| 00:00.31 | starseeker | is looking, but doesn't see any immediate indication that anyone has pdksh compiling with MSVC... |
| 00:03.33 | brlcad | http://en.wikipedia.org/wiki/Korn_shell says some versions of windows already include a posix-compliant ksh, and http://en.wikipedia.org/wiki/UWIN |
| 00:03.43 | CIA-133 | BRL-CAD: 03starseeker * r46620 10/brlcad/trunk/ (CMakeLists.txt misc/CMake/test_srcs/time.c.in): Yay! RFC2822-ify the DATE stamp code, and go with Sean's idea of linking the DATE regeneration to the COUNT regeneration. |
| 00:04.23 | brlcad | the main portability issue is getting a pseudo terminal interface, which was the other part of that equation |
| 00:04.24 | starseeker | urm. CPL? not very familiar with that license |
| 00:04.54 | brlcad | you need a terminal to run a shell, if you have that, porting the underlying shell is pretty trivial |
| 00:05.25 | starseeker | winces |
| 00:05.52 | starseeker | so I guess it boils down to whether implementing a terminal emulator in Windows is easer than the sh->tclsh port |
| 00:06.18 | brlcad | meh |
| 00:06.24 | brlcad | the shell scripts are all developer infrastructure |
| 00:06.56 | starseeker | would be nice to do regress/bench with MSVC |
| 00:07.00 | brlcad | not a big deal if you can't test some parts of the system on Windows, you can everywhere else |
| 00:07.11 | brlcad | bench is a diff issue, already on it |
| 00:07.51 | starseeker | ah, cool :-) |
| 00:08.27 | brlcad | sure it would be nice to have all of regress on windows, but not "absolutely must have necessary" .. and there is cygwin/mingw if we really want to be pedantic, it's pretty trivially possible |
| 00:09.39 | brlcad | let a windows diehard rewrite them :) |
| 00:11.21 | starseeker | heh |
| 00:21.46 | starseeker | brlcad: how does that new setup look? |
| 00:22.02 | starseeker | seems to be behaving OK here, so far |
| 00:33.42 | brlcad | been working on getting the final nurbs rerun kicked off again for another 10-20 hours of processing |
| 00:35.30 | brlcad | now that it's done and churning, hopefully it'll be a clean run and the stats can be distributed as a baseline |
| 00:36.07 | brlcad | i'll be kicking off some builds here in a couple minutes so i'll let you know how it goes :0 |
| 02:44.46 | CIA-133 | BRL-CAD: 03173.234.97.161 07http://brlcad.org * r3149 10/wiki/User:108.62.162.154: New page: [http://www.galeriabali.pl/pl/Ryby owoce morza] Szeroki wybor w karcie dan spowoduje, ze znajdziesz cos wyjatkowego a ciekawosc jak smakuja inne dania sprawi, ze nie beziesz mogl sie docze... |
| 03:17.31 | CIA-133 | BRL-CAD: 03Sean 07http://brlcad.org * r0 10/wiki/Special:Log/delete: deleted "[[User:108.62.162.154]]" |
| 03:17.32 | CIA-133 | BRL-CAD: 03Sean 07http://brlcad.org * r0 10/wiki/Special:Log/block: blocked [[User:173.234.97.161]] with an expiry time of infinite (anonymous users only, account creation disabled): Spamming links to external sites |
| 03:27.47 | CIA-133 | BRL-CAD: 03starseeker * r46621 10/brlcad/trunk/src/libged/CMakeLists.txt: need to ignore the simulate files if we're not building them |
| 03:32.13 | starseeker | brlcad: I'm a bit confused about how to approach an aspect of the distcheck rule - I can ignore files that the build logic doesn't know about and not include them in the tarball, but how do I distinguish between "junk" files that should be ignored and files that the developer genuinely forgot to include? (like the simulate files if bullet isn't found) |
| 03:41.38 | CIA-133 | BRL-CAD: 03starseeker * r46622 10/brlcad/trunk/misc/CMake/distcheck_buildsys.cmake.in: Improvements to the distcheck logic - haven't quite got to activting the CPack tweaking logic, but close. In the meantime, better error reporting. |
| 03:54.30 | CIA-133 | BRL-CAD: 03starseeker * r46623 10/brlcad/trunk/misc/CMake/distcheck_buildsys.cmake.in: more TODO notes for dist |
| 04:02.02 | brlcad | starseeker: the autotools build checks the svn manifest -- if they're in svn, then they should be in the dist |
| 04:02.51 | starseeker | urm. So I need to parse the svn manifests. |
| 04:04.00 | brlcad | 12-liner coming: |
| 04:04.03 | brlcad | dist-hook: |
| 04:04.03 | brlcad | <PROTECTED> |
| 04:04.07 | brlcad | <PROTECTED> |
| 04:04.09 | brlcad | <PROTECTED> |
| 04:04.12 | brlcad | <PROTECTED> |
| 04:04.14 | brlcad | <PROTECTED> |
| 04:04.17 | brlcad | <PROTECTED> |
| 04:04.19 | brlcad | <PROTECTED> |
| 04:04.22 | brlcad | <PROTECTED> |
| 04:04.24 | brlcad | <PROTECTED> |
| 04:04.27 | brlcad | <PROTECTED> |
| 04:04.29 | brlcad | <PROTECTED> |
| 04:04.55 | brlcad | that's how autotools does it -- gets the list of names from the ".svn/entries" files, then looks to see if they're in the source dist that was built |
| 04:05.10 | brlcad | if missing, reports then halts |
| 04:05.39 | brlcad | that script-fu should convert pretty simply to cmake funcs |
| 04:06.31 | starseeker | those entries files are per .svn directory? |
| 04:06.40 | brlcad | yes |
| 04:07.04 | brlcad | they're simple text, pop one up |
| 04:07.52 | brlcad | basically xml blocks with name="filename" identifiers |
| 04:08.18 | starseeker | what does it do without the .svn directories then? |
| 04:09.06 | brlcad | no manifest, no way to know what's actually missing |
| 04:09.09 | brlcad | so it does nothing |
| 04:09.12 | starseeker | ah |
| 04:09.24 | starseeker | that's actually the case I have implemented then :-/ |
| 04:09.38 | brlcad | it could report what files were found, but meh .. 99% of the time, we have the list |
| 04:10.10 | starseeker | nods - I'll leave what I have for that case (no .svn entities files found) and figure out the .svn manifest part |
| 04:10.30 | brlcad | yeah, that's probably very reasonable behavior -- no svn files, report and could even halt |
| 04:10.56 | brlcad | but if there is a manifest, it probably shouldn't halt on the unknowns |
| 04:11.29 | starseeker | nods |
| 04:11.38 | CIA-133 | BRL-CAD: 03starseeker * r46624 10/brlcad/trunk/CMakeLists.txt: Echo some messages to identify various stages of distcheck. |
| 04:11.39 | brlcad | arguably useful to even report them since our actual "correctness" rule is only to make sure what's in svn is in the dist |
| 04:11.41 | starseeker | that's what I was figuring - with svn manifests, it's sane to continue |
| 04:12.04 | brlcad | but even reporting them could be good sanity too |
| 04:12.18 | starseeker | <shrugs> at least flags the trouble spot |
| 04:12.28 | brlcad | the biggest catch is finding things you *know* are in svn but are missing |
| 04:12.31 | brlcad | yeah |
| 04:13.15 | starseeker | I doubt I'll be ready with an svn based distcheck for this release :-( |
| 04:13.36 | brlcad | that's fine |
| 04:13.45 | brlcad | planning on distchecking with both anyways |
| 04:13.57 | brlcad | if you want to get fancy -- one thing I never got around to fixing is if you svn delete a file, the entry remains (with a deleted flag in the entries record) |
| 04:14.45 | starseeker | nods - sounds worth doing |
| 04:14.55 | brlcad | rarely hit it in practice so not a big deal, but one more place to knock the autotools build down a notch |
| 04:15.24 | brlcad | would be a bigger issue if we deleted/renamed files more frequently between releases (and not just whole dirs) |
| 04:16.17 | starseeker | nods |
| 04:16.29 | brlcad | oh wow, so I thought it was still relinking everything .... |
| 04:17.04 | brlcad | but turns out it's just that slow to walk over all 400+ targets to evaluate and print "Built target ..." |
| 04:17.24 | starseeker | O.o |
| 04:17.41 | brlcad | cool, so looks like it works |
| 04:17.47 | brlcad | to do nothing: |
| 04:17.48 | brlcad | Elapsed compilation time: 1 minute 7 seconds |
| 04:18.07 | starseeker | what's the autotools time to do nothing? |
| 04:18.39 | starseeker | actually, with all the docbook targets present (even without pdf) I think it's over a thousand targets... |
| 04:18.57 | brlcad | yeah, it is |
| 04:19.08 | starseeker | that was fun in Visual Studio |
| 04:19.14 | brlcad | good question, I don't recall off the top of my head but was less |
| 04:19.38 | brlcad | counts the number of targets |
| 04:19.40 | starseeker | bets it's about 50-60%, mostly due to faster docbook |
| 04:20.14 | CIA-133 | BRL-CAD: 03108.62.166.91 07http://brlcad.org * r3150 10/wiki/User:108.62.173.114: New page: [http://www.willa-sloneczna.euroadres.pl noclegi kazimierz dolny] .Kiedy mysli sie nad slowami dluzszy pobyt to prawdopodobnie rozwazasz o kilku tygodniach lub miesiacach. W rzeczywistosc... |
| 04:20.22 | brlcad | sunufa! |
| 04:20.38 | brlcad | looks like I know what I'm doing tomorrow |
| 04:20.56 | CIA-133 | BRL-CAD: 03Sean 07http://brlcad.org * r0 10/wiki/Special:Log/delete: deleted "[[User:108.62.173.114]]" |
| 04:21.03 | starseeker | oddly enough, I'll be doing something similar (cleaning bathrooms ;-) |
| 04:21.06 | CIA-133 | BRL-CAD: 03Sean 07http://brlcad.org * r0 10/wiki/Special:Log/block: blocked [[User:108.62.166.91]] with an expiry time of infinite (anonymous users only, account creation disabled): Spamming links to external sites |
| 04:21.48 | brlcad | good thing is that it takes less than 30 seconds to revert and block .. but still, pita! |
| 04:21.52 | brlcad | 1175 targets :) |
| 04:22.32 | starseeker | <evil laugh>MWA_HAHAHAHAHA</evil laugh> |
| 04:23.09 | starseeker | hmm - time to do nothing on my Linux box here - 4 seconds |
| 04:24.04 | starseeker | can't wait to post a table somewhere with Linux, Mac and Windows build times side by side |
| 04:25.03 | starseeker | brlcad: you're feeding it the make -j# for the empty case? |
| 04:26.22 | starseeker | tries on a mac once... |
| 04:26.35 | brlcad | I had the first time, but not that timed one |
| 04:27.00 | starseeker | it should make a difference |
| 04:27.43 | brlcad | probably, but not a huge one .. this laptop only has two procs and the disks are too fast and os x is an i/o bottleneck |
| 04:27.52 | starseeker | ah |
| 04:27.53 | brlcad | 1min9sec |
| 04:28.02 | starseeker | phooy |
| 04:28.03 | brlcad | so .. bout the same |
| 04:33.16 | starseeker | brlcad: so it looks like we're closing in - svn manifests for distcheck and turning the MSVC stuff into tests are the two biggies I know about |
| 04:33.28 | starseeker | oh, and finishing updating the docs |
| 04:33.38 | brlcad | msvc into tests? |
| 04:33.46 | starseeker | config_win going byebye |
| 04:33.55 | brlcad | ah great |
| 04:35.17 | brlcad | should probably hit the docs first, the sooner the deprecation statements get added, the sooner we can remove the build in a couple months |
| 04:35.17 | starseeker | is considering making a small convenience NSIS installer for the Windows xsltproc |
| 04:36.39 | starseeker | knows you had a list of things to add statements to - configure obviously, and autogen.sh |
| 04:36.50 | starseeker | is that in a TODO somewhere? |
| 04:37.01 | starseeker | (just so I know what to hit) |
| 04:37.41 | brlcad | those two are the main ones |
| 04:38.22 | starseeker | did you want to add notes in the current documentation files, or just swap them out when the time comes? |
| 04:38.46 | brlcad | just swap, if it's deprecated then there's no need to educate on the old system |
| 04:39.07 | brlcad | the statements are just added to the remnants visible for those that did learn the old way |
| 04:39.09 | starseeker | nods - oh, is the new Bundled/System/Auto option setup to your liking? |
| 04:39.23 | brlcad | haven't dove that deep again |
| 04:39.31 | starseeker | ah, k |
| 04:39.37 | brlcad | hm, libpc bustage |
| 04:41.52 | starseeker | probably the new boost |
| 04:41.57 | brlcad | yep |
| 04:42.26 | starseeker | grim? |
| 04:43.09 | brlcad | trivial |
| 04:43.13 | CIA-133 | BRL-CAD: 03brlcad * r46625 10/brlcad/trunk/configure.ac: reflect new location of boost headers |
| 04:43.37 | starseeker | ah :-) |
| 04:45.07 | brlcad | the new normalized rays/s is pretty cool |
| 04:45.52 | starseeker | oh, in your benchmark? |
| 04:46.04 | brlcad | yeah |
| 04:46.25 | starseeker | tries clang as long as he's on the mac... |
| 04:46.42 | starseeker | how're nurbs stacking up? |
| 04:47.03 | brlcad | pulls rays/s from the summary for a given trace, but then also calculates how many pixels actually involved geometry and weights the rtfm value |
| 04:47.25 | starseeker | sweet |
| 04:47.32 | starseeker | was that the math you were scripting earlier? |
| 04:48.00 | brlcad | related, but that was actually for the projected area and volume cases |
| 04:48.12 | starseeker | yow |
| 04:48.47 | brlcad | they have to do something a bit more tricky since I'm trying to report a relative deviation as percentage error |
| 04:49.00 | starseeker | kinda sounds like stuff mged should have had tools for |
| 04:49.44 | brlcad | mmm, dunno |
| 04:49.52 | brlcad | not seeing general utility |
| 04:50.01 | brlcad | for the pix comparisons, sure |
| 04:51.00 | brlcad | but there the math is easy, just 1 - (num_pixels_wrong / num_pixels) |
| 04:51.33 | starseeker | ah - guess I was thinking something like having rt report the "pixels involving geometry" metric |
| 04:51.56 | brlcad | ah, maybe but that was also pretty easy to script |
| 04:52.01 | brlcad | with existing tools |
| 04:52.18 | starseeker | cool |
| 04:52.18 | brlcad | the hard one is the volume and error |
| 04:52.48 | starseeker | hmm: src/conv/g-vrml.c:99:21: warning: using extended field designator is an extension [-pedantic] |
| 04:52.50 | brlcad | amount_vol_wrong / amount_vol doesn't give what you want |
| 04:53.20 | brlcad | well, at least doesn't give what *I* want :) .. that just reports the % difference |
| 04:53.27 | starseeker | nods |
| 04:53.39 | starseeker | sounds like a problem for our resident math genius ;-) |
| 04:53.58 | brlcad | ended up with this little bit of goodness: dc -e "1k 100.0 $bvol $vol - d * v $bvol / 100.0 * - d [0] sa 0.0 >a p" |
| 04:54.08 | starseeker | O.o |
| 04:55.25 | starseeker | watches SdaiCONFIG_CONTROL_DESIGN.h barf warnings and reflects that he REALLY needs to try the newer SCL code... |
| 04:57.00 | brlcad | which is basically: 100.0 - abs(actual_volume - real_volume) / actual_volume * 100.0 but then clamping the value to 0.0 if it goes negative |
| 04:57.30 | starseeker | nods |
| 04:57.47 | brlcad | actual is tiny and real is huge, then the "difference factor" is going to be huge |
| 04:58.28 | starseeker | so it's amplifying the difference? |
| 04:58.29 | brlcad | basically means objects that are smaller than the baseline will have an error value that is the percentage of the actual volume |
| 04:58.49 | brlcad | an object half as big as it's supposed to be is going to report a volume error of 50% |
| 04:58.59 | starseeker | ah, gotcha |
| 05:00.43 | brlcad | but then for objects larger than they're supposed to be, it'll report 50% when it's 50% larger and 100% error when it's double (*or* more) in size |
| 05:00.58 | brlcad | that's the case that was hard to figure out |
| 05:01.13 | brlcad | since it can be unboundedly larger than the baseline |
| 05:01.46 | brlcad | good stuff |
| 05:01.52 | starseeker | are you seeing cases much larger than baseline? |
| 05:02.00 | starseeker | is curious how that could come about... |
| 05:02.37 | brlcad | there are some clear failures, but most are slight larger/smaller deviations |
| 05:02.42 | brlcad | bots |
| 05:02.58 | brlcad | the calcs will work for any case though |
| 05:03.10 | starseeker | nods - that's cool :-) |
| 05:04.44 | starseeker | aw - clang build busted on obj-g_new.c |
| 05:05.26 | starseeker | ok, I gotta get out of here |
| 05:05.59 | starseeker | brlcad: let me know if you see an problems with the new COUNT setup, but I think that will do the trick for most situations we're likely to see |
| 05:11.10 | CIA-133 | BRL-CAD: 03starseeker * r46626 10/brlcad/trunk/src/conv/obj-g_new.c: size_t -> size_t * for comparison |
| 05:11.20 | starseeker | and we build with clang again :-) |
| 05:11.27 | starseeker | hits the road |
| 05:12.09 | brlcad | it's looking great |
| 07:43.37 | *** join/#brlcad ibot (~ibot@rikers.org) | |
| 07:43.37 | *** topic/#brlcad is BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || #brlcad logs: http://ibot.rikers.org/%23brlcad/ || BRL-CAD release 7.20.2 is posted (20110701) || BRL-CAD is participating in the ESA Summer of Code in Space! | |
| 09:07.01 | *** join/#brlcad packrat (~packrator@99-67-225-40.lightspeed.livnmi.sbcglobal.net) | |
| 09:50.29 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 12:07.35 | *** join/#brlcad abhi2011 (~chatzilla@wlan-145-94-185-071.wlan.tudelft.nl) | |
| 13:28.28 | starseeker | oh, beautiful - apparently a Makefile doesn't know what N is in make -j N |
| 13:28.33 | starseeker | http://old.nabble.com/MAKEFLAGS-var-does-not-show-%22-j%22-param-----td15983337.html |
| 13:36.33 | brlcad | starseeker: that's not what that thread says |
| 13:38.55 | brlcad | says make-w32 doesn't implement the parallel build with the "job server" mechanism, so it redoes the parameter for some other mechanism on windows |
| 13:45.14 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) | |
| 14:02.59 | *** join/#brlcad abhi2011_ (~chatzilla@wlan-145-94-185-071.wlan.tudelft.nl) | |
| 14:06.32 | starseeker | brlcad: not the windows part, the part where $(MAKEFLAGS) will never show the -jN option in parallel |
| 14:07.09 | starseeker | needs to introspect in the Makefile what the -j option passed in was, and it apparently can't be accessed at all |
| 14:08.25 | starseeker | just gives back the --jobserver-fds=3,4 -j contents regardless |
| 14:08.29 | starseeker | confirms that in tests here |
| 14:16.16 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 14:20.01 | brlcad | starseeker: ah, I see what you're getting at |
| 14:20.12 | *** join/#brlcad n_reed (~molto_cre@BZ.BZFLAG.BZ) | |
| 14:20.38 | brlcad | that's undoubtedly because the jobserver is being entrusted with that detail |
| 14:25.08 | brlcad | starseeker: couldn't you just set it up so that it just toggles off a cmake variable? cmake -DJOBS=3 |
| 14:25.38 | brlcad | then you can access that to add -j$JOBS during dist or wherever |
| 14:27.01 | starseeker | that would mean doing something like cmake -P distcheck.cmake -DJOBS=3 instead of make -j3 distcheck |
| 14:27.18 | starseeker | workable, but it takes distcheck outside the make system |
| 14:27.39 | brlcad | not necessarily |
| 14:28.30 | brlcad | so you could run "cmake -DJOBS=3 path/to/src" during setup, but it doesn't actually do anything with it other than stash it |
| 14:28.56 | starseeker | oh, set the job count at cmake configure time instead of make time? |
| 14:29.04 | brlcad | then "make distcheck" would use the stashed value, or at worst maybe "make distcheck JOBS=3" instead |
| 14:29.18 | brlcad | to override at make-time |
| 14:29.28 | starseeker | isn't sure the override would work... |
| 14:29.33 | brlcad | another options is just "-j" |
| 14:29.46 | brlcad | -j without a number means "go hog wild" |
| 14:29.50 | starseeker | heh |
| 14:30.30 | starseeker | even that's tricky though - accessing make values to feed to CMake isn't really something the CMake generated scripts are set up for |
| 14:30.47 | starseeker | has an email into the CMake list to ask about it |
| 14:31.31 | starseeker | ideal case would be to add the ability to CMake generated make files to respect some variable (JOBS or CPUS) passed in at make time |
| 14:32.53 | starseeker | maybe have a ${CMAKE_CPU_COUNT} variable at CMakeLists.txt level that could be added to the --build line, e.g. ${CMAKE_COMMAND} --build -j${CMAKE_CPU_COUNT} and then have the CMake makefiles do the right thing |
| 14:35.51 | brlcad | that's kind of what I mean, just s/CMAKE_CPU_COUNT/JOBS/ |
| 14:36.15 | starseeker | nods - that would take support on the CMake generator backend |
| 14:36.46 | brlcad | CMAKE_NCPU would be a "proper" prefixed form I guess :) |
| 14:36.52 | starseeker | hehe |
| 14:36.57 | starseeker | will suggest that |
| 14:37.26 | starseeker | make NCPU=3 isn't make -j3, but it is probably as close as make will let us get |
| 15:04.07 | CIA-133 | BRL-CAD: 03brlcad * r46630 10/brlcad/trunk/ (include/bu.h src/libbu/fnmatch.c src/librt/search.c): |
| 15:04.08 | CIA-133 | BRL-CAD: expose the documented minmum inteface for bu_fnmatch() including the various |
| 15:04.08 | CIA-133 | BRL-CAD: flags that a caller might set as well as the nomatch return value. remove the |
| 15:04.08 | CIA-133 | BRL-CAD: other BU_-prefixed symbols from the implementation since they're not public api. |
| 15:04.08 | CIA-133 | BRL-CAD: renamed BU_CASEFOLD to BU_FNMATCH_CASEFOLD in the process for bu api |
| 15:04.08 | CIA-133 | BRL-CAD: consistency. |
| 15:18.35 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 15:33.54 | brlcad | starseeker: do you remember why you made the plans in librt/search.c void* instead of db_plan_t* ? |
| 15:35.25 | starseeker | might have had something to do with how I was calling things from libged |
| 15:37.58 | CIA-133 | BRL-CAD: 03starseeker * r46631 10/brlcad/trunk/ (CMakeLists.txt autogen.sh configure.ac): Make the depends explicit for these initial commands - looks like ninja actually caught one out. |
| 15:38.04 | starseeker | ah crud |
| 15:38.24 | starseeker | brlcad: well, how do those look for autogen.sh and configure.ac deprecation warnings? |
| 15:40.00 | CIA-133 | BRL-CAD: 03brlcad * r46632 10/brlcad/trunk/src/librt/search.c: lets not add 50 new functions to librt's API. mark all of the implementation functions as HIDDEN. |
| 15:44.05 | CIA-133 | BRL-CAD: 03brlcad * r46633 10/brlcad/trunk/src/librt/search.c: ws indent consistency cleanup |
| 15:44.38 | CIA-133 | BRL-CAD: 03brlcad * r46634 10/brlcad/trunk/include/raytrace.h: migrated command doc from impl to header for db_search_formplan() |
| 15:53.30 | CIA-133 | BRL-CAD: 03brlcad * r46635 10/brlcad/trunk/include/common.h: better document why we use HIDDEN, emphasize that it's not really appropriate for front-end code but is expected for libraries. |
| 15:53.55 | CIA-133 | BRL-CAD: 03brlcad * r46636 10/brlcad/trunk/src/libged/search.c: mark the private functions as HIDDEN |
| 15:55.26 | CIA-133 | BRL-CAD: 03starseeker * r46637 10/brlcad/trunk/ (autogen.sh configure.ac): these got sucked in by mistake |
| 15:56.28 | brlcad | starseeker: good start, but a few changes -- s/may be/will be/ |
| 15:56.39 | starseeker | ah, right |
| 15:56.44 | brlcad | and "Please use .." doesn't really say anything |
| 15:57.14 | brlcad | they're using what they know, so we should tell them what the new command is AND where to read for more info |
| 15:57.39 | starseeker | k |
| 16:33.50 | CIA-133 | BRL-CAD: 03starseeker * r46638 10/brlcad/trunk/CMakeLists.txt: Try a recommendation from Dave Cole of CMake - this is (more or less) what they do in their ExternalProject code to handle subbuilding |
| 16:44.25 | brlcad | starseeker: is ninja one of the cmake devs or some system?? |
| 17:02.12 | CIA-133 | BRL-CAD: 03brlcad * r46639 10/brlcad/trunk/src/libbu/ (argv.c vls.c): move bu_argv_from_string() from vls.c to argv.c |
| 17:06.12 | starseeker | oh, sorry - new CMake generater for this: http://martine.github.com/ninja/manual.html |
| 17:06.45 | brlcad | ah |
| 17:07.05 | starseeker | in development at the moment |
| 17:11.33 | starseeker | http://www.cmake.org/pipermail/cmake/2011-September/046145.html |
| 17:11.53 | brlcad | nods |
| 17:12.11 | starseeker | faster building ftw, if it actually works |
| 17:15.37 | CIA-133 | BRL-CAD: 03brlcad * r46640 10/brlcad/trunk/ (include/bu.h src/libbu/argv.c): make bu_argv_from_string() work with size_t instead of int for the size parameters. move the decl over with the other argv functions. need ctype.h header for isspace() too. |
| 17:31.19 | CIA-133 | BRL-CAD: 03bob1961 * r46641 10/brlcad/trunk/src/tclscripts/lib/TkTable.tcl: Added a -singleSelectCallback option. If specified this essentially puts the table widget into a mode where only one row at a time can be selected. |
| 17:35.03 | *** join/#brlcad abhi2011 (~chatzilla@wlan-145-94-185-071.wlan.tudelft.nl) | |
| 18:06.25 | CIA-133 | BRL-CAD: 03173.234.121.94 07http://brlcad.org * r3151 10/wiki/Haha_Spoil_her_reputation_81: New page: [[Image:reputation_management_1273.jpg|thumb|]] ya estamos listas pa salir al shows miados aki en Tijuana, agarrensen poke este pedo va a estar mamalon x 1 vez "las you you" !!!! NCBI ROF... |
| 18:12.34 | CIA-133 | BRL-CAD: 03Sean 07http://brlcad.org * r0 10/wiki/Special:Log/delete: deleted "[[Haha Spoil her reputation 81]]" |
| 18:12.45 | CIA-133 | BRL-CAD: 03Sean 07http://brlcad.org * r0 10/wiki/Special:Log/block: blocked [[User:173.234.121.94]] with an expiry time of infinite (anonymous users only, account creation disabled): Spamming links to external sites |
| 18:32.37 | *** join/#brlcad abhi2011_ (~chatzilla@wlan-145-94-185-071.wlan.tudelft.nl) | |
| 18:37.23 | brlcad | so there are now a couple new wiki extensions installed |
| 18:38.16 | brlcad | one is a new blacklist extension that pulls the massive lists maintained by mediawiki and wikipedia folks (they are separate lists) that block posts that contain blacklisted url/content |
| 18:39.06 | brlcad | another is a Q&A tension -- anyone care to suggest some good hard cad/math-related questions? |
| 18:44.40 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 18:45.45 | n_reed | not sure i understand the notion of a Q&A extension |
| 18:45.56 | n_reed | is the idea to post questions and answers |
| 18:46.04 | n_reed | or to get people to answer questions |
| 18:47.12 | brlcad | we pose a question, they have to answer it correctly to modify the wiki |
| 18:47.40 | brlcad | like "What is the name of this software?" answer: BRL-CAD |
| 18:48.21 | n_reed | so like a captcha, but harder |
| 18:49.14 | brlcad | what is the square root of 1000-100? |
| 18:49.15 | brlcad | sure |
| 18:50.25 | n_reed | are you asking for suggesstions because you want to populate a list of options? |
| 18:53.29 | brlcad | that is the idea |
| 18:54.52 | n_reed | i'll ask the obvious math person to send you some |
| 18:56.15 | brlcad | mere (CAD) mortals should be able to answer them |
| 18:56.38 | brlcad | I already have about 10, probably enough -- more just whether others wanted to add a couple |
| 19:22.31 | CIA-133 | BRL-CAD: 0368.34.98.23 07http://brlcad.org * r3152 10/wiki/Testing_something_bogus: New page: This is a test. |
| 19:23.26 | CIA-133 | BRL-CAD: 0368.34.98.23 07http://brlcad.org * r3153 10/wiki/Testing_something_bogus: |
| 19:28.22 | starseeker | phew - parallel distcheck |
| 19:28.55 | CIA-133 | BRL-CAD: 03bob1961 * r46642 10/brlcad/trunk/ (11 files in 5 dirs): Added more editing support for pipes (i.e. append, prepend, delete, move, select and scale). |
| 19:31.18 | starseeker | re-enables the "make your life miserable" parts until he gets proper svn manifest checking working |
| 19:40.42 | CIA-133 | BRL-CAD: 03bob1961 * r46643 10/brlcad/trunk/src/libged/Makefile.am: Added edpipe.c to libged/Makefile.am |
| 19:41.02 | CIA-133 | BRL-CAD: 0368.34.98.23 07http://brlcad.org * r3154 10/wiki/Testing_something_bogus: |
| 19:42.14 | CIA-133 | BRL-CAD: 03Sean 07http://brlcad.org * r0 10/wiki/Special:Log/delete: deleted "[[Testing something bogus]]": just testing |
| 19:44.36 | brlcad | so we'll see if that helps -- recaptcha is no longer enabled and logged in users still have no captcha |
| 19:49.33 | starseeker | brlcad: nice, thanks! |
| 19:52.40 | starseeker | pwd |
| 19:52.42 | starseeker | whoops |
| 20:03.26 | starseeker | hmm: http://www.opencascade.org/about/news/issue173/ |
| 20:03.40 | starseeker | wonders if that constitutes a glimmer of hope or not |
| 20:05.49 | CIA-133 | BRL-CAD: 03starseeker * r46644 10/brlcad/trunk/CMakeLists.txt: turn back on the repo checks. distcheck will need to become its own CMake file I think - this is getting too complex for one custom target. |
| 20:10.23 | brlcad | probably in response to OCE |
| 20:11.15 | starseeker | almost certainly - question will be whether it's just words or if they actually change |
| 20:11.33 | brlcad | they need to start with the license itself |
| 20:11.37 | starseeker | would be awesome if they went LGPL |
| 20:11.59 | brlcad | or better |
| 20:13.32 | starseeker | dunno how technically compatible we would be even if the license were ironed out, but they do have some bits we don't right now... |
| 20:41.23 | brlcad | and vice versa, but cleanly integrating pieces from two large codebases is arguably more work than implementing from scratch for each |
| 20:41.54 | brlcad | better to collaborate on the pieces that can be fully abstracted from both (like SCL) |
| 20:42.19 | CIA-133 | BRL-CAD: 03r_weiss * r46645 10/brlcad/trunk/include/bn.h: |
| 20:42.19 | CIA-133 | BRL-CAD: Updated include file 'bn.h' to enable the prototype versions of functions |
| 20:42.19 | CIA-133 | BRL-CAD: bn_isect_lseg3_lseg3, bn_isect_line3_line3, and bn_isect_line_lseg and remove |
| 20:42.19 | CIA-133 | BRL-CAD: the original versions. This change is the first of three. The remaining changes |
| 20:42.19 | CIA-133 | BRL-CAD: will be to files 'plane.c' and 'nmg_tri.c'. |
| 20:44.27 | CIA-133 | BRL-CAD: 03r_weiss * r46646 10/brlcad/trunk/src/libbn/plane.c: Updated file 'plane.c' to enable the prototype versions of functions bn_isect_lseg3_lseg3, bn_isect_line3_line3, and bn_isect_line_lseg. The original functions are removed. |
| 20:47.28 | CIA-133 | BRL-CAD: 03r_weiss * r46647 10/brlcad/trunk/src/librt/primitives/nmg/nmg_tri.c: Updated file 'nmg_tri.c' to change function calls from the prototype names of functions bn_isect_lseg3_lseg3, and bn_isect_line3_line3 back to their original names. |
| 21:04.07 | brlcad | http://vim.wikia.com/wiki/Remove_unwanted_spaces |
| 21:04.07 | brlcad | http://vim.wikia.com/wiki/Highlight_unwanted_spaces |
| 21:04.23 | brlcad | possibly of interest to some of the vimragers in here |
| 21:08.09 | CIA-133 | BRL-CAD: 03brlcad * r46648 10/brlcad/trunk/src/ (70 files in 70 dirs): ws consistency removing trailing ws and much more (see sh/ws.sh) |
| 21:21.36 | CIA-133 | BRL-CAD: 03brlcad * r46649 10/brlcad/trunk/src/ (12 files in 12 dirs): more ws consistency cleanup |
| 21:22.09 | brlcad | starts to see a pattern |
| 21:33.21 | CIA-133 | BRL-CAD: 03bob1961 * r46650 10/brlcad/trunk/src/libged/edpipe.c: Collapse the functionality of ged_append_pipept() and ged_prepend_pipept() into _ged_append_pipept_common(). |
| 21:36.33 | CIA-133 | BRL-CAD: 03r_weiss * r46651 10/brlcad/trunk/src/librt/primitives/nmg/ (8 files): Updated files nmg_class.c, nmg_fuse.c, nmg_info.c, nmg_inter.c, nmg_mk.c, nmg_mod.c, nmg_rt_isect.c and nmg_tri.c to turn on the prototype triangulation code of nmg primitives. |
| 21:38.48 | CIA-133 | BRL-CAD: 03bob1961 * r46652 10/brlcad/trunk/src/libtclcad/tclcad_obj.c: |
| 21:38.48 | CIA-133 | BRL-CAD: Collapse the functionality of to_mouse_append_pipept() and |
| 21:38.48 | CIA-133 | BRL-CAD: to_mouse_prepend_pipept() into to_mouse_append_pipept_common(). Added a call to |
| 21:38.48 | CIA-133 | BRL-CAD: ged_snap_to_grid() in to_mouse_append_pipept_common for snap-to-grid behavior |
| 21:38.48 | CIA-133 | BRL-CAD: when adding/inserting pipe points. |
| 21:40.38 | CIA-133 | BRL-CAD: 03r_weiss * r46653 10/brlcad/trunk/include/raytrace.h: Updated raytrace.h to turn on the prototype triangulation of nmg primitives. |
| 21:40.50 | brlcad | r_weiss' change really begs for some thorough retesting with the conversion script |
| 21:48.33 | CIA-133 | BRL-CAD: 03r_weiss * r46654 10/brlcad/trunk/src/librt/primitives/ars/ars.c: Updated ars.c to turn on changes which were disabled until the prototype version of nmg triangulation was enabled. |
| 21:52.00 | CIA-133 | BRL-CAD: 03n_reed * r46655 10/brlcad/trunk/src/external/CMakeLists.txt: syntax error |
| 21:59.04 | starseeker | brlcad: the pattern being my files having busted whitespace? |
| 22:00.05 | starseeker | hides |
| 22:02.26 | starseeker | tried to set his editors to behave, really he did... |
| 22:08.29 | starseeker | winces - did Richard see your source release announcement? |
| 22:09.47 | starseeker | makes a note to look into astyle sometime http://astyle.sourceforge.net/ |
| 22:35.12 | CIA-133 | BRL-CAD: 03r_weiss * r46656 10/brlcad/trunk/src/librt/primitives/tor/tor.c: Updated file tor.c to enable removing of zero length edges from torus primitives. This was disabled until the prototype nmg triangulation was made available. |
| 23:09.00 | *** part/#brlcad n_reed (~molto_cre@BZ.BZFLAG.BZ) | |
| 23:45.57 | starseeker | huh - of course it's not quite the same since the tcl pkgIndex building macros aren't firing, but if I turn off docbook and compare: |
| 23:46.50 | starseeker | make: 7m8s, ninja: 6m13s to build all of BRL-CAD on my gentoo box |
| 23:47.13 | starseeker | (debug config) |