| 00:05.21 | starseeker | confound it |
| 00:05.37 | starseeker | tested the build on FreeBSD i386, and the off_t stuff is buggered |
| 00:06.29 | starseeker | not sure at the moment if it's all 32bit setups or just FreeBSD, but the build fails and tweaking it resulted in a failed regress-red |
| 00:10.32 | Notify | 03BRL-CAD:starseeker * 55504 brlcad/trunk/TODO: Add TODO item for 32bit issue with off_t on FreeBSD (at least) |
| 00:22.46 | starseeker | tries to figure out how to install gcc4 on haiku... |
| 00:44.10 | *** join/#brlcad zero_level (0e8b5206@gateway/web/freenode/ip.14.139.82.6) |
| 00:46.19 | zero_level | brlcad: I was studying about the feasibility and Hope the current load is much balanced. Need your suggestions here. |
| 00:46.53 | zero_level | the algorithm to balance the load was as follows. |
| 00:47.14 | zero_level | found the current implementation of those functions (in the tools) |
| 00:47.42 | zero_level | checked out the places required for changes. |
| 00:48.04 | zero_level | Also considered the numbers of functions (for documentation purpose) |
| 01:42.32 | starseeker | ah hah - setgcc x86 gcc4 |
| 01:44.50 | brlcad | starseeker: not at all surprising |
| 01:45.31 | starseeker | nods |
| 01:47.24 | brlcad | the approach taken is a rather buggered one |
| 01:48.27 | starseeker | looks like we'll have to straighten it out, after all... I'll know for sure once I get a 32 bit linux image of some sort set up |
| 01:48.43 | starseeker | (before release, I mean...) |
| 01:57.19 | brlcad | starseeker: any retest on the static analyzer? |
| 01:57.58 | starseeker | brlcad: not yet, sorry - computer resources being pretty well consumed by virtual machines at the moment :-/ |
| 01:57.59 | brlcad | exciting possibilities to see one in action based off the AST |
| 01:58.23 | starseeker | I do have it built though - I'll see if I can set up an overnight build |
| 01:58.47 | starseeker | one of the annoying parts is having to build src/other with the analyzer on too - Tcl/Tk is a bit... noisy |
| 01:59.08 | brlcad | you can't precompile src/other? |
| 01:59.37 | starseeker | not without mixing and matching compilers in the same CMake build - that's... a little tricky |
| 02:00.04 | brlcad | I was thinking more path trickery |
| 02:00.11 | starseeker | ah. possibly |
| 02:01.03 | brlcad | make a ccanal symlink to clang, compile src/other, then point to clang analyzer |
| 02:01.12 | brlcad | then finish the build |
| 02:01.47 | starseeker | probably worth isolating a test case based on our macroized bu_magic stuff and sending it to the analyzer devs as a bug report - would be handy to be able to spot cases where we genuinely had something wrong in that sort of logic |
| 02:02.03 | starseeker | brlcad: yeah, that would probably work |
| 02:02.50 | brlcad | it might have found a bonefide issue, but would have to see the output in detail |
| 02:03.35 | starseeker | remind me and I can show you next week, if I don't re-generate it here sooner |
| 02:03.38 | starseeker | bugger |
| 02:03.48 | starseeker | Haiku appears to have found an issue with non-X11 builds |
| 02:03.58 | starseeker | and another with M_LIBRARY |
| 02:06.48 | brlcad | nice to see you exploring the world of real operating system variability :) |
| 02:08.58 | starseeker | heh - new (powerful) toy that can run VirtualBox successfully == big trouble |
| 02:09.07 | brlcad | wants to give smart os a try |
| 02:09.15 | starseeker | ? |
| 02:09.38 | brlcad | http://smartos.org/ |
| 02:09.49 | brlcad | basically illumos |
| 02:10.06 | starseeker | ah - I grabbed an openindiana image, but haven't tried it yet |
| 02:10.52 | starseeker | looks like you can still download the sun studio compiler (depending on what all the legalize I haven't looked at yet says) in case we want to beat up on ourselves some more |
| 02:12.02 | starseeker | wonders what would happen with a plan9 build attempt... <shudder> |
| 02:13.48 | Notify | 03BRL-CAD:starseeker * 55505 brlcad/trunk/src/libdm/CMakeLists.txt: Wrap all the X11 libraries in the X11 conditional protection. |
| 02:15.54 | brlcad | i'm excited about it just because it's a major different kernel |
| 02:16.43 | brlcad | which usually means substantial differences in the underlying c standard lib, regardless of the compiler |
| 02:16.56 | brlcad | which is the best way to weed out assumptions |
| 02:17.47 | brlcad | zero_level: hi, sorry for the delay but not sure what you're telling me :) |
| 02:19.06 | Notify | 03BRL-CAD:starseeker * 55506 brlcad/trunk/src/libtclcad/CMakeLists.txt: Wrap X11 libs in feature check |
| 02:20.06 | brlcad | starseeker: if you can set up remote access, I'd take a stab at fixing the libregex off_t mess |
| 02:20.30 | brlcad | need access to a problematic system |
| 02:24.38 | zero_level | brlcad: think i couldnt write that well.. Anyways which part ? |
| 02:25.51 | starseeker | brlcad: urm - I only saw it on my virtualbox image |
| 02:26.14 | starseeker | brlcad: do you have virtualbox on your mac? Can just grab the FreeBSD i386 image and set it up... |
| 03:24.56 | *** join/#brlcad cstirk_ (~quassel@c-71-56-216-45.hsd1.co.comcast.net) |
| 03:28.19 | *** join/#brlcad IriX64 (~IriX64@70.49.197.250) |
| 03:49.14 | Notify | 03BRL-CAD:starseeker * 55507 (brlcad/trunk/src/other/CMakeLists.txt brlcad/trunk/src/other/libpng/CMakeLists.txt): Haiku doesn't need or support -lm for the math library - adjust build logic accordingly. Need to send the patch back to libpng. |
| 03:49.22 | Notify | 03BRL-CAD:brlcad * 55508 brlcad/trunk/NEWS: keith checked in his NURBS shaded display tessellation support in r55067. this does not yet include water-tight tessellation, but it is well-suited for visualization already. |
| 04:12.02 | *** join/#brlcad zero_level (0e8b5206@gateway/web/freenode/ip.14.139.82.6) |
| 04:17.36 | zero_level | brlcad:hi.. sry had to go smwh |
| 04:56.19 | starseeker | shamefacedly rebuilds his Haiku image with enough disk space to actually try compiling this time... |
| 05:13.48 | starseeker | OK, Haiku is data point #2 for the off_t issue |
| 05:15.40 | starseeker | winces - zlib failed on openindiana |
| 05:15.44 | brlcad | ~smwh |
| 05:16.21 | brlcad | zero_level: is everyone supposed to know what smwh means? |
| 05:16.45 | brlcad | seems rather obscure if you meant 'somewhere' |
| 05:17.33 | Notify | 03BRL-CAD:brlcad * 55509 brlcad/trunk/src/libbu/getopt.c: revert r54186 that intended to prevent confusion of negative numbers and number options. the change introduced a bug with parsing of numeric options that don't have a space after the option and multi-option strings (e.g., -s128 -8abc). the comment in the patch also incorrectly stated that options could only be a-zA-Z0-9 (punctuation is also allowed). need a |
| 05:17.35 | Notify | proper regression test case that we can compare against expected behavior from system getopt (if available) like we do for some of the other libbu unit tests. |
| 05:18.05 | brlcad | woot, I'll have to do one more pass, but that's the last commit-related issue I had noted for release |
| 05:19.09 | starseeker | smwh? |
| 05:19.52 | zero_level | brlcad: yes it meant somehwere :) |
| 05:21.28 | brlcad | zero_level: we're going to need to work on your communication skills a little bit buddy |
| 05:21.43 | brlcad | your message earlier is a train wreck :) |
| 05:22.04 | brlcad | punctuation is just misleading |
| 05:22.20 | brlcad | "random" use of case |
| 05:22.46 | brlcad | algorithm in following ... and no algorithm follows |
| 05:23.07 | zero_level | brlcad: I am naive on IRC i guess :) |
| 05:23.07 | brlcad | and now obscure shorthand |
| 05:23.24 | brlcad | this has absolutely nothing to do with IRC... |
| 05:23.27 | starseeker | blegh - openindiana only has gcc3 by default |
| 05:23.33 | starseeker | might explain a few things... |
| 05:24.09 | brlcad | it's how you're using text to convey information |
| 05:24.19 | brlcad | starseeker: we should still build on 3 |
| 05:24.53 | zero_level | may be on all text chats. Did you see my addition in the proposal ? |
| 05:25.07 | starseeker | brlcad: really? I thought even early versions of 4 were problematic these days... |
| 05:25.45 | brlcad | starseeker: I guess it depends on the issue -- it's certainly not been tested in a while, but I did do a gcc3 compile around 2 years ago |
| 05:25.49 | brlcad | took some cleanup, but worked |
| 05:25.57 | starseeker | nods |
| 05:26.25 | starseeker | well, barring some compiler installing pain, it looks like we're about to find out |
| 05:26.34 | brlcad | we might have adopted some mess with boost updates since then, but would need to see an error |
| 05:27.08 | starseeker | brlcad: the Haiku image might make a good potted example of the off_t bug, fwiw |
| 05:27.14 | brlcad | the bigger problem is more likely going to be the glibc that accompanies it |
| 05:28.07 | brlcad | we very well may be assuming libc features that earlier glibc wasn't so hot about |
| 05:29.56 | brlcad | zero_level: social chat is a completely different domain, this is international open source software development |
| 05:30.05 | brlcad | zero_level: you need to be prepared to be precise, complete, and clear in your communications |
| 05:30.30 | starseeker | brlcad: http://brlcad.org/~starseeker/openindiana_gcc3_build_errors.txt |
| 05:31.42 | starseeker | that's where I'm at as of this evening |
| 05:32.17 | starseeker | need some sleep - be back later |
| 05:33.20 | brlcad | starseeker: http://sourceforge.net/p/brlcad/discussion/362510/thread/ec2f5d8b |
| 05:33.44 | brlcad | interesting that he got past that same libz error just by re-running the autotools build |
| 05:36.18 | starseeker | ah - http://www.ilkda.com/compile/Build_Problems.htm - probably the linker |
| 05:37.15 | brlcad | ah, that would explain the autogen.sh fix |
| 05:38.04 | brlcad | means there's probably a flag that can be added/removed from the compile to get it to work |
| 05:38.19 | starseeker | doesn't have a lot of linker validation in the CMake build, IIRC - looks like I'll need to figure that out |
| 05:38.21 | brlcad | going to be some linker setting like -Wl,-E |
| 05:40.23 | starseeker | you can see the off_t errors there there too |
| 05:40.29 | brlcad | yeah |
| 05:40.35 | brlcad | I really expected that to happen |
| 05:41.08 | brlcad | we should not be defining standard types like that |
| 05:42.24 | brlcad | the only platform we do this on is windows and it's only because they have historically provided equivalent functionality, just with a different (underscored) name |
| 05:52.17 | Notify | 03BRL-CAD:brlcad * 55510 brlcad/trunk/NEWS: richard modified gqa in r52353 to report all missed regions, not just the first missed region (so users won't have to keep re-running gqa to get any objects further down the shotline. |
| 05:56.22 | zero_level | brlcad: hope i will learn in days ahead |
| 06:01.24 | *** join/#brlcad zero_level_ (0e8b5206@gateway/web/freenode/ip.14.139.82.6) |
| 06:57.36 | *** join/#brlcad jasleen (~chatzilla@202.164.53.118) |
| 07:03.41 | *** join/#brlcad jasleen (~chatzilla@202.164.53.117) |
| 10:31.59 | *** join/#brlcad vladbogo (~vlad@188.25.237.147) |
| 11:18.34 | *** join/#brlcad caen23 (~caen23@92.85.91.26) |
| 11:36.35 | *** join/#brlcad mpictor (~mark@99-93-104-202.lightspeed.iplsin.sbcglobal.net) |
| 12:21.08 | *** join/#brlcad kesha (~kesha@49.249.1.180) |
| 13:03.54 | *** join/#brlcad zero_level (0e8b5206@gateway/web/freenode/ip.14.139.82.6) |
| 13:20.57 | *** join/#brlcad caen23 (~caen23@92.81.183.7) |
| 14:45.48 | *** join/#brlcad phoenixyjll (8c71fd66@gateway/web/freenode/ip.140.113.253.102) |
| 15:11.29 | *** join/#brlcad d_rossberg (~rossberg@66-118-151-70.static.sagonet.net) |
| 15:23.58 | Notify | 03BRL-CAD:starseeker * 55511 brlcad/trunk/TODO: More details on where we do and don't see the off_t issue |
| 15:25.25 | *** join/#brlcad caen23 (~caen23@92.81.183.7) |
| 15:43.07 | *** join/#brlcad caen23_ (~caen23@92.83.182.2) |
| 16:49.43 | *** join/#brlcad kesha (~kesha@49.249.1.226) |
| 16:58.07 | *** join/#brlcad jasleen (~chatzilla@117.253.202.139) |
| 17:27.05 | *** join/#brlcad kesha (~kesha@49.249.1.226) |
| 18:06.26 | *** join/#brlcad caen23 (~caen23@92.83.185.228) |
| 18:43.09 | *** join/#brlcad kesha (~kesha@49.249.19.215) |
| 23:35.04 | *** join/#brlcad mpictor (~mark@99-93-104-202.lightspeed.iplsin.sbcglobal.net) |