| 00:34.22 | *** join/#brlcad stevegt_ (~stevegt@cislunar.TerraLuna.Org) |
| 01:44.47 | louipc | wowzarz |
| 01:47.31 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) |
| 02:13.19 | *** join/#brlcad Ralith (~ralith@S010600221561996a.vc.shawcable.net) |
| 04:23.24 | brlcad | ``Erik: autogen/configure seems to be happy on automake 1.6 |
| 04:25.59 | *** join/#brlcad stevegt_1 (~stevegt@c-69-181-134-76.hsd1.ca.comcast.net) |
| 10:12.36 | *** join/#brlcad mafm (~mafm@193.153.52.142) |
| 12:03.57 | starseeker | tries distcheck on gentoo, crosses fingers - maybe this will be an easy cycle and we can sync to STABLE soon |
| 12:05.34 | starseeker | auuugh |
| 12:05.36 | starseeker | brlcad-7.17.0/src/other/boost/spirit/home/support/iterators/detail/buffering_input_iterator_policy.hpp: file name is too long (max 99); not dumped |
| 12:06.20 | starseeker | will revert that for release then - not in use currently, and we can sort it out later |
| 12:23.18 | *** join/#brlcad R0b0t1 (~Enigma@64-136-219-55.dyn.everestkc.net) |
| 12:23.18 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |
| 12:38.06 | CIA-48 | BRL-CAD: 03starseeker * r41251 10/brlcad/trunk/src/other/boost/ (1147 files in 136 dirs): Revert the boost upgrade - resulting in filenames too long for tar. Will deal with this after the release. |
| 12:38.30 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) |
| 12:39.08 | starseeker | sigh |
| 12:39.11 | starseeker | tries again |
| 12:50.50 | brlcad | heh |
| 12:51.53 | brlcad | starseeker: you should be good to go for release once everything is tested |
| 12:51.57 | brlcad | make sure all user-visible commits since last release are documented (ChangeLog is good for that) |
| 12:52.26 | brlcad | I took care of everything in my queue, but I usually double-check the log |
| 12:54.40 | starseeker | ah, cool :-) |
| 12:54.52 | starseeker | hmm, looks like something weird here: |
| 12:55.02 | starseeker | ../../../include/conf/COUNT:1:1: error: invalid suffix "n" on integer constant |
| 12:55.48 | starseeker | aand... sure enough, there is an n after the 1 |
| 12:55.55 | _psilva_ | so what are you guys using spirit for? |
| 12:56.01 | starseeker | libpc |
| 12:56.15 | starseeker | parametric constraints |
| 12:56.22 | _psilva_ | parsing them? |
| 12:58.35 | brlcad | starseeker: that's something erik's seen on a bsd system |
| 12:58.51 | brlcad | results from configure setting a wrong value for ECHO |
| 12:59.38 | starseeker | ah |
| 12:59.46 | brlcad | printf "%d\n" 123 |
| 12:59.48 | starseeker | _psilva_: evaluating them |
| 12:59.52 | brlcad | vs echo "123\n" |
| 13:00.06 | starseeker | so it's my particular autotools's fault? |
| 13:00.15 | brlcad | *shrug* |
| 13:00.16 | starseeker | hopes |
| 13:00.28 | brlcad | did you autogen.sh on that platform? |
| 13:00.34 | starseeker | I beleve so |
| 13:00.41 | starseeker | will do it totally clean to be sure |
| 13:01.25 | starseeker | oo - might have some static files, hang on |
| 13:02.05 | brlcad | grep ECHO include/conf/Makefile |
| 13:03.19 | starseeker | _psilva_: http://brlcad.org/wiki/Libpc |
| 13:03.46 | starseeker | or whoops, http://brlcad.org/wiki/Libpg_:_A_parametrics/constraint_library |
| 13:03.56 | starseeker | second one looks better |
| 13:04.37 | starseeker | not sure why it's Libpg there... |
| 13:05.15 | brlcad | original name, "parametric geometry" |
| 13:05.27 | brlcad | he was going to make separate geometry objects |
| 13:05.33 | brlcad | told him that was nfg |
| 13:06.32 | brlcad | so it was reworked to be libpc, which librt uses |
| 13:17.26 | brlcad | thinks the GCI list is going to make a good "how to get started contributing to BRL-CAD" list even if we're not accepted |
| 13:18.17 | brlcad | starseeker: oh yeah, note the outstanding TODO item for Windows |
| 13:28.53 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r2332 10/wiki/Google_Code_In/Project_Ideas: BU jumping |
| 13:29.39 | starseeker | brlcad: yeah, ECHO = printf %s\n |
| 13:30.11 | starseeker | so it's not our fault :-) |
| 13:30.30 | starseeker | and the failure was from a clean autogen |
| 13:31.33 | ``Erik | I've seen the issue on my mac, too |
| 13:31.53 | ``Erik | and on a linux box iirc |
| 13:33.24 | ``Erik | it also shows up at the end of the build |
| 13:33.26 | ``Erik | ---nRun 'make test' to run the BRL-CAD Test SuitenRun 'make benchmark' to run the BRL-CAD Benchmark Suitenn**********************************************************n BRL-CAD 7.17.0 is now installed into /usr/brlcad/HEADn Be sure to add /usr/brlcad/HEAD/bin to your PATHn**********************************************************nnmake[2]: Entering directory `/var/tmp/erikg/brlcadbuild' |
| 13:33.51 | starseeker | looks like this may be related: http://www.mail-archive.com/bug-autoconf@gnu.org/msg02903.html |
| 13:35.14 | starseeker | hmm: http://www.mail-archive.com/bug-autoconf@gnu.org/msg02911.html |
| 13:38.15 | starseeker | yeah, this is not a BRL-CAD specific issue from the looks of it - recent libtool changed something |
| 13:39.40 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r2333 10/wiki/Google_Code_In/Project_Ideas: SCLstring -> std::string |
| 13:40.22 | ``Erik | from 02912: The bug is not in 'echo', but in the improper use of $(ECHO) within `` |
| 13:40.25 | ``Erik | inside the Makefile |
| 13:43.11 | starseeker | let me see if we can override it by force in configure.ac - there's a place where we check if $ECHO was defined at all |
| 13:43.27 | starseeker | perhaps a check there for this printf statement, and an override if it is present, would get it working |
| 13:44.52 | ``Erik | echo is posix (echo -n is NOT, however), where does $(ECHO) actually buy us anything? |
| 13:45.47 | starseeker | shrugs - dunno. I'm just looking for the most minimal change to get us past distcheck right now, removing $(ECHO) would be a fair bit of work |
| 13:46.26 | ``Erik | it would? O.o |
| 13:46.50 | starseeker | not in and of itself, but looking for what prompted us to make it a variable and confirm it's no longer an issue |
| 13:54.07 | starseeker | bah, overriding it does nothing |
| 13:54.23 | starseeker | ``Erik: OK, I was wrong - perhaps we do need to replace $(ECHO) |
| 13:55.19 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r2334 10/wiki/Google_Code_In/Project_Ideas: quell verbose warnings in src/util |
| 14:06.49 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r2335 10/wiki/Google_Code_In/Project_Ideas: clean up the 'analyze' command |
| 14:08.35 | CIA-48 | BRL-CAD: 03starseeker * r41252 10/brlcad/trunk/ (5 files in 5 dirs): Try replacing $(ECHO) with echo in the Makefile.am files - the newest libtool is using a printf expression for $(ECHO) that is resulting in extra n characters at the end of lines. |
| 14:11.23 | starseeker | Ubuntu is going to try Wayland instead of X.org???? |
| 14:14.42 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r2336 10/wiki/Google_Code_In/Project_Ideas: research impact of setting CPU affinity |
| 14:15.31 | starseeker | tries to figure out what Wayland is... |
| 14:15.34 | starseeker | ah |
| 14:17.29 | brlcad | the reason it's $(ECHO) at all is because there are/were platforms where "echo" is not found |
| 14:17.55 | brlcad | so you needed to rely on configure detecting the echo mechanism for that platform |
| 14:18.16 | starseeker | apparently modern libtool is going to gum that up but good |
| 14:19.40 | brlcad | so how to address both problems? |
| 14:19.51 | starseeker | CMake? :-P |
| 14:20.04 | starseeker | what platforms lack echo? |
| 14:21.11 | brlcad | you'd have to search the logs |
| 14:21.37 | starseeker | If they're still relevant today we've got a problem |
| 14:25.17 | brlcad | from that mailing list, it looks like it's recent libtool that's causing the problem and they're going to fix it |
| 14:27.14 | starseeker | that was back in august though |
| 14:29.38 | starseeker | earliest item I see in the first cut is r23717 |
| 14:33.21 | _psilva_ | hm Vala looks interesting |
| 14:38.02 | starseeker | brlcad: for include/conf/Makefile.am the change to $(ECHO) came at r31964 |
| 14:40.03 | starseeker | for bench/Makefile.am it was r24110 |
| 14:40.18 | starseeker | I don't see a motivation in eather of those... |
| 14:42.03 | starseeker | in the case of the toplevel Makefile.am it was using @ECHO@ |
| 14:45.12 | starseeker | first appearance in db/Makefile.am was r30442 |
| 14:45.45 | starseeker | and step I'm sure inherited it from BRL-CAD's file |
| 14:46.40 | starseeker | does not appear to be mentioned in any toplevel file (TODO/NEWS/etc.) |
| 14:49.35 | brlcad | no matter then, you can just do the solaris build when they're back up and running ;) |
| 14:49.49 | starseeker | heh :-) |
| 14:49.59 | brlcad | they're probably the closest platform to causing any modern issue |
| 14:50.46 | starseeker | they're posix though, correct? |
| 14:52.58 | brlcad | traditionally, the strictest |
| 14:53.03 | brlcad | even where posix has flaws |
| 14:53.46 | brlcad | e.g., if posix just says echo has to be on the system somewhere, that doesn't mean it's in your path or is built-in or is available within Makefiles .. lots of variables in play |
| 14:55.01 | starseeker | ah |
| 14:56.22 | starseeker | well, fwiw, the gentoo build is progressing well |
| 14:56.32 | starseeker | ``Erik: did that get your Mac working? |
| 14:59.37 | starseeker | sweet - after the echo change, distcheck passes on gentoo |
| 15:00.54 | ``Erik | seems to all be good, lemme purge and reconfigure/compile |
| 15:01.15 | starseeker | anybody know of a test farm somewhere where we could set up a Solaris build? |
| 15:08.10 | starseeker | may have to poke at the GCC farm to see if they have something viable and would consider allowing us on there... |
| 15:08.23 | starseeker | or perhaps we could use virtualbox/qemu and set something up... |
| 15:19.05 | starseeker | alrightie, distcheck passes here, heading in |
| 15:23.20 | ``Erik | cliff |
| 15:38.38 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r2337 10/wiki/Google_Code_In/Project_Ideas: g-iges + iges-g changes |
| 15:40.54 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r2338 10/wiki/Google_Code_In/Project_Ideas: new website solicitation |
| 15:56.30 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r2339 10/wiki/Google_Code_In/Project_Ideas: lots of missing manual pages, provide an easy and hard task |
| 16:18.58 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r2340 10/wiki/Google_Code_In/Project_Ideas: command spreadsheets |
| 16:24.39 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r2341 10/wiki/Google_Code_In/Project_Ideas: reword for generality |
| 17:05.15 | *** join/#brlcad mafm (~mafm@193.153.52.142) |
| 17:29.39 | ``Erik | fresh configure/make on fbsd8 and osX.6 both succeed |
| 17:41.01 | starseeker | sweet |
| 17:47.54 | CIA-48 | BRL-CAD: 03brlcad * r41253 10/brlcad/trunk/HACKING: |
| 17:47.54 | CIA-48 | BRL-CAD: that first step should really also include notifying the brlcad-devel mailing |
| 17:47.54 | CIA-48 | BRL-CAD: list with a simple message letting others know that someone has begun release |
| 17:47.54 | CIA-48 | BRL-CAD: steps so that others know to keep the commits at ease for a couple days |
| 17:48.20 | starseeker | distcheck passed on Redhat |
| 17:53.28 | brlcad | starseeker: not to throw a wrench into the release, but there needs to be a strong change that distinguishes pushing out the minor or it should be a merged 7.16.12 off stable |
| 17:53.42 | brlcad | at least an api change (deprecations, removal of obsoletes) |
| 17:55.25 | brlcad | I'm thinking we can hit up the pre 7 and 7.12 deprecations quickly |
| 17:56.59 | starseeker | k, cool |
| 17:57.00 | brlcad | otherwise, we can declare archer in official alpha status if it's stable enough, but I didn't get a chance to talk to bob |
| 17:57.21 | starseeker | he's in today, feeling better - he should be in next week |
| 17:57.41 | brlcad | ask him what he thinks about an alpha archer |
| 17:58.06 | brlcad | presuming we all are on the same page as to what that means .. :) |
| 17:58.10 | starseeker | I'm willing to go the 16.12 route, but what does "merged 7.16.12 off of stable" mean? sync to stable without updating the version numbers? |
| 17:58.30 | starseeker | brlcad: first, define what it means to you :-) |
| 17:58.57 | brlcad | well, head is already 7.17 so it technically shouldn't regress back to 7.16 but stable is still at 7.16 so it can move forward a minor and get tagged from there |
| 17:59.53 | starseeker | we probably can't declare it alpha without turning opengl in the default builds |
| 17:59.55 | brlcad | not a big deal with 7.17 since it's still dev, it could be committed briefly as 7.16.12 for release purpose then jumped back (to 7.17.1) |
| 18:00.23 | starseeker | nods - that might be a way to go |
| 18:01.47 | brlcad | alpha means we're including it enabled in all binary installs now and ACTIVELY soliciting users to check it out, provide feedback, and that we're otherwise done adding features from our end |
| 18:02.19 | starseeker | ah, we are not done adding features |
| 18:02.22 | brlcad | it's not quite a feature-lock yet, but new features should be mostly in response to user feedback |
| 18:02.24 | starseeker | so the answer is a definite now |
| 18:02.26 | starseeker | er no |
| 18:03.17 | brlcad | then beta is then when it goes into lock-down and the only new features are in response to bugs or problems (e.g. usability) |
| 18:03.28 | starseeker | sketch editing, pipe editing, finishing comb editing... |
| 18:04.05 | starseeker | nods - yeah, we're a ways away then |
| 18:04.10 | brlcad | finishing /refining features is okay during alpha |
| 18:04.25 | brlcad | adding new/missing ones usually isn't |
| 18:05.08 | starseeker | we might (generously) call comb editing a finishing/refining, but not sketch and pipe - they're not present at all ATM |
| 18:05.52 | brlcad | so then the question is whether they are release critical |
| 18:06.06 | starseeker | sketch probably not, pipe I would say yes |
| 18:06.23 | brlcad | what about via libged? |
| 18:06.31 | brlcad | command-line editing, is that working for them? |
| 18:07.04 | starseeker | bob says "in theory yeah, using the adjust command - may not be practical for big pipes" |
| 18:07.17 | brlcad | k |
| 18:08.00 | brlcad | so not quite ready, and release is then either API deprecations/removals or a minor release |
| 18:08.56 | brlcad | searchable command manual pages in mged would be a good minor justification too |
| 18:09.44 | starseeker | not sure we can whip that off in a day or two - what searching mechanism did you have in mind? |
| 18:09.50 | brlcad | keyword |
| 18:09.53 | starseeker | hunts up the deprecations file |
| 18:10.17 | starseeker | I mean, what mechanism to look through the files? grep? a tcl regex match? etc. |
| 18:10.27 | brlcad | *shrug* |
| 18:11.38 | starseeker | bob says he might be able to have someing servicable for pipe in a week or so |
| 18:11.51 | brlcad | ideally like pdf searching in Preview |
| 18:12.45 | brlcad | type some word(s) and get a list of the matching line with surrounding context in a summary list with the word bolded that jumps you to that line on that page |
| 18:13.05 | starseeker | oh, I see - yeah, that's not a quickie |
| 18:13.40 | brlcad | so it'd be a straight-up grep, but the good usability aspect is the results while you type, the context, and jumping you to the match |
| 18:15.09 | starseeker | right - sounds good, just a bunch of details like the mechanics of highlighting words in tkhtml would take a bit of time (for me at least :-/) |
| 18:15.57 | brlcad | that part of the search pane wouldn't need to be tkhtml |
| 18:16.09 | brlcad | you'd just need snippets of text and a list view |
| 18:18.36 | starseeker | It's probably not hard, but I still wouldn't bet on pulling it together prior to release starting cold |
| 18:23.20 | starseeker | I'd say rather than force it, let's go for 7.16.12 and then take steps to make sure the next one is 7.18 |
| 18:23.28 | brlcad | probably less than a day's work for sure, but it's fine either way -- clearing out the deprecations will take a few hours too |
| 18:23.52 | brlcad | that's fine |
| 18:24.05 | starseeker | are those the ones in raytrace.h and whatnot? |
| 18:24.15 | brlcad | the doc file |
| 18:24.27 | brlcad | the 7.12's and the pre7's |
| 18:24.34 | starseeker | oh, I see them |
| 18:24.46 | starseeker | so just go through and make sure our code doesn't use them? |
| 18:25.12 | brlcad | and ideally deprecating anything else we see (particularly libbu/libbn/libwdb/librt) .. should give notice about libnmg |
| 18:25.19 | brlcad | right |
| 18:25.39 | brlcad | make sure we don't use them, and remove their decls from headers or the headers |
| 18:25.39 | starseeker | libnmg... oh, you mean that it'll be moving back to its own lib? |
| 18:25.58 | brlcad | right -- technically, that's a body of librt API that's being removed from librt |
| 18:26.10 | *** join/#brlcad stevegt_ (~stevegt@cislunar.TerraLuna.Org) |
| 18:26.12 | brlcad | minor, but worth documenting |
| 18:26.33 | starseeker | yow. how much detail do we need to document for that one? full list of functions and vars that will be going away? |
| 18:27.13 | brlcad | no, just categorically say the nmg_* functions/structures/defines or "the nmg code" or similar |
| 18:27.24 | starseeker | (aside) with distcheck passing now on multiple platforms, I'm gonna go ahead and try a stable sync |
| 18:27.34 | starseeker | cool |
| 18:27.40 | brlcad | and then marking them in the headers |
| 18:27.51 | brlcad | it's 500+ functions, so no sense listing them all |
| 18:27.56 | starseeker | so that'll be 7.20 or 7.22? |
| 18:29.28 | brlcad | deprecations announced in 7.16.12 can be moved in 7.20 |
| 18:29.41 | brlcad | deprecations in 7.18 are 8.22 |
| 18:29.44 | brlcad | er, 7.22 ;) |
| 18:30.01 | starseeker | crosses fingers, toes, etc. and starts the merge command... |
| 18:30.22 | brlcad | that's why it's more important to mark deprecated than it is to remove for obsoletion |
| 18:30.32 | starseeker | brlcad: now you're giving us an incentive to make this 7.16.12 :-P |
| 18:30.33 | brlcad | obsoletion just becomes a good justification for minor |
| 18:31.02 | brlcad | yesh -- deprecation pushes patch, obsolete pushes minor |
| 18:31.31 | starseeker | is dizzy |
| 18:31.40 | brlcad | though libnmg is arguably "minimally impacting" |
| 18:32.04 | starseeker | I doubt many people call nmg functions direct... |
| 18:32.05 | brlcad | so it could happen at any time since it's a regex fix to a build file s/-lrt/-lrt -lnmg/ |
| 18:32.33 | starseeker | oh, gotcha - make into libnmg, but preserve API |
| 18:32.47 | brlcad | right, if the api doesn't change, it's still available |
| 18:33.05 | starseeker | was under the impression the API could stand a little TLC... |
| 18:33.16 | brlcad | doesn't matter if we think *nobody* uses it -- it matters if we "publicly published" in some manner |
| 18:33.41 | starseeker | true |
| 18:33.58 | starseeker | yeah, I think that counts as minor then |
| 18:34.00 | brlcad | every API could use some TLC but it's actually one of the more consistent ones .. just very terse |
| 18:34.38 | brlcad | it counts as a minor, but it may still be minimally impacting .. have to see what all is impacted when it's moved |
| 18:34.56 | brlcad | the move can still happen and it get bundled into librt as a LIBADD |
| 18:35.08 | brlcad | then removed in 7.20 |
| 18:35.20 | starseeker | nods |
| 18:36.20 | starseeker | wonders if it might be better to do this sync in stages... |
| 19:48.22 | _psilva_ | we change APIs in alpha :p |
| 19:52.51 | CIA-48 | BRL-CAD: 03starseeker * r41254 10/brlcad/branches/cmake/src/other/tcl/CMake/CheckSystemFunctionality.cmake: |
| 19:52.51 | CIA-48 | BRL-CAD: Start trying to make a more generic version of the CMake functionality for |
| 19:52.51 | CIA-48 | BRL-CAD: checking things. This one will hopefully cover both the config.h case and the |
| 19:52.51 | CIA-48 | BRL-CAD: CFLAGS case, although a lot of rewiring is needed to test it properly. |
| 20:04.39 | brlcad | nothing wrong with an API changing in alpha, it's supposed to change, even in beta I'd expect maybe some changes |
| 20:08.35 | CIA-48 | BRL-CAD: 03starseeker * r41255 10/brlcad/branches/cmake/src/other/tcl/ (CMake/tcl.cmake CMakeLists.txt): Shift tcl build to using changed macros. |
| 20:19.02 | CIA-48 | BRL-CAD: 03starseeker * r41256 10/brlcad/branches/cmake/src/other/tk/library/CMakeLists.txt: As with tcl, put the tk scripts where they should be for a local run (hopefully) |
| 21:08.33 | CIA-48 | BRL-CAD: 03starseeker * r41257 10/brlcad/branches/cmake/src/other/incrTcl/itcl/ (4 files in 3 dirs): Get closer to a proper itcl build - not there yet, need to compare with a standard itcl build and install. |
| 21:13.50 | *** join/#brlcad packrat (~packrator@99-67-225-40.lightspeed.livnmi.sbcglobal.net) |
| 21:54.49 | CIA-48 | BRL-CAD: 03starseeker * r41258 10/brlcad/branches/cmake/src/other/ (incrTcl/itcl/CMakeLists.txt tcl/CMakeLists.txt): Fix a couple bugs, add a little extra foo to make itcl locatable in the build dir. |
| 22:46.46 | CIA-48 | BRL-CAD: 03starseeker * r41259 10/brlcad/branches/cmake/src/bwish/main.c: BWISH is not cooperating - tweak it some, but so far it only starts if run from exactly the build dir. |
| 23:02.26 | starseeker | meh - tclsh runs successfully now, but btclsh doesn't (at least, not in the build) |
| 23:03.56 | starseeker | drills into tclsh to see why it's succeeding when we're failing... |
| 23:10.24 | *** join/#brlcad Ralith (~ralith@S010600221561996a.vc.shawcable.net) |
| 23:26.55 | starseeker | hmm, tcl_library isn't set right |
| 23:31.01 | starseeker | checks to make sure tclsh works after install... |
| 23:34.08 | starseeker | yep, that's it - somehow, tcl_library in the CMake btclsh build is getting set to a (wrong) relative path |
| 23:34.40 | starseeker | will check the autotools build for foo he missed when doing the initial setup |