| 00:09.37 | *** join/#brlcad akafubu (n=akafubu@unaffiliated/akafubu) |
| 00:17.55 | ``Erik | both were const, I was also seeing it in a recursive function where it called itself with the exact same parameter... |
| 00:18.03 | ``Erik | there's something fruity going on with my bsd box I think |
| 00:18.23 | ``Erik | but at the moment, I'm cooking dinner, so *shrug* I don't care :) |
| 00:21.02 | *** join/#brlcad cosurgi (n=cosurgi@atak.bl.pg.gda.pl) |
| 00:57.24 | ``Erik | <colbert> so in order to influence people, I have to appear in their facebook pictures? I don't think I can vomit in that many photos |
| 00:57.27 | ``Erik | *snrkt* |
| 01:32.23 | *** join/#brlcad R0b0t1 (n=Enigma@unaffiliated/r0b0t1) |
| 01:33.11 | *** join/#brlcad R0b0t1 (n=Enigma@unaffiliated/r0b0t1) |
| 01:49.47 | brlcad | ``Erik: both are const, it's const char * vs const char [6] .. should be promotable, but there must be some obscure reason gcc feels the need to bitch about it |
| 01:53.16 | ``Erik | hm, I thought I was seeing a couple cases of "const point_t p[6]" not maching "const point_t p[6]", and one case of "const point_t *p" not maching "const point_t *p" |
| 01:53.33 | ``Erik | (think it was point_t... not plane_t... hrm...) |
| 01:53.57 | ``Erik | still in the office? |
| 01:59.13 | Ralith | Anyone familiar with X input contexts? |
| 01:59.46 | brlcad | yeah, it was plane_t ``Erik |
| 02:00.57 | brlcad | maybe some spec issue where a char [] parameter is always passed by pointer address so you get a warning if the prototype has [] even if it's a [] to begin with |
| 02:01.01 | brlcad | dunno |
| 02:01.02 | brlcad | something obscure for sure |
| 02:01.18 | ``Erik | http://www.youtube.com/watch?v=4nigRT2KmCE |
| 02:01.41 | brlcad | Ralith: have to be more specific to answer that ;) |
| 02:01.42 | ``Erik | it's weird, did a -E and it all looked correct to me, saw it in metaball.c, too |
| 02:01.57 | ``Erik | I'm thinkin' I may've busted something on my mp3 player *shrug* :) |
| 02:02.23 | Ralith | brlcad: just trying to get basic internationalized text input working. |
| 02:02.31 | Ralith | Xutf8LookupString wants an XIC |
| 02:02.35 | Ralith | and I can't work out where to get one |
| 02:03.18 | Ralith | the X manpages are just short of completely useless. |
| 02:03.40 | ``Erik | you don't have the 10 volume X programmer manual on your shelf, ralith? :D |
| 02:03.48 | Ralith | :P |
| 02:04.00 | Ralith | wish I had a shelf that big. |
| 02:04.47 | ``Erik | why are you touching that directly instead of using a wrapper like sdl, qt, gtk, oid, ...? |
| 02:07.40 | Ralith | I'm writing a wrapper like ... ^^ |
| 02:11.55 | ``Erik | brlcad: still get the error, replicated on crit |
| 02:12.42 | ``Erik | metaball does it on a "const point_t *" |
| 02:13.24 | ``Erik | arb8 on "const plane_t *" |
| 02:14.26 | Ralith | ooh, I think I found a PDF of a relevant book. |
| 02:15.24 | ``Erik | removing the const seems to fix it |
| 02:17.34 | brlcad | Ralith: XCreateIC() |
| 02:18.34 | brlcad | ``Erik: *shrug* .. need to see the snippet -- I replicated one earlier and fixed it just by making it a const * param |
| 02:18.41 | brlcad | instead ofa [] param |
| 02:18.56 | brlcad | something with it being a sized array not being constable |
| 02:18.59 | Ralith | brlcad: that seems to need a boatload of parameters and I can't find a single doc on what they are. |
| 02:19.16 | brlcad | perhaps the ambiguity that the array itself can't be protected as const |
| 02:20.28 | ``Erik | perhaps |
| 02:21.46 | brlcad | Ralith left, boo hiss |
| 02:21.49 | ``Erik | http://pastebin.bzflag.bz/dfda925e |
| 02:22.00 | ``Erik | (this is withOUT strict flags, btw) |
| 02:22.05 | brlcad | k |
| 02:22.47 | brlcad | huh, are you up to date? |
| 02:22.53 | brlcad | because I changed arb8 |
| 02:22.56 | ``Erik | yes |
| 02:23.14 | *** join/#brlcad Ralith (n=ralith@69.90.48.97) |
| 02:23.40 | ``Erik | the arb8 refers to "const plane_t *p" instead of "const plane_t p[6]" |
| 02:23.54 | brlcad | Ralith: when in doubt.. |
| 02:23.59 | brlcad | look at someone else's code |
| 02:24.01 | brlcad | http://svn.tomasu.org/index.cgi/allegro/checkout/allegro/branches/4.9/src/x/xkeyboard.c?rev=12203 |
| 02:24.26 | brlcad | allegro uses it exactly as I had in mind |
| 02:25.15 | brlcad | gotta get used to the hell that is X11 programming.. verbose and with many layers, like an onion |
| 02:25.18 | brlcad | not a parfait |
| 02:25.34 | ``Erik | hehehe |
| 02:25.47 | ``Erik | X is the result of an API appearing on "will it blend" O:-) |
| 02:26.30 | brlcad | "hmm.. what ELSE can we abstract away in the API into it's own type??" |
| 02:27.09 | ``Erik | what was the um, "if windowing systems were a car" or something? X had a dozen steering wheels, and no two worked alike? |
| 02:30.47 | brlcad | yea |
| 02:31.12 | brlcad | and you can use them all simultaneously |
| 02:31.24 | brlcad | and in some cases are required to use multiple simultaneously |
| 02:31.31 | brlcad | okay, so this is bizzare |
| 02:31.56 | brlcad | plane_t *foo = someplane_array; /* no problem */ |
| 02:32.09 | brlcad | const plane_t *bar = foo; /* bitches * |
| 02:33.32 | brlcad | unrolling, plane_t is a typedef for a fastf_t[4], so somehow related |
| 02:33.39 | ``Erik | yup |
| 02:33.50 | ``Erik | and -E doesn't show anything fruity happening to const or anything |
| 02:34.14 | ``Erik | wonder if some change somewhere is setting something that's putting gcc in a weird mode |
| 02:35.16 | brlcad | const fastf_t *bar = foo[0]; works fine as I'd hope |
| 02:38.53 | CIA-38 | BRL-CAD: 03brlcad * r37181 10/brlcad/trunk/src/librt/primitives/arb8/arb8.c: see if this helps, make the pointer non-temporary |
| 02:38.54 | brlcad | see if that does it |
| 02:39.18 | ``Erik | succeeded |
| 02:39.38 | brlcad | welp, there's a pattern to follow at least |
| 02:39.42 | brlcad | seems retaded |
| 02:39.50 | brlcad | retaaaded |
| 02:39.54 | ``Erik | ghah, ughly |
| 02:43.29 | ``Erik | but functional :( |
| 02:45.47 | CIA-38 | BRL-CAD: 03erikgreenwald * r37182 10/brlcad/trunk/src/librt/primitives/metaball/metaball.c: Use temp const to pass instead of casting |
| 02:49.04 | ``Erik | "metro-curious" heh |
| 02:52.36 | Ralith | brlcad: cool, thanks |
| 03:01.26 | *** join/#brlcad akafubu (n=akafubu@unaffiliated/akafubu) |
| 04:55.06 | CIA-38 | BRL-CAD: 03brlcad * r37183 10/brlcad/trunk/src/libged/ (ged_private.h rtcheck.c wdb_obj.c): quell the log() shadow warnings, rename to logstr. |
| 05:08.00 | CIA-38 | BRL-CAD: 03brlcad * r37184 10/brlcad/trunk/src/libged/ (3ptarb.c adc.c analyze.c attr.c bev.c bigE.c): quell a slew of compilation warnings including type mismatches, exact floating point comparisons, and param checks. |
| 05:36.19 | *** join/#brlcad louipc (n=louipc@archlinux/trusteduser/louipc) |
| 11:45.09 | *** join/#brlcad mafm2 (n=mafm@119.Red-81-34-12.dynamicIP.rima-tde.net) |
| 15:05.49 | CIA-38 | BRL-CAD: 03brlcad * r37185 10/brlcad/trunk/src/mged/chgview.c: fix the usage reporting of the drawing commands (E/draw/e/etc) so that usage printing (GED_HELP) is not considered an error. |
| 16:23.44 | *** join/#brlcad archivist (n=archivis@host81-149-119-172.in-addr.btopenworld.com) |
| 17:37.11 | *** join/#brlcad jesica__ (n=jesica@168.226.178.192) |
| 18:04.55 | CIA-38 | BRL-CAD: 03brlcad * r37186 10/brlcad/trunk/ (NEWS src/libged/aet.c src/libged/quat.c src/libged/ypr.c): |
| 18:04.55 | CIA-38 | BRL-CAD: fix a libged refactor bug in view quat and view ypr where you couldn't set quat |
| 18:04.55 | CIA-38 | BRL-CAD: or ypr due to a wrong argc count (it was assuming "view .." as an additional |
| 18:04.55 | CIA-38 | BRL-CAD: arg). this fixes both quat and ypr so you can now set their values as before. |
| 18:14.10 | *** join/#brlcad mafm2 (n=mafm@119.Red-81-34-12.dynamicIP.rima-tde.net) |
| 18:46.57 | CIA-38 | BRL-CAD: 03brlcad * r37187 10/brlcad/trunk/src/tclscripts/mged/garbage_collect.tcl: (log message trimmed) |
| 18:46.57 | CIA-38 | BRL-CAD: complete overhaul of the 'garbage_collect' command. rewrite the command to take |
| 18:46.57 | CIA-38 | BRL-CAD: a lot more care to not accidentally destroy the users data if something |
| 18:46.57 | CIA-38 | BRL-CAD: unexpected happens. instead of copying over the current database, use the |
| 18:46.59 | CIA-38 | BRL-CAD: working keep file as a testing copy and just rename files around as they are |
| 18:47.01 | CIA-38 | BRL-CAD: processed and verified. add more verification checks to make sure the view is |
| 18:47.03 | CIA-38 | BRL-CAD: restored, objects are saved, and that processing failures won't leave the user |
| 18:55.30 | CIA-38 | BRL-CAD: 03brlcad * r37188 10/brlcad/trunk/src/tclscripts/mged/garbage_collect.tcl: if the file is really tiny, it might actually get bigger and that's okay. likewise, the percentage decrease can be dramatic if it's only got a couple objects. |
| 18:56.47 | CIA-38 | BRL-CAD: 03brlcad * r37189 10/brlcad/trunk/src/tclscripts/mged/garbage_collect.tcl: clarify the percentage |
| 19:01.03 | CIA-38 | BRL-CAD: 03brlcad * r37190 10/brlcad/trunk/NEWS: |
| 19:01.03 | CIA-38 | BRL-CAD: the garbage_collect command gets a nearly complete rewrite with considerably |
| 19:01.03 | CIA-38 | BRL-CAD: more data validation checks and care taken to not leave the users .g in a |
| 19:01.03 | CIA-38 | BRL-CAD: damaged state. if various conditions are encountered, the users original .g |
| 19:01.03 | CIA-38 | BRL-CAD: file is left intact. |
| 19:04.31 | CIA-38 | BRL-CAD: 03brlcad * r37191 10/brlcad/trunk/src/tclscripts/mged/garbage_collect.tcl: heed warning to the users to back up their geometry, even if the command is considerably more careful now to not irrevocably damage their data. |
| 20:39.01 | *** join/#brlcad Phurl (n=mdupont@ip-81-210-245-60.unitymediagroup.de) |
| 21:17.55 | *** join/#brlcad cosurgi (n=cosurgi@atak.bl.pg.gda.pl) |
| 23:37.34 | *** join/#brlcad jesica__ (n=jesica@168.226.178.192) |