| 00:12.56 | *** join/#brlcad akafubu (~akafubu@unaffiliated/akafubu) |
| 02:06.58 | CIA-73 | BRL-CAD: 0368.37.119.2 07http://brlcad.org * r2232 10/wiki/MGED_CMD_knob: put the arguments in a table instead of a list; remove superfluous whitespace; put the program name in code tags |
| 02:26.41 | starseeker | is there any good way to get libtool to cough up the name of the actual .so or .dylib library file name? |
| 02:28.15 | juub | stat, maybe? |
| 02:29.05 | juub | nah, probably not what you need |
| 02:29.27 | juub | well, maybe... |
| 02:29.42 | juub | on my machine: $ stat /lib/libc.so.6 File: `/lib/libc.so.6' -> `libc-2.10.1.so |
| 02:29.55 | juub | for example... |
| 02:30.18 | juub | n/m, sorry |
| 02:30.28 | juub | completely misread what you were after. |
| 02:37.19 | *** join/#brlcad IriX64 (~mario_dul@bas2-sudbury98-1177680356.dsl.bell.ca) |
| 02:44.25 | *** part/#brlcad IriX64 (~mario_dul@bas2-sudbury98-1177680356.dsl.bell.ca) |
| 03:07.39 | starseeker | brlcad: this is annoying - to provide something to a tcl load command that can actually load, I have to provide the .so or .dylib or whatever filepath |
| 03:08.18 | starseeker | but that's precisely what libtool doesn't want to tell me |
| 03:08.52 | starseeker | is there something like "libtool --realname librt.la" that will return librt.so or librt.dylib as the case may be? |
| 03:09.51 | starseeker | the TEA stuff would give me that but we don't want to suck that into our main build |
| 03:12.20 | ``Erik | grep ^dlname file.la | cut -d \' -f 2 |
| 03:12.21 | ``Erik | :D |
| 03:15.39 | starseeker | can live with that |
| 03:15.49 | starseeker | see if it works in the Makefile.am... |
| 03:16.25 | ``Erik | could make a little shell script to source the .la and print the variable, if you don't trust the system to have grep and cut |
| 03:17.25 | ``Erik | (all the .la should have is posix shell compliant variable setting) |
| 03:18.11 | starseeker | that would be great if you could |
| 03:18.20 | starseeker | apparently has week sed foo |
| 03:18.25 | ``Erik | heh, I meant YOU could :D |
| 03:18.29 | ``Erik | it'd be something like, uh |
| 03:18.29 | starseeker | or sh foo, whatever... |
| 03:18.32 | ``Erik | #!/bin/sh |
| 03:18.35 | ``Erik | source $1 |
| 03:18.37 | ``Erik | echo $dlname |
| 03:18.41 | ``Erik | or so |
| 03:19.30 | ``Erik | yup, that works |
| 03:20.42 | starseeker | hah - cool! |
| 03:20.49 | starseeker | thanks ``Erik |
| 03:20.56 | ``Erik | np |
| 03:22.15 | starseeker | ponders if that should be stuck in misc or can function in some way in the Makefile.am itself... |
| 03:22.54 | ``Erik | libtool might have a 'print variable' feature *shrug* |
| 03:23.24 | starseeker | was trying to find that |
| 03:23.45 | starseeker | gets the sense that the whole point of using libtool is that you never care about the suffix |
| 03:23.58 | starseeker | which, of course, is diametrically opposed to how tcl views the world |
| 03:24.16 | ``Erik | yeah |
| 03:24.33 | starseeker | heck with it - I'll be silly/stupid if it works... |
| 03:28.08 | ``Erik | I can't think of a good way to make it as an inline in Makefile.am, you don't want to pollute an existing shell while building, and anything in a makefile would be a convoluted attempt to fake a text stream for a subshell, like sh <<EOF style |
| 03:29.09 | starseeker | nods |
| 03:29.11 | ``Erik | (and of course using a shell script is gonna make things weird for msvc and mebbe xcode) |
| 03:29.35 | starseeker | well, we need a different solution for msvc anyway |
| 03:29.41 | ``Erik | lithp/cffi seems to have a similar issue, btw |
| 03:29.46 | starseeker | and I dunno if we support xcode or not... |
| 03:29.50 | starseeker | nods |
| 03:30.23 | ``Erik | <-- been poking at it for okra and mebbe writing an adrt cffi O.o |
| 03:30.27 | starseeker | it's not like TEA is any better in that regard |
| 03:31.19 | ``Erik | mwahahaha, src/adrt/adrt.asd :D |
| 03:43.07 | *** join/#brlcad ibot (~ibot@rikers.org) |
| 03:43.07 | *** 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 is now available on Gentoo! (20100225) |
| 03:44.57 | *** join/#brlcad 50UAAP4VJ (~stevegt@c-69-181-134-76.hsd1.ca.comcast.net) |
| 04:44.29 | CIA-73 | BRL-CAD: 03starseeker * r39164 10/brlcad/trunk/sh/ (Makefile.am libtoolfilename.sh): Add trivial sh script to report actual library file associated with a libtool archive (thanks Erik) |
| 04:47.17 | CIA-73 | BRL-CAD: 03starseeker * r39165 10/brlcad/trunk/sh/libtoolfilename.sh: Whoops - remove pwd, that's not supposed to be there. |
| 04:49.38 | CIA-73 | BRL-CAD: 03starseeker * r39166 10/brlcad/trunk/doc/docbook/Makefile.am: Fix docbook makefile typo. |
| 05:05.38 | CIA-73 | BRL-CAD: 03starseeker * r39167 10/brlcad/trunk/configure.ac: Add in RENDER and TIE variable definitions for the src/adrt libs |
| 05:08.38 | starseeker | should sleep now... |
| 05:12.08 | *** join/#brlcad Nohla (~jesica@201.255.237.95) |
| 09:00.34 | *** join/#brlcad cosurgi (~cosurgi@atak.bl.pg.gda.pl) |
| 09:48.18 | *** join/#brlcad Nohla (~jesica@201.255.251.127) |
| 10:14.43 | *** join/#brlcad Nohla (~jesica@201.255.251.127) |
| 11:20.24 | d-lo | Mernin! |
| 11:35.06 | brlcad | howdy |
| 11:37.17 | brlcad | starseeker: repeat the command but remove the --silent options |
| 11:37.33 | brlcad | it will echo the actual compile/link commands used after the libtool line |
| 11:41.27 | CIA-73 | BRL-CAD: 03brlcad * r39168 10/brlcad/trunk/autogen.sh: s/considered/consider/ typo noticed by bullet_catcher |
| 11:46.44 | *** join/#brlcad louipc (~louipc@archlinux/trusteduser/louipc) |
| 11:48.10 | louipc | mornin |
| 12:01.36 | brlcad | starseeker: the libtool way for loadable modules is that those libraries get declared as modules in the Makefile.am, then you get to specify exactly what name you want/need -- but you don't do that for all libraries, only loadable ones and the idea is that your loadable libs are different from the libs used to link programs |
| 12:02.18 | brlcad | i.e., librt would not be a "plugin" or module, but erik's dlopen bits for adrt definitely would be |
| 12:39.20 | starseeker | nods - librt.la was just the first libtool archive I thought of for a test case |
| 12:50.11 | ``Erik | hm, but how do ya explain to libtool that "no, I don't want .so or .dylib or .dll, I really DO want .plugin"? |
| 12:55.30 | brlcad | that's the module method |
| 12:55.35 | brlcad | it lets you name it however you want |
| 13:08.14 | brlcad | yeah, -shrext plugin |
| 13:16.09 | brlcad | the 1.5 libtool on Mac is busted wrt shrext, but newer versions should work fine |
| 13:16.29 | brlcad | there's workarounds for that too |
| 13:47.15 | *** join/#brlcad stevegt_1 (~stevegt@c-69-181-134-76.hsd1.ca.comcast.net) |
| 13:47.57 | ``Erik | hm, since my macs are using 2.2.6b, nifty! |
| 13:48.24 | ``Erik | once I have my home directory back, I'll have to adjust the demo plugin in isst |
| 13:50.15 | starseeker | had created a test Makefile.am last night, and now finds he accidently erased it and saved the Makefile.in |
| 13:50.19 | starseeker | kicks himself |
| 13:50.38 | starseeker | and starts tring to do it the "right way"... maybe it's just as well |
| 14:01.17 | brlcad | starseeker: most of the Makefile.am decls will be in the .in file |
| 14:01.32 | starseeker | nods |
| 14:01.34 | brlcad | blocked together in groups |
| 14:01.58 | starseeker | well, I need to try the module/plugin thing anyway |
| 14:02.38 | brlcad | ``Erik: note that we still require support of 1.5 on the brlcad module, so the work-around would have to go in there, but it's like two lines |
| 14:04.34 | starseeker | brlcad: do we have it in the tree anywhere? |
| 14:04.37 | ``Erik | if it goes into brlcad/brlcad/trunk... :D brlcad/isst/trunk already diverges from that paradigm |
| 14:04.57 | starseeker | ``Erik: so I'm noticing (shakes fist) |
| 14:06.16 | ``Erik | heh, something doesn't work with ancient versions of libtool already? I meant more like the src/other inclusion stuff |
| 14:24.31 | starseeker | ``Erik: no, I mean isst build logic diverges more generally |
| 15:02.47 | CIA-73 | BRL-CAD: 03davidloman * r39169 10/rt^3/trunk/ (23 files in 2 dirs): Fixed compile errors that stemmed from not cascading 'const' ness down from NetMsg (when I implemented operator==) |
| 15:04.21 | CIA-73 | BRL-CAD: 03davidloman * r39170 10/rt^3/trunk/src/libNetwork/ (GenericOneStringMsg.cxx NetMsg.cxx TypeOnlyMsg.cxx): Removed debugging statements from NetMsg hierarchy. No longer needed. |
| 15:08.13 | *** join/#brlcad akafubu (~akafubu@unaffiliated/akafubu) |
| 15:13.08 | CIA-73 | BRL-CAD: 03davidloman * r39171 10/rt^3/trunk/ (include/NetMsgTypes.h src/libNetwork/NetMsgFactory.cxx): Add NetMsgTypes for the generic NetMsg classes and wire them into the NetMsgFactory. All to support testing. |
| 15:47.03 | CIA-73 | BRL-CAD: 03davidloman * r39172 10/rt^3/trunk/src/libNetwork/FailureMsg.cxx: Testing picked up a MsgType error in FailureMsg. Fixed. |
| 15:59.10 | CIA-73 | BRL-CAD: 03davidloman * r39173 10/rt^3/trunk/src/libNetwork/GeometryManifestMsg.cxx: Implemented GeometryManifest::_equals(...) to (initially) support testing. |
| 16:14.49 | CIA-73 | BRL-CAD: 03davidloman * r39174 10/rt^3/trunk/tests/libNetwork/CMakeLists.txt: Drop the job library from the include list for this network test. Not used. |
| 16:16.41 | CIA-73 | BRL-CAD: 03davidloman * r39175 10/rt^3/trunk/tests/libNetwork/netMsgSerialTest.cxx: Reworked and cleaned up the NetMsg serialization test to verify the operation of the NetMsg subclasses. Currently tests for equality only. Later work should include verification of inequality. |
| 16:18.02 | CIA-73 | BRL-CAD: 03davidloman * r39176 10/rt^3/trunk/tests/libNetwork/netMsgFactoryTest.cxx: Drop netMsgFactoryTest. NetMsgFactory is now fully tested via netMsgSerialTest. |
| 16:33.34 | brlcad | starseeker: do we have what in tree? |
| 16:33.54 | brlcad | and I meant brlcad/brlcad vs brlcad/isst .. the latter can do it's own thing |
| 16:34.05 | brlcad | but to bring into the prior will need the workaround |
| 16:40.57 | CIA-73 | BRL-CAD: 03davidloman * r39177 10/rt^3/trunk/ (include/Gateway.h src/libNetwork/Gateway.cxx): Implement Gateway::stopListening() to allow a Gateway to close its listening socket while maintaining existing connections. Implemented to support Gateway restarting. |
| 16:43.53 | CIA-73 | BRL-CAD: 03davidloman * r39178 10/rt^3/trunk/tests/libNetwork/PrintingMsgHandler.h: Implement PrintingMsgHandler. PrintingMsgHandler implements the interface INetMsgHandler. Provides simple printing to console functionality for any/all NetMsgs passed to this handler. To be used for Gateway testing. |
| 16:46.30 | CIA-73 | BRL-CAD: 03davidloman * r39179 10/rt^3/trunk/tests/libNetwork/ (. CMakeLists.txt gatewayTest.cxx netPortalManagerTest.cxx): Rename 'netPortalManagerTest' to more appropriate 'gatewayTest'. Updated CMakeLists.txt accordingly. |
| 17:13.11 | CIA-73 | BRL-CAD: 03davidloman * r39180 10/rt^3/trunk/tests/libNetwork/gatewayTest.cxx: (log message trimmed) |
| 17:13.11 | CIA-73 | BRL-CAD: Update to GatewayTest. Looks as though QT's signals and slots can only operate |
| 17:13.11 | CIA-73 | BRL-CAD: from an event loop launched from a QApplication or QCoreApplication, NOT a |
| 17:13.11 | CIA-73 | BRL-CAD: QThread's event loop. Furthermore, a QApplication/QCoreApplication event loop |
| 17:13.11 | CIA-73 | BRL-CAD: can only be launched from an executable's main thread. Since |
| 17:13.12 | CIA-73 | BRL-CAD: QCoreApplication.exec() blocks, there can only be one event loop at any given |
| 17:13.13 | CIA-73 | BRL-CAD: time. Thus, my attempts to contain QT-ness to the lower levels of libNetwork |
| 17:14.47 | d-lo | oh? CIA has a character limit? |
| 17:17.18 | CIA-73 | BRL-CAD: 03davidloman * r39181 10/rt^3/trunk/ (4 files in 2 dirs): WS, Formatting. |
| 17:28.43 | starseeker | brlcad: an example of the libtool 1.5 workaround |
| 17:37.55 | starseeker | d-lo: it trims long messages, yeah |
| 17:39.35 | ``Erik | wraps |
| 17:39.45 | ``Erik | oh, trims, too |
| 17:41.56 | ``Erik | RENDER and TIE aren't actually used yet? |
| 17:42.32 | starseeker | not in the trunk - I'm using them in my test code |
| 17:46.02 | *** join/#brlcad Stattrav (~Stattrav@117.96.29.161) |
| 17:52.21 | *** join/#brlcad louipc (~louipc@archlinux/trusteduser/louipc) |
| 17:54.07 | CIA-73 | BRL-CAD: 03erikgreenwald * r39182 10/isst/trunk/sdl/ (event.c isst.h): store time delta and framerate in isst struct. fix resolution sensitivity issue |
| 18:05.47 | starseeker | scowls at bsd |
| 18:19.37 | CIA-73 | BRL-CAD: 03erikgreenwald * r39183 10/isst/trunk/sdl/ (event.c isst.h main.c): chintzy fold-away stuff (should probably use glFrustum and actual angles or something... the texture doesn't map quite right) |
| 18:41.18 | CIA-73 | BRL-CAD: 03davidloman * r39184 10/rt^3/trunk/ (8 files in 3 dirs): |
| 18:41.18 | CIA-73 | BRL-CAD: Attempt at using QCoreApplication to spin off a QThread that handles all generic |
| 18:41.18 | CIA-73 | BRL-CAD: executable needs did not work. QT's automatic Parent/Child relationships does |
| 18:41.18 | CIA-73 | BRL-CAD: not allow for a Parent to be in a different thread than a child. gatewayTest |
| 18:41.18 | CIA-73 | BRL-CAD: currently compiles and runs, but is not functional. |
| 19:06.51 | brlcad | starseeker: if RENDER and TIE aren't used, there's no point in setting and subst'ing them |
| 19:07.11 | starseeker | brlcad: the plan is to use them soon |
| 19:07.30 | starseeker | I can take them out until I'm actually using them if it's an issue |
| 19:08.45 | brlcad | nah, just noting |
| 19:09.11 | starseeker | I can commit now, but there's no actual CAD functionality in place yet |
| 19:09.24 | brlcad | there's a script that kicks out a report of zero-referenced AC_SUBST variables .. it came up |
| 19:09.55 | CIA-73 | BRL-CAD: 03bob1961 * r39185 10/brlcad/trunk/src/libfb/fbserv_obj.c: fbs_open was not returning the correct port. |
| 19:10.02 | brlcad | eek |
| 19:12.46 | brlcad | how was that bug exposed? |
| 19:13.01 | brlcad | mged and archer? just archer? |
| 19:13.20 | starseeker | Bob found it |
| 19:13.34 | starseeker | showed in customer code, then archer |
| 19:22.32 | CIA-73 | BRL-CAD: 03erikgreenwald * r39186 10/isst/trunk/sdl/ (event.c main.c): Add command entry mode. Shrink display to corner instead of folding it back. |
| 19:34.05 | brlcad | how was is exposed? |
| 19:35.52 | starseeker | dunno - Bob has details |
| 19:37.34 | brlcad | more looking for succinct summary for NEWS |
| 19:41.26 | *** join/#brlcad PrezKennedy (~Prez@96.31.84.96) |
| 19:44.51 | CIA-73 | BRL-CAD: 03erikgreenwald * r39187 10/isst/trunk/sdl/event.c: fix sticky key issue. fix framerate display when capping. |
| 19:46.49 | CIA-73 | BRL-CAD: 03starseeker * r39188 10/brlcad/trunk/ (80 files in 9 dirs): |
| 19:46.49 | CIA-73 | BRL-CAD: Start roughing out a way to get isst and tk working together. Using togl widget |
| 19:46.49 | CIA-73 | BRL-CAD: and gears example as a starting point - the following allows 'bwish gears.tcl' |
| 19:46.49 | CIA-73 | BRL-CAD: to work succesfully, and links in libtie and librender (though it doesn't do |
| 19:46.49 | CIA-73 | BRL-CAD: anything with them yet. From here, convert from the gear demo code to the ISST |
| 19:46.49 | CIA-73 | BRL-CAD: opengl code as seen in the isst sdl based gui. |
| 19:52.29 | CIA-73 | BRL-CAD: 03Velociostrich 07http://brlcad.org * r2233 10/wiki/BRL-CAD_Commands: created the "rendering" subsection and added rtwizard and rtedge to it; arranged entries alphabetically; fixed spelling typo |
| 19:56.54 | *** join/#brlcad velociostrich (~nsd@c-68-37-119-2.hsd1.nj.comcast.net) |
| 20:04.58 | CIA-73 | BRL-CAD: 03Velociostrich 07http://brlcad.org * r2234 10/wiki/Developer_Documents: changed irc server (I tried irc.brlcad.org, but it didn't work; the hacking file says freenode) |
| 20:08.36 | CIA-73 | BRL-CAD: 03Velociostrich 07http://brlcad.org * r2235 10/wiki/Hex: expanded installation instructions a wee bit |
| 20:19.18 | CIA-73 | BRL-CAD: 03erikgreenwald * r39189 10/brlcad/trunk/src/other/togl/Makefile.in: add missing Makefile.in |
| 20:31.15 | *** join/#brlcad stevegt_ (~stevegt@cislunar.TerraLuna.Org) |
| 20:38.40 | CIA-73 | BRL-CAD: 03bob1961 * r39190 10/brlcad/trunk/src/tclscripts/lib/Ged.tcl: Added an alias for opendb. |
| 20:46.29 | ``Erik | hm, the package require for Togl doesn't seem to work in that gears.tcl for me |
| 20:46.43 | ``Erik | on fbsd... and togl doesn't compile at all on my mac O.o |
| 21:27.55 | CIA-73 | BRL-CAD: 03bob1961 * r39191 10/brlcad/trunk/misc/win32-msvc8/ (38 files in 38 dirs): It looks like fb.h now includes tk.h, so added include paths for tk. |
| 21:30.02 | CIA-73 | BRL-CAD: 03bob1961 * r39192 10/brlcad/trunk/ (include/fb.h src/mged/fbserv.c src/rt/viewarea.c): More tweaks for compiling on Windows. |
| 21:44.02 | CIA-73 | BRL-CAD: 03bob1961 * r39193 10/brlcad/trunk/src/tclscripts/mged/bindings.tcl: If not on a Mac, include the focus command in the default mouse button bindings. |
| 21:46.29 | starseeker | ``Erik: urm. It worked here |
| 21:46.41 | starseeker | what happens if you feed it --with-x11 ? |
| 21:49.07 | *** join/#brlcad Nohla (~jesica@201.255.250.157) |
| 21:54.07 | CIA-73 | BRL-CAD: 03brlcad * r39194 10/brlcad/trunk/src/libtclcad/ged_obj.c: collapse the case-switched dm types calling DM_close_existing() instead just calling the general fb_close_existing() routine. reduction of complexity, the DM types shouldn't be public API. |
| 21:54.39 | CIA-73 | BRL-CAD: 03brlcad * r39195 10/brlcad/trunk/include/fb.h: there shouldn't be any code outside of libdm (outside of fb_generic.c) that calls wgl_close_existing(). |
| 21:55.57 | brlcad | s/DM/FB/ |
| 22:03.20 | CIA-73 | BRL-CAD: 03brlcad * r39196 10/brlcad/trunk/NEWS: |
| 22:03.20 | CIA-73 | BRL-CAD: bob fixed a bug in the framebuffer server object that prevented multiple |
| 22:03.20 | CIA-73 | BRL-CAD: framebuffers from running concurrently. additional framebuffer servers would |
| 22:03.20 | CIA-73 | BRL-CAD: store the wrong port number (due to a refactoring consolidation bug (of |
| 22:03.20 | CIA-73 | BRL-CAD: unix+windows code)). |
| 22:15.00 | CIA-73 | BRL-CAD: 03r_weiss * r39197 10/brlcad/trunk/src/conv/obj-g_new.c: adding functions to fuse vertices outside of nmg creation functions, this is necessary for surface closure testing outside nmg functions |
| 22:37.57 | CIA-73 | BRL-CAD: 03brlcad * r39198 10/brlcad/trunk/ (include/raytrace.h src/librt/comb/db_comb.c): make db_tree_nleaves() return a size_t instead of int for proper typing. cascade of quellage changes to follow. |
| 22:54.41 | CIA-73 | BRL-CAD: 03starseeker * r39199 10/brlcad/trunk/ (configure.ac src/Makefile.am): Turn off the togl stuff by default - not nearly stable enough to have on yet. |
| 22:58.27 | ``Erik | heh |
| 22:59.46 | starseeker | ``Erik: you know what? I get the Cannot find file: ./toglStubInit.c on BSD only with gmake |
| 23:00.13 | starseeker | that's what happened I typed make early on, then switched to gmake - whoopsie |
| 23:02.05 | brlcad | got /bin/sh: line 17: cd: togl: No such file or directory |
| 23:02.26 | starseeker | at what stage? |
| 23:02.28 | brlcad | on out-of-dir build .. presumably line missing from configure.ac |
| 23:02.30 | brlcad | make |
| 23:02.46 | starseeker | confound it |
| 23:02.50 | brlcad | just updated with r29199 |
| 23:02.55 | brlcad | er 39199 |
| 23:03.03 | brlcad | so testing again |
| 23:09.48 | starseeker | what does the tcl/tk world have against out-of-dir builds? |
| 23:17.34 | CIA-73 | BRL-CAD: 03brlcad * r39200 10/brlcad/trunk/configure.ac: remove jove from the summary since it can be made obsolete on the next minor |
| 23:22.20 | CIA-73 | BRL-CAD: 03brlcad * r39201 10/brlcad/trunk/ (3 files in 3 dirs): |
| 23:22.20 | CIA-73 | BRL-CAD: initial stab at documenting deprecated old forms of attributes for region id, |
| 23:22.20 | CIA-73 | BRL-CAD: material id, aircode, color, and shader. intent is to make attributes |
| 23:22.20 | CIA-73 | BRL-CAD: case-insensitive yet case-preserving and to clarify what the reserved words |
| 23:22.20 | CIA-73 | BRL-CAD: actually are (possibly putting them into a cad:: namespace at some point) |
| 23:22.27 | brlcad | looks like the disable did the trick, though you should distcheck |
| 23:37.47 | CIA-73 | BRL-CAD: 03brlcad * r39202 10/brlcad/trunk/ (11 files in 6 dirs): make db_mkgift_tree() take a size_t, even though internally it needs a signed type the way it decrements in a loop past zero. this lets us cascade size_t's all over the place. |
| 23:43.44 | *** join/#brlcad dtidrow (~dtidrow@c-71-238-51-148.hsd1.mi.comcast.net) |