAD wrote an entire modeling system based on quaternions called "nurbana", brilliant dude... far more "crash and burn" than me, but I ended up adopting a few of his projects
20:04.21 andromeda-galaxy interesting...
20:06.26 ``Erik well, more nurbs than quaternions, but highly related
20:06.38 Stragus Was that Justin?
20:07.03 ``Erik yeah, don't tell him I said that...
20:07.24 Stragus Eheh, it's noted
20:07.33 Stragus I haven't heard of him in years
20:07.57 ``Erik he got a pilot license and a registered glider
20:08.20 MarcTannous_ notices the chat took quite a turn in the past half an hour
20:08.22 ``Erik also; married, bought a large house out in farmland, ...
20:08.51 Stragus So no more rockets to launch a home-made satellite from his garage
20:09.20 maths22 ``Erik: written by mike
20:09.25 maths22 http://brlcad.org/fisheye/changelog/BRL-CAD?cs=2074
20:09.34 andromeda-galaxy hmm...it looks kinda like the interpolation equivalent of a cubic bezier curv...
20:09.34 ``Erik hehehe, I've no idea! he doesn't talk to me anymore, he's busy with his own stuff
20:09.35 maths22 Or at least committed by him
20:10.11 maths22 ``Erik: committed 25 years ago, and essentially unchanged
20:10.25 maths22 It got moved, but the body of the function is identical
20:11.25 Stragus andromeda-galaxy: Quaternions are neat, it's the sane way to interpolate between rotations in 3D (or interpolate between "matrices" if you wish)
20:11.49 maths22 andromeda-galaxy: http://caig.cs.nctu.edu.tw/course/CA/Lecture/slerp.pdf seems to address sberp
20:11.55 maths22 just not by that name
20:12.02 Notify 02GCI:sharan1998 * 5821990405406720 Model a toy tank in 3D using BRL-CAD #4: Ready for review - The work on this task is ready to be reviewed.
20:12.07 maths22 Page 16?
20:12.11 ``Erik maths22: not terribly surprising that mike muuss had an interest... iirc, the quaternion story goes back to a constitutinal by hamilton in the 1820's, and he carved the gist in a bridge in scotland?
20:12.13 andromeda-galaxy maths22: that's the one that I saw to... it doesn't go into to much depth, as far as i can tell...
20:12.36 Stragus I heard that story about the bridge engraving
20:12.40 maths22 https://ece.uwaterloo.ca/~dwharder/C++/CQOST/Bezier/
20:12.54 maths22 I'm not sure if that really explains it, but it might
20:13.20 andromeda-galaxy looking at the source & considering the definition fo a Bezier curve (and Paragraph 6 of the ece.uwaterloo.ca link, I think I've got it...
20:13.46 maths22 andromeda-galaxy: http://run.usc.edu/cs520-s13/quaternions/quaternions-cs520-6up.pdf
20:13.53 ``Erik woops, 1840's, my bad
20:14.52 ``Erik and an irish bridge http://en.wikipedia.org/wiki/Broom_Bridge
20:17.32 andromeda-galaxy brl-cad does quaternions as fastf_t[4]={x,y,z,w}, right?
20:17.51 ``Erik yes, where x,y,z are i,j,k
20:18.10 andromeda-galaxy right, it's x+yi+zj+wk
20:18.23 andromeda-galaxy the quaternion group is kinda fun to...
20:18.44 ``Erik uh, I think ix, jy, kz, w
20:18.57 andromeda-galaxy oh is it... yeah, my bad
20:18.58 andromeda-galaxy sorry
20:19.09 andromeda-galaxy I'm used to seing them written as
20:19.18 andromeda-galaxy <1,i,j,k> for some reason
20:19.42 ``Erik :) as long as the term correlate in the end
20:20.09 andromeda-galaxy indeed...
20:20.24 Stragus Mathematicians and programmers often like to write things out differently
20:20.34 Stragus i,j,k,1 makes a lot more sense for a programmer
20:21.01 andromeda-galaxy Stragus: why?
20:21.41 Stragus Because it maps directly to points/vertices/vectors of x,y,z,w. Therefore you can easily do arithmetics with a loop or vectorized with SSE and so on
20:21.48 ``Erik Stragus thinks computers were invented in 1970, so he cound funny
20:21.56 ``Erik counts
20:22.44 Stragus Eh. Try to optimize quaternion maths written stored as 1,i,j,k with SSE, you'll have extra pointless shuffles
20:23.09 ``Erik yes, sse is built on ideas from 1970
20:23.39 ``Erik right along with indexing counting from 0
20:23.47 Stragus And it's still around and kicking today, hence why programmers like to write things differently from mathematicians
20:26.01 ``Erik hey, I'm not saying there's anything wrong with it :) just recognize the history... fortran counts from 1 because computers counted from 1... until 1970... with the pdp1... to crap-talk fortran for not complying to specs that showed up 15 years after its inception, well, that'd be lame... :D
20:26.06 ``Erik O:-)
20:26.28 ``Erik (not saying fortran was right, just saying it wasn't wrong at the time)
20:26.42 Notify 03BRL-CAD:starseeker * 63705 brlcad/trunk/src/libbn/chull3d.c: Break up the task of convex hull calculation into calculating the hulls of subsets of the input vertex array, and then calculating hulls on the sets of hulls to build up to the final hull. While the previous setup crashed on the Stanford dragon after around 9 hours running, this approach succeeded in about 30 seconds. It doesn't address whatever the
20:26.44 Notify underlying issues are with the hull implementation itself and large data sets, just pre-processes the inputs to avoid dealing with them *as* a large data set. This will not work in cases where large numbers of vertices are legitimately part of the convex hull - in that case it actually increases run times - and it's still not fast on large models even when it helps, but since large bots of that nature already aren't
20:26.46 Notify practical with this code and this does handle at least some situations better go with it for now.
20:26.52 ``Erik shakes fist a bit, then steps off the soap box
20:27.23 Stragus :) I need to look up on these old computers you speak of, it still makes no sense to count from 1
20:27.45 ``Erik early computers also counted base 10
20:28.06 Stragus I want to access the element 1 from an array. Is it at memory[index*stride] or is it at memory[(index-1)*stride]?
20:28.35 ``Erik element one is memory[1]
20:28.46 ``Erik boom, done, let the machine math it out, that's the job
20:28.47 Stragus And how do you compute the offset to it?
20:29.10 Stragus It requires an extra operation to subtract 1, totally superfluous due to a flawed design
20:29.50 Stragus In fact, it's the human languages that are broken. The zeroth element should be the first element
20:30.00 Stragus steps down ``Erik's soap box
20:30.18 ``Erik Stragus: your argument is erroneous... we do things different now because we learned better... things were different back then
20:31.09 Stragus Fine, so we learned a better way, and that's why computers today count from zero
20:31.10 ``Erik I will flippin' beat you with your own bicycle, er, um, uh, I mean, uh, early computers did not behave as a C dev would expect them...
20:31.24 ``Erik the pdp1 was a huge progression, and it is what C is
20:31.32 Stragus nods to that
20:32.29 ``Erik still wants to beat Stragus with a bicycle thought :D
20:32.34 ``Erik though
20:32.58 Stragus I'll watch out the next time I bring mine to Maryland :)
20:33.36 ``Erik any idea when that will be? I owe you a good dinner :/ that last one was embarassing
20:34.00 Stragus Ahah. No idea yet, SURVICE seems fond of doing Google hangouts lately
20:35.11 Stragus Did the kid start building without instructions already?
20:35.50 Stragus If not, I'll have to bring a tub with no instructions
20:35.58 ``Erik he's still hesitant, I think I'm going to buy him a big box of generic parts with no instructions
20:36.01 ``Erik hah
20:36.15 Stragus Yup, that should be the plan
20:36.36 ``Erik don't you worry about it, I'll take care of it :)
20:37.13 ``Erik https://www.google.com/search?safe=off&es_sm=91&tbm=isch&q=lego+bucket&revid=1367562469&sa=X&ei=61GLVO_YEeG1sATFqoCIBA&ved=0CCMQ1QIoAQ&dpr=1&biw=1232&bih=705#facrc=_&imgdii=_&imgrc=sAb0Il7ATJ5XeM%253A%3BaOy-HTQ4wuwhJM%3Bhttp%253A%252F%252Fecx.images-amazon.com%252Fimages%252FI%252F916yU0qxE8L._SY355_.jpg%3Bhttp%253A%252F%252Fwww.amazon.com%252FLEGO-Bricks-10662-Creative-Bucket%252Fdp%252FB00AWBLS80%3B323%3B355
20:37.43 ``Erik http://www.amazon.com/LEGO-Bricks-10662-Creative-Bucket/dp/B00AWBLS80 to be more exact
20:38.23 Stragus That's overpriced though
20:38.39 Stragus But yes, something like that
20:39.57 andromeda-galaxy I can't find a mat_t equivalent of HNEAR_EQUAL... oh well
20:56.39 ``Erik overpriced? can you give me a better spot?
20:57.11 ``Erik I'm itching to give him something without instructions :-)
20:57.37 ``Erik I think I saw a similar thing for ~30
20:59.12 Stragus Walmart U.S. had a 1600 parts tub for $30 for Black Friday
20:59.24 Stragus Let me look around a little
20:59.39 Notify 02GCI:brlcad * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s)...
20:59.45 Notify 02GCI:brlcad * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: Deadline extended - The deadline of the task has been extended with 2 days and 2 hours.
21:00.31 andromeda-galaxy ``Erik: sorry to interrupt, but what exactly does quat_make_nearest do?
21:03.01 Notify 02GCI:Melange * 4797643658100736 Model "BRL-CAD" in BRL-CAD #2: No more Work can be submitted - Melange has detected that the deadline has passed and no more work can be submitted. The submitted...
21:04.06 Notify 02GCI:brlcad * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: prose - Olivia, sorry for not replying sooner. This is looking great. The only bit is reducing the prose,...
21:05.10 MarcTannous_ ``Erik: for wait()
21:05.19 MarcTannous_ I need to #include <sys/wait.h> right?
21:05.41 ``Erik yes
21:05.51 MarcTannous_ also #include <sys/types.h> ?
21:05.58 ``Erik probably
21:06.15 MarcTannous_ can I get some feedback on a simple function test?
21:06.21 ``Erik sure
21:06.26 andromeda-galaxy ``Erik: do you know where there is some good documentation on quat_make_nearest?
21:06.40 MarcTannous_ http://pastebin.com/kpA8w4v4
21:06.58 ``Erik andromeda-galaxy: how do you mean? zomfg mathematical, or just code?
21:07.11 andromeda-galaxy almost any... ideally not the code that's in qmath.c...
21:07.18 andromeda-galaxy for the test...
21:08.06 Notify 02GCI:brlcad * 6423086379302912 Write manual page documentation (for g-dot): Task Closed - Congratulations, this task has been completed successfully.
21:08.51 ``Erik I can try to convince the fiance to give me some time tomorrow to read the code and grok the math...
21:08.54 MarcTannous_ ``Erik: is that one okay, should I try and compile it?
21:08.57 Notify 03BRL-CAD:brlcad * 63706 brlcad/trunk/doc/docbook/system/man1/en/CMakeLists.txt: apply gci patch http://www.google-melange.com/gci/task/view/google/gci2014/6423086379302912 from shardulc that adds a new manual page for the g-dot exporter.
21:10.08 *** join/#brlcad chick_ (~capslock@41.205.22.41)
21:10.09 ``Erik MarcTannous_: try to compile it.
21:10.27 chick_ clear
21:10.27 ``Erik and run it, and check it against tests
21:10.41 MarcTannous_ if it works
21:10.43 MarcTannous_ you're the man
21:11.23 Notify 02GCI:brlcad * 6423086379302912 Write manual page documentation (for g-dot): nicely done! - Shardul, this looks much better. I was able to successfully apply your changes and they compiled cleanly....
21:11.37 MarcTannous_ the whole proces in a process thingie is great
21:12.05 ``Erik no, dude, don't do that to me... the gf's bdat is just wrapping up, my bday is starting, every bit of time I can give is outside of norm...
21:12.50 MarcTannous_ happy birthday!
21:13.20 ``Erik not until sunday, but thanks!
21:13.40 MarcTannous_ it compiled :O
21:13.52 MarcTannous_ if the test works, I'm gonna go to sleep so happy
21:13.54 Notify 03BRL-CAD:brlcad * 63707 (brlcad/trunk/AUTHORS brlcad/trunk/NEWS): credit Shardul for his latest gci contribution, manpage for g-dot
21:14.00 Notify 02GCI:brlcad * 5238032693198848 Implement new API function to reduce duplication #4: Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned...
21:14.11 ``Erik crosses his fingers for mark
21:14.54 MarcTannous_ what was the syntax for echo again?
21:15.01 MarcTannous_ echo $? seems to not work
21:16.09 MarcTannous_ IT WORKS AND THE TEST WORKs
21:17.49 andromeda-galaxy ``Erik: not sure if you saw this above, but I'm fine with any description of the make_nearest function you can find other than the bn implementation; mathematical would be great...
21:18.16 Notify 02GCI:brlcad * 5238032693198848 Implement new API function to reduce duplication #4: er, reduction? - Perhaps I missed the task/discussion where the need for this API addition came up? I'm not sure...
21:18.46 *** join/#brlcad Ch3ck_ (~localhost@41.205.22.13)
21:21.17 Notify 02GCI:izak * 4797643658100736 Model "BRL-CAD" in BRL-CAD #2: Deadline extended - The deadline of the task has been extended with 1 days and 0 hours.
21:21.32 Notify 02GCI:o7p9bxbnyj * 5238032693198848 Implement new API function to reduce duplication #4: Discussion on IRC - There was a big discussion on IRC with me, Stragus, and starseeker, last night around...
21:24.01 Notify 02GCI:o7p9bxbnyj * 5238032693198848 Implement new API function to reduce duplication #4: Reason for submission under this task - I was originally planning on submitting this under fixing the last...
21:24.06 Notify 02GCI:o7p9bxbnyj * 5238032693198848 Implement new API function to reduce duplication #4: Ready for review - The work on this task is ready to be reviewed.
21:25.48 andromeda-galaxy MarcTannous: no, but I have a patch that I'm pretty sure will be acceptable, waiting to submit
21:25.52 Notify 02GCI:brlcad * 6353662997692416 Design a BRL-CAD splash screen #3: mukul - Mukul, Did you design that background? If you did not, you should provide attribution (ie, credit) when you create...
21:26.08 andromeda-galaxy oh wait, sorry, looking at something from last ngiht :-)
21:26.47 Notify 02GCI:brlcad * 6353662997692416 Design a BRL-CAD splash screen #3: 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...
21:26.57 Notify 02GCI:brlcad * 6353662997692416 Design a BRL-CAD splash screen #3: Deadline extended - The deadline of the task has been extended with 2 days and 2 hours.
21:35.42 *** join/#brlcad merzo (~merzo@245-112-133-95.pool.ukrtel.net)
21:36.41 Notify 02GCI:tannousmarc * 6449379632218112 Create an utility library (LIBBU) API unit test for bomb.c: About - Looked into the bombs.c file, the two functions I came up with that required testing were : ...
21:36.46 Notify 02GCI:tannousmarc * 6449379632218112 Create an utility library (LIBBU) API unit test for bomb.c: Ready for review - The work on this task is ready to be reviewed.
21:37.42 Notify 03BRL-CAD:starseeker * 63708 brlcad/trunk/src/libbn/CMakeLists.txt: Make chull3d.cpp a C++ file
21:38.31 Notify 03BRL-CAD:starseeker * 63709 brlcad/trunk/src/libbn/chull3d.cpp: Try using a std::map for point lookup when assembling hull points
21:38.57 Notify 02GCI:brlcad * 5238032693198848 Implement new API function to reduce duplication #4: testing infrastructure - I would assert that this should still not become public LIBBU api. It could live in the...
21:39.02 Notify 02GCI:brlcad * 5238032693198848 Implement new API function to reduce duplication #4: Task Needs More Work - One of the mentors has sent this task back for more work. Talk to the mentor(s) assigned...
21:40.40 MarcTannous_ I'm done for the day guys, have a great night
21:41.33 Notify 02GCI:brlcad * 5221860329390080 Design a new website favicon! #8: please describe - Rohit, Would you please describe how you made this favicon (in detail)? Thanks!
21:42.43 Notify 02GCI:veskopos * 6424327624851456 Animate our logo! #2: Ready for review - The work on this task is ready to be reviewed.
21:43.28 Notify 02GCI:veskopos * 6424327624851456 Animate our logo! #2: NEW COOL DESIGN - ACCEPT IT PLS - it looks better now. ;)
21:43.58 Notify 02GCI:o7p9bxbnyj * 5238032693198848 Implement new API function to reduce duplication #4: which tests subdirectory? - I did consider moving it into a tests subdirectory, but couldn't because tests...
21:46.03 Ch3ck_ Ljudusika, the photo looks good
21:46.10 Ch3ck_ i'm trying to test the .g file
21:46.22 Ch3ck_ is fixing his brlcad_build directory
21:56.31 Ljudusika OK
21:57.46 Notify 02GCI:o7p9bxbnyj * 5238032693198848 Implement new API function to reduce duplication #4: Ready for review - The work on this task is ready to be reviewed.
22:02.25 Notify 02GCI:Melange * 6424327624851456 Animate our logo! #2: 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...
22:07.31 Notify 02GCI:o7p9bxbnyj * 5238032693198848 Implement new API function to reduce duplication #4: More tests - The new version of the diff has several more tests using it, which I believe includes most of...
22:07.52 Notify 03BRL-CAD:carlmoore * 63710 brlcad/trunk/src/util/pixblend.c: revise the Usage
22:11.16 Ch3ck takes over, Ch3ck_ retires
22:14.35 Notify 03BRL-CAD:starseeker * 63711 brlcad/trunk/src/libbn/chull3d.cpp: update comments
22:16.38 starseeker well, it still took 12 minutes, but it worked: http://brlcad.org/~starseeker/lucy_chull.png
22:17.33 andromeda-galaxy starseeker; great!
22:22.28 starseeker with std::map in the mix it took 12 minutes, without it lucy took about 115 minutes
22:25.21 starseeker definitely someone will need to read up on quickhull and fix this up to work the right way (papers that time convex hulls of even large models use milliseconds for their units)
22:25.48 andromeda-galaxy hmm...
22:26.14 Stragus Mesh reconstruction from point clouds?
22:26.26 starseeker Stragus: no, this is just convex hull
22:26.30 Stragus Ah.
22:26.59 starseeker we're a lot better off than we were (we had no capability for it at the beginning of the week) but it should be a lot faster than it is
22:27.55 starseeker I'm seeing a few indications in the literature that the quickhull algorithm is a refinement of the Clarkson approach, but since we're avoiding looking at qhull that means pulling the paper and seeing what would need to change
22:28.39 starseeker it's one of those things that might prove to be fairly simple once someone understands what needs to be done, or it could be A Lot Of Work - no way to tell in advance
22:29.40 *** join/#brlcad winghouchan (~winghouch@2.120.176.79)
22:30.33 starseeker I'd have to check with brlcad to see if he thought it's even possible to chop something like that up into GCI tasks - it would have to go something like review the quickhull algorithm as published in the paper and compare it to what we've got in chull3d to identify key differences, then (if they are tractable) chop up the implementation update into incremental pieces
22:31.03 starseeker suspects it's too much, especially given how much else needs to be done that is of more immediate benefit
22:34.44 Stragus It seems very difficult to chop such a task into tiny GCI tasks
22:35.13 Stragus That's a task for one programmer to do up to the end
22:35.29 starseeker yeah, that's my thought too
22:43.50 starseeker however, now that it *is* working (however slowly) we can try porting the geometrictools minimumvolumebox3 routine to libbn: http://www.geometrictools.com/GTEngine/Include/GteMinimumVolumeBox3.inl
22:51.31 starseeker reads... we've got the 2d minimum box routine
22:53.11 starseeker hmm... that's actually a case I'm not set up for - to handle when the 3d routine gets fed something coplanar or otherwise degenerate.
22:56.33 Ch3ck looks like i can't join #brlcad-gcimentors?
22:59.35 Ch3ck starseeker: is the chull3d.c in src? looks like my cmake is failing as a result of this
23:01.21 Ch3ck here is a compilation error i'm getting http://ur1.ca/j3jbz
23:02.24 starseeker Ch3ck: it should be updated to be looking for chull3d.cpp
23:02.29 starseeker try an svn update
23:03.16 Notify 02GCI:ch3ck * 4797643658100736 Model "BRL-CAD" in BRL-CAD #2: None - I like the png file, hoped if you could show other images rendering with different colors. still working on the .g file.
23:03.16 starseeker huh
23:03.23 starseeker now I'm seeing it too
23:03.24 starseeker one sec
23:04.00 starseeker Oh!
23:04.02 starseeker I see it now
23:04.41 Notify 03BRL-CAD:starseeker * 63712 brlcad/trunk/src/libbn/CMakeLists.txt: No longer need to ignore this, and the file is renamed anyway.
23:04.47 starseeker Ch3ck: that should fix it
23:04.57 starseeker sorry about that
23:07.03 Ch3ck starseeker: aight
23:09.58 Ch3ck starseeker: i still get this error http://ur1.ca/j3jdn
23:11.26 Notify 03BRL-CAD:starseeker * 63713 brlcad/trunk/doc/docbook/system/man1/en/CMakeLists.txt: No g-dot.xml file uploaded?
23:14.09 Notify 02GCI:o7p9bxbnyj * 5238032693198848 Implement new API function to reduce duplication #4: Alternate diff - The -v3-alternate diff (second one; the first one was missing a file, sorry!) file includes...
23:15.51 Ch3ck i have this problem http://ur1.ca/j3jf6
23:16.08 starseeker working it
23:16.23 Ch3ck looks like time headers are messed up
23:16.54 Ch3ck feeling sleepy, will surely review this task tomorrow
23:17.22 starseeker think I got it
23:17.26 Notify 03BRL-CAD:starseeker * 63714 (brlcad/trunk/src/libbn/tests/bn_chull3d.c brlcad/trunk/src/libbu/y2038/time64.c): rename variables to avoid shadowing warnings
23:17.32 Notify 02GCI:ch3ck * 4797643658100736 Model "BRL-CAD" in BRL-CAD #2: Deadline extended - The deadline of the task has been extended with 0 days and 10 hours.
23:19.08 starseeker Ch3ck: builds here now
23:26.39 Notify 02GCI:icyjuban * 5173865613361152 Design a web page that overviews the BRL-CAD contributors: Is this shortened enough? - The BRL-CAD development team would like to take this opportunity to thank...
23:28.30 Notify 03BRL-CAD:starseeker * 63715 brlcad/trunk/src/libbn/chull3d.cpp: Avoid segfaulting, but the correct answer is actually to return the set of points the way we would for bn_3d_coplanar_chull - in fact, this may wind up replacing that function once we handle it properly...
23:32.14 Notify 03BRL-CAD:starseeker * 63716 brlcad/trunk/src/libbn/chull3d.cpp: return the dimension of the convex hull as the return value of bn_3d_chull.
23:53.49 Notify 03BRL-CAD:starseeker * 63717 (brlcad/trunk/include/bn/chull.h brlcad/trunk/src/libbn/chull3d.cpp and 2 others): Add some handling for the 2d case in chull3d - rather than replace bn_3d_coplanar_chull, just use it to assemble the vertices into the polygon hull we're looking for. Update return code expectations
23:56.04 Notify 03BRL-CAD:starseeker * 63718 brlcad/trunk/src/libbn/chull3d.cpp: update comment

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