| 00:00.07 | andrei | the PBCData inside nurbs.h isn't identical to the other one? |
| 00:01.01 | caen23 | let me check |
| 00:01.10 | andrei | if you haven't added the file to CMakelist it should at worst not find it, I believe. |
| 00:02.33 | caen23 | it's not the same. in nurbs.h, it's a typedef struct pbc_data {} PBCData; also, the attributes vary |
| 00:03.36 | caen23 | but what i don't understand is why it conflicts when i put it in a separate file, but it works when it's inside _ext :-? |
| 00:05.27 | andrei | where did you put the file you created? |
| 00:06.21 | caen23 | inside src/libnurbs/ |
| 00:07.27 | andrei | I ll try rebuilding and see what happens |
| 00:07.57 | caen23 | ok, thanks |
| 00:17.05 | andrei | it takes a bit to rebuild, I m at 75% |
| 00:17.38 | caen23 | and it works? with my file? because for me, it fails at 27% if it fails :-? |
| 00:18.32 | andrei | caen23: I first rebuilded the svn copy, I needed to be sure I have the latest version. |
| 00:19.01 | caen23 | oh, ok then |
| 00:19.16 | andrei | if you don't include nurbs.h in your file it will most likely solve the conflict, but I want to see if it doesn't break functionality. |
| 00:22.03 | Notify | 03BRL-CAD:n_reed * 53827 brlcad/trunk/src/librt/primitives/pipe/pipe.c: more parameter reduction and descriptive renaming |
| 00:24.31 | caen23 | btw, is there no way i could simply compile nurbs alone whenever i make a change? |
| 00:25.12 | andrei | there is, if you re running on linux |
| 00:25.18 | andrei | sudo make libnurbs |
| 00:25.28 | caen23 | i am (mac). thanks |
| 00:28.28 | *** join/#brlcad GrantMercer015 (~GrantMerc@ip72-193-246-126.lv.lv.cox.net) | |
| 00:28.55 | GrantMercer015 | where can i get the source code and installation instructions for brl-cad? i cant seem to find it on the google code in site |
| 00:29.01 | *** join/#brlcad n_reed (~molto_cre@BZ.BZFLAG.BZ) | |
| 00:29.21 | caen23 | GrantMercer015 check this out http://brlcad.org/wiki/Compiling |
| 00:29.27 | andrei | <PROTECTED> |
| 00:32.01 | brlcad | alcon: fyi, there's an edward coming by to upgrade our drupal -- I'll have to set up an account for him, so just have him to wait if I'm not around :) |
| 00:32.22 | brlcad | GrantMercer015: also, some info here: http://brlcad.org/wiki/Deuces |
| 00:32.37 | brlcad | in the getting started and code sections |
| 00:32.47 | brlcad | http://brlcad.org/wiki/Deuces#Code |
| 00:33.59 | caen23 | andrei i ran `sudo make libnurbs` and it failed with the same error, even after removing nurbs.h from my file :( |
| 00:35.22 | andrei | yes, it does fail for me aswell. I m currently looking to see what's wrong |
| 00:37.49 | caen23 | perhaps the typedef should be included in the brlcad namespace inside nurbs.h? |
| 00:38.06 | brlcad | caen23: you do not need sudo |
| 00:38.29 | brlcad | that would only be for install and even that isn't needed just to test compilation |
| 00:38.48 | caen23 | i understand, but this can't be why it fails |
| 00:39.08 | brlcad | certainly not |
| 00:39.17 | brlcad | just a very dangerous habit |
| 00:40.23 | brlcad | it's like if you saw someone directing traffic with a shotgun, might not affect how effectively you're directing traffic but it sure is distracting and dangerous :) |
| 00:41.18 | caen23 | i don't usually build with sudo, but i didn't really pay attention this time :D |
| 00:41.29 | andrei | caen23 : after some research I believe you need to remove the typedef and merge the struct with the class. |
| 00:41.57 | caen23 | so the typedef is redundant? |
| 00:42.17 | andrei | <PROTECTED> |
| 00:42.38 | brlcad | that's going down a dangerous route for just breaking out classes/structs from the source files .. merging objects implies needing to change code, not just moving things around |
| 00:42.43 | brlcad | typedefs are not just for C |
| 00:43.14 | brlcad | quite common in C++ actually |
| 00:43.25 | brlcad | especially for dealing with stl classes |
| 00:43.46 | brlcad | bhlegm: which project is yours? |
| 00:44.03 | andrei | brlcad: from what he said, he hasn't taken any yet. |
| 00:44.06 | brlcad | ahh |
| 00:44.17 | brlcad | hasn't caught up with all the backlog |
| 00:44.33 | andrei | the issue that caen23 experiences is that he cannot name a class with a typedef-name |
| 00:44.44 | andrei | there probably is a workaround for this but I don't see it. |
| 00:45.28 | brlcad | so the first question is why is '''tl;dr: read below for etiquette then join with http://webchat.freenode.net/ or get a [http://www.ircreviews.org/clients/ real IRC client program]''' |
| 00:45.34 | brlcad | ep, wrong paste |
| 00:45.44 | brlcad | starting over |
| 00:46.04 | brlcad | so the first question is why the PBCData class that was extracted needs to include nurbs.h |
| 00:46.07 | *** join/#brlcad GrantMercer015 (~GrantMerc@ip72-193-246-126.lv.lv.cox.net) | |
| 00:46.22 | brlcad | and why that wouldn't also be an issue when it's still in the .cpp file |
| 00:47.42 | caen23 | i was wrong to include nurbs.h in my file. i removed it and compilation still fails. why it works when it's inside the _ext file, that i don't know, but i'm guessing it has something to do with cake. not sure. |
| 00:50.15 | brlcad | nah, not a cmake issue |
| 00:50.34 | brlcad | caen23: when you run into oddities like that, have to take a step back |
| 00:50.39 | brlcad | because it fails a sanity check |
| 00:50.43 | *** join/#brlcad andrei_ (~andrei_@188.25.27.167) | |
| 00:50.53 | brlcad | look around line 146, what do you see |
| 00:51.02 | caen23 | in which file? |
| 00:51.05 | brlcad | ext |
| 00:51.11 | brlcad | cpp |
| 00:51.34 | caen23 | line 148: namespace brlcad { |
| 00:51.34 | brlcad | something "different" should jump out at you |
| 00:51.44 | brlcad | bingo |
| 00:51.50 | caen23 | line 145: end of void brep_newton_iterate |
| 00:52.03 | brlcad | everything after that point is contained within a namespace |
| 00:52.09 | brlcad | have you learned namespaces? |
| 00:52.42 | caen23 | i am familiar with them in other languages, but i haven't written much cpp |
| 00:52.50 | brlcad | might as well pretend that line says: |
| 00:52.53 | brlcad | class brlcad { |
| 00:53.05 | caen23 | i understand how they work :D |
| 00:53.18 | brlcad | how does that change your understanding of PBCData that was contained within there |
| 00:54.10 | caen23 | pbcdata (the class) is contained within the brlcad namespace? |
| 00:54.20 | brlcad | yep |
| 00:54.32 | caen23 | so i should wrap it inside that namespace in my file, too? |
| 00:54.37 | brlcad | you don't want to remove the class from a namespace, that's changing structure and .... can cause a naming conflict ;) |
| 00:54.40 | brlcad | bingo |
| 00:54.48 | brlcad | no more conflict |
| 00:54.55 | andrei | Java ! |
| 00:55.00 | caen23 | now i feel stupid for not figuring it out earlier |
| 00:55.04 | caen23 | thank you |
| 00:55.24 | brlcad | caen23: no worries, that's why there's folks to help ;) |
| 00:55.50 | brlcad | you could have written c++ for 5 years and still stumbled around on that for hours |
| 00:56.59 | caen23 | my main problem is that this is the first time i try collaborating on a project, and it's very intimidating at first, even navigating around the code :( |
| 00:57.34 | brlcad | no worries, relax and have fun with it ;) |
| 00:57.48 | caen23 | will do :) |
| 00:58.20 | brlcad | moving a class definition and encountering a name conflict should raise red flags |
| 00:58.51 | brlcad | that's clearly BAD that we also have a PBCData (which is one hell of a retarded obscure name) in a public header |
| 00:59.12 | brlcad | maybe another GCI to merge them or rename both |
| 00:59.13 | *** join/#brlcad GrantMercer015 (~GrantMerc@ip72-193-246-126.lv.lv.cox.net) | |
| 01:00.00 | caen23 | after i finish this task, i could have a stab at solving this problem, too :-?? |
| 01:02.32 | brlcad | absolutely |
| 01:03.20 | caen23 | well… it now seems to compile properly |
| 01:03.33 | brlcad | great |
| 01:03.40 | brlcad | so which have you busted out? |
| 01:04.07 | brlcad | (classes/files) |
| 01:05.07 | caen23 | i started from scratch after i first got this error, so right now, i only have bspline and pbcdata. but it should be easier from now on |
| 01:05.18 | brlcad | at some point, you should wrap up and we can create another task for other work that remains |
| 01:05.48 | brlcad | the effort is what counts, but you might feel better with incremental progress ;) |
| 01:06.02 | brlcad | looking through ext cpp i'm reminded of the other classes |
| 01:06.37 | caen23 | i could finish the current task, and then try working on the typedef issue as a new task. i think it's safer to merge than rename, but i could be mistaken |
| 01:06.57 | brlcad | the ones defined in there should get broken out, like SurfaceTree.cpp and CurveTree.cpp |
| 01:07.04 | brlcad | but they may be tasks on their own |
| 01:07.16 | brlcad | they just happened to be declared in include/nurbs.h |
| 01:07.28 | brlcad | because they're part of libnurbs public API |
| 01:08.32 | caen23 | alright |
| 01:09.25 | GrantMercer015 | having trouble finding the right QtSDK download, does anyone have the right link where they downloaded the sdk? |
| 01:09.29 | brlcad | you can decide whether to include them now or wait till we write up another task |
| 01:09.40 | brlcad | GrantMercer015: er, on the qt website? |
| 01:09.46 | brlcad | what for? |
| 01:10.10 | GrantMercer015 | whoops, wrong chatroom, sorry! |
| 01:10.21 | *** part/#brlcad GrantMercer015 (~GrantMerc@ip72-193-246-126.lv.lv.cox.net) | |
| 01:10.27 | brlcad | heh |
| 01:11.15 | andrei | O.o |
| 01:14.03 | *** join/#brlcad Skriptkid (~Skriptkid@117.202.103.43) | |
| 01:16.44 | caen23 | i've broken GCPData into two files, a .h and a .cpp. now i feel that i should modify CMakeLists, but i'm not sure how |
| 01:20.27 | andrei | well, look where opennurbs_ext.cpp is in CMakeLists |
| 01:20.44 | andrei | where the sources are set |
| 01:21.08 | caen23 | yep |
| 01:21.15 | caen23 | do i simply add my cpp? |
| 01:21.29 | andrei | yes, try that. If it doesn't work I ll check |
| 01:27.40 | brlcad | ``Erik: tasks for upgrading both drupal and mediawiki are already claimed.. :) |
| 01:45.52 | starseeker | glares at dvec.h |
| 01:48.32 | caen23 | brlcad i think i'm ready to submit my work. i tried `svn diff > ~/t1.patch`, but as far as i can see, it does not include information about the new files i wrote :-?? |
| 01:50.48 | *** join/#brlcad moinnadeem (409521fc@gateway/web/freenode/ip.64.149.33.252) | |
| 01:50.59 | andrei | can you put a paste somewhere with t1.patch ? |
| 01:51.27 | andrei | it should be a text file containing something very similar to the diff command. |
| 01:51.35 | moinnadeem | Is Sean from Google Code-In here? |
| 01:51.47 | andrei | yes, you can find him by "brlcad" |
| 01:52.52 | caen23 | andrei: it does contain something very similar to the diff command, but i ran a search for "triangle.h" (it's one of my new files), and i don't find the contents of my new file there |
| 01:53.24 | caen23 | https://gist.github.com/4151884 |
| 01:53.32 | andrei | ah, caen23 |
| 01:54.17 | andrei | caen23 you need to use svn add if you created new files |
| 01:54.26 | andrei | I ll give you an example in a second |
| 01:54.46 | caen23 | done :D |
| 01:54.47 | caen23 | thanks |
| 01:54.53 | andrei | ah, nice :) |
| 01:55.17 | Notify | 03BRL-CAD:starseeker * 53828 brlcad/trunk/include/dvec.h: Put back Sean's version of dvec.h |
| 02:05.11 | Notify | 03BRL-CAD:starseeker * 53829 brlcad/trunk/include/vector_x86.h: Probably not fully correct adaptations to vector_x86.h, but this does get things building when using the newer dvec.h |
| 02:05.44 | starseeker | brlcad: I doubt that's right, but perhaps its a step in the right direction |
| 02:15.40 | brlcad | moinnadeem: hi, give me a few minutes then we can chat |
| 02:17.11 | moinnadeem | brlcad; okay, I still have a bit of homework to finish up, take as long as you need. |
| 02:19.35 | brlcad | starseeker: yeah, at a glance, that doesn't look right but I'd have to run it through the unit test |
| 02:21.21 | starseeker | brlcad: is that the test_dvec test? |
| 02:21.29 | brlcad | yeah |
| 02:21.35 | starseeker | blinks |
| 02:21.40 | starseeker | uh... passes here |
| 02:21.51 | brlcad | it just runs, the testing portion is commented out |
| 02:21.57 | starseeker | ah |
| 02:22.03 | brlcad | that was to make sure it kept compiling ;) |
| 02:22.17 | brlcad | I didn't get a chance to look at the test code yet |
| 02:22.26 | starseeker | no problem |
| 02:22.33 | brlcad | you should at least see whether performance is affected |
| 02:22.49 | brlcad | where'd the 8 byte alignement come from? |
| 02:23.26 | brlcad | that would seem to be a problem right away |
| 02:24.00 | brlcad | (otherwise at best performance is hosed, at worst you get wrong values) |
| 02:24.09 | starseeker | urm. I probably did something nonsensical... |
| 02:24.42 | starseeker | (a sphere brep raytrace did succeed, but other than that it was just a compilation thing) |
| 02:26.13 | brlcad | if you have r53825 handy, you could see if test_dvec is slower/faster/unchanged |
| 02:26.51 | brlcad | whole point of the fpu version is to be faster, right? :) |
| 02:27.14 | starseeker | not sure, actually |
| 02:27.26 | starseeker | fpu looks like the fallback... |
| 02:27.28 | brlcad | that's why it was originally written, trying to implement the paper |
| 02:27.43 | brlcad | supposed to be the other way around |
| 02:28.01 | starseeker | take a look at dvec.h around line 125 |
| 02:28.20 | brlcad | oh I believe you, just saying |
| 02:29.32 | starseeker | so... if the fpu version was getting used everywhere and works, why bother with vector_x86.h? |
| 02:30.34 | brlcad | oh right, sorry -- the files are named funny |
| 02:31.14 | brlcad | x86 == vectorized |
| 02:31.22 | brlcad | so it's right |
| 02:31.54 | brlcad | fpu is just use the cpu, sans SSE |
| 02:32.05 | starseeker | doing a distcheck at the moment - soon as the machine's free I'll try the sph brep raytrace with both |
| 02:33.24 | brlcad | should be vector_sse.h instead of vector_x86.h |
| 02:33.40 | brlcad | and vector_cpu.h or leave it as vector_fpu.h |
| 02:33.59 | brlcad | fpu vs sse is the distinction |
| 02:35.23 | brlcad | moinnadeem: since you're on a web interface, you may want to get a real irc client if you plan on taking on other tasks like this ;) |
| 02:37.37 | moinnadeem | brlcad: The CPU on my laptop failed (off-the-shelf HP, go figure.), so didn't want to install another one on my desktop. May just take your advice and have to ;) |
| 02:39.03 | brlcad | moinnadeem: so have you done this sort of upgrade before? |
| 02:39.27 | moinnadeem | brlcad: Server upgrades, yes. Drupal in specific, no. |
| 02:39.33 | brlcad | okay |
| 02:39.57 | brlcad | it shouldn't be too bad, but I want to make sure of two things |
| 02:40.47 | brlcad | 1) that you don't spend more time on this than you have available -- it might take an hour or a couple days but just let us know how things are going -- we can always break the task up into multiple pieces if it's going to take WAY more time due to something unforseen |
| 02:40.58 | moinnadeem | I'm going to backup everything on a test repository and attempt an upgrade on there first so I can have an idea of what to do |
| 02:41.47 | brlcad | 2) that we end up with forward progress, either learn/document what we need to do next or have it all up and running happy |
| 02:42.15 | moinnadeem | Alright, it shouldn't hopefully, but I'll contact you guys if there are any problems. |
| 02:42.55 | moinnadeem | My goal is to hopefully have everything good, but modules (and breaking it.) are my main worry. How many modules do you have installed? |
| 02:43.05 | brlcad | not many |
| 02:44.03 | moinnadeem | The wiki seems pretty straightforward, I assume your server is running Linux? |
| 02:45.38 | brlcad | do you have a shell preference? |
| 02:45.48 | brlcad | actually the server is running freebsd, but shouldn't matter |
| 02:45.55 | starseeker | and distcheck-full passed |
| 02:46.22 | moinnadeem | Used to Ubuntu, but it shouldn't be much different. |
| 02:48.01 | moinnadeem | What version of Drupal are you currently running, and which do you want upgraded to? |
| 02:48.32 | *** join/#brlcad minichrispy (b82f2f47@gateway/web/freenode/ip.184.47.47.71) | |
| 02:50.31 | starseeker | brlcad: nothing in the way of performance difference to speek of raytracing a sphere |
| 02:50.36 | starseeker | (brep sphere) |
| 02:52.00 | starseeker | however |
| 02:52.36 | starseeker | using fpu, test_dvec takes 1.37 seconds |
| 02:53.27 | starseeker | using x86, test_dvec takes 1.00 seconds |
| 02:54.52 | brlcad | heh |
| 02:55.12 | brlcad | bump the constant in test_dvec up with two more 00's |
| 02:55.19 | brlcad | see if it's still that much |
| 02:55.42 | starseeker | you mean COUNT? |
| 02:56.09 | brlcad | if it is, then either you did make it unusably slower or it's a failed experiment we should eliminate |
| 02:56.12 | brlcad | yeah |
| 02:56.24 | caen23 | brlcad could you review the work on my task? :D |
| 02:56.33 | brlcad | I bumped it down way low so it didn't slow down make test, but to a value too slow for comparisons |
| 02:56.33 | bhlegm | b/c there's 2 gallery tasks, does that mean you want both? |
| 02:56.43 | starseeker | brlcad: so test_dvec should take longer? |
| 02:56.49 | brlcad | caen23: working on it ;) |
| 02:56.55 | caen23 | thanks |
| 02:57.03 | brlcad | bhlegm: depends which looks better |
| 02:57.08 | bhlegm | ah |
| 02:57.32 | brlcad | gallery 3 has lots of interesting features but our gallery 2 install has been a pita slow p.o.s. |
| 02:58.07 | brlcad | v3 might have fixed all our problems, maybe not -- but upgrade should be really simple there |
| 02:58.29 | brlcad | then it'd be nice to evaluate an alternative, see if we want to switch |
| 02:58.37 | brlcad | there are lots of follow-on tasks too |
| 02:59.23 | brlcad | starseeker: to use it for testing performance, it needs to take at least 60 seconds to overcome process startup overhead |
| 02:59.42 | starseeker | nods |
| 02:59.48 | starseeker | running longer test now |
| 02:59.59 | brlcad | moinnadeem: account created but I have a better/slightly different approach than was originally intended |
| 03:00.17 | moinnadeem | Yes? |
| 03:00.28 | brlcad | moinnadeem: instead of tackling the upgrade of our primary site, you can try it on one of our secondary sites (more.brlcad.org) |
| 03:00.32 | brlcad | that'll be a good testing ground |
| 03:00.39 | brlcad | almost the same set of issues |
| 03:01.16 | brlcad | and if anything, you can keep a transcript of what you do and turn that into a how-to (which we'll add as another gci task) |
| 03:01.17 | starseeker | vector_fpu(no SSE): 136.22 sec |
| 03:01.33 | starseeker | vector_x86(SSE2): 100.88 sec |
| 03:01.41 | brlcad | starseeker: before changes or after? |
| 03:01.54 | starseeker | after - that's with current trunk |
| 03:01.58 | moinnadeem | Much better, do you still think it'll be a good idea to upgrade on a backup as a first attempt? |
| 03:02.07 | brlcad | moinnadeem: absolutely |
| 03:02.23 | brlcad | how do you want to proceed with that? |
| 03:03.27 | moinnadeem | brlcad: I was intending on following the instructions from the wiki: http://drupal.org/node/770402 |
| 03:03.28 | brlcad | your account is created, I can put the files and db in a tarball for you to play with, or you can hop on and try in a copy somewhere |
| 03:04.11 | moinnadeem | is thinking |
| 03:04.43 | brlcad | how about this -- I'll make a copy of the database and you can just log in and give it a try |
| 03:04.54 | brlcad | give me a few minutes |
| 03:05.10 | moinnadeem | That works. |
| 03:07.03 | starseeker | calls it a night - brlcad, feel free to undo my fumblings - I was just trying to fix the build with both fpu and x86 - I don't have enough fundamental knowledge of the issue dvec.h and friends are addressing to work intelligently |
| 03:07.32 | brlcad | moinnadeem: see PMs |
| 03:08.09 | brlcad | starseeker: that's actually somewhat reassuring -- it's somewhat faster |
| 03:08.57 | brlcad | "should" be like 13s, but it's good |
| 03:09.21 | starseeker | except there's no excuse at all for it's even being functional :-P |
| 03:10.17 | brlcad | well, you said it still renders, so that's promising too |
| 03:10.32 | brlcad | dunno if it's exercised, but hopefully |
| 03:10.40 | starseeker | not sure |
| 03:11.39 | brlcad | could run in gdb and set a breakpoint, but it sounds good to me |
| 03:12.18 | starseeker | probably will break when built as float instead of double though |
| 03:13.00 | starseeker | that might be an interesting test to add to distcheck-full once its working again, if that's something that can be controlled with configuration options |
| 03:14.06 | starseeker | jots down a low priority note to self to try BRL-CAD with clang -Wdocumentation, just to see what happens... |
| 03:14.47 | brlcad | starseeker: that looked like a proposal to implement that feature, not that it was done no? |
| 03:15.38 | starseeker | brlcad: it seems to be implemented, although I'm not sure it's on by default yet - what got me hunting docs on it was a note in the Ninja git logs about them using it to test their code |
| 03:16.01 | starseeker | and I've seen remarks about the clang/llvm guys using it on their own code base |
| 03:16.57 | starseeker | http://lists.cs.uiuc.edu/pipermail/llvmdev/2012-August/052887.html |
| 03:17.46 | starseeker | that presentation on Nov 8th looks like it was presenting actual results, not just a proposal |
| 03:21.27 | starseeker | this looks potentially interesting for us vim guys: https://github.com/gribozavr/clang_complete |
| 03:39.22 | brlcad | that's even more awesome then |
| 04:01.42 | *** join/#brlcad caen23 (~cezar@92.81.166.138) | |
| 04:46.54 | *** join/#brlcad Alex__ (b82f2f47@gateway/web/freenode/ip.184.47.47.71) | |
| 04:48.09 | brlcad | caen23: ALMOST |
| 04:48.19 | brlcad | reviewed and looks great |
| 04:48.25 | brlcad | only two minor issues |
| 04:49.02 | brlcad | per our developer guide, local dir headers should be indicated with ./ (e.g., #include "./bspline.h") |
| 04:49.27 | brlcad | and headers should also be listed in the CMakeLists.txt file |
| 04:50.03 | brlcad | do that and it'll be golden |
| 04:53.20 | Alex__ | Hi, I had a few questions regarding my task on Google Code-In and was wondering if this was the place to get some feedback |
| 04:59.03 | *** join/#brlcad GrantMercer015 (~GrantMerc@ip72-193-246-126.lv.lv.cox.net) | |
| 04:59.19 | GrantMercer015 | Hello! where can i find the brl-cad source download? |
| 04:59.40 | Alex__ | http://brlcad.org/d/download |
| 04:59.51 | GrantMercer015 | thanks! |
| 05:00.09 | Alex__ | no problem! |
| 05:02.56 | *** join/#brlcad jarray52 (~bigbear@unaffiliated/jarray52) | |
| 05:04.57 | *** join/#brlcad harmanpreet (~chatzilla@202.164.53.117) | |
| 05:09.59 | starseeker | brlcad: it's in there all right, at least in the latest trunk srcs: http://pastebin.mozilla.org/1962435 |
| 05:10.24 | bhlegm | hi, could i get some help w/ the virtual image? in virtual box, do i just use the ova file in Import Appliance? |
| 05:10.48 | starseeker | bhlegm: um. I'm embarassed to say I'm not sure |
| 05:11.10 | bhlegm | alright :P |
| 05:11.33 | starseeker | ah http://www.virtualbox.org/manual/ch01.html |
| 05:11.50 | starseeker | select "File" -> "Import appliance" from the Manager window. In the file dialog that comes up, navigate to the file with either the .ovf or the .ova file extension. |
| 05:12.05 | starseeker | bhlegm: looks like yes |
| 05:12.45 | bhlegm | oh dear |
| 05:14.12 | *** join/#brlcad elf_ (~elf@p5.eregie.pub.ro) | |
| 05:20.47 | bhlegm | http://pastie.org/5440846 should i try redownloading it? ... |
| 05:21.44 | starseeker | um. Yeah, possibly - does your filesize match that of the file on sf? |
| 05:26.31 | *** join/#brlcad archivist (~archivist@host81-149-189-98.in-addr.btopenworld.com) | |
| 06:19.49 | *** part/#brlcad jarray52 (~bigbear@unaffiliated/jarray52) | |
| 08:13.15 | *** join/#brlcad elf_ (~elf@p5.eregie.pub.ro) | |
| 08:14.17 | *** join/#brlcad harmanpreet (~chatzilla@202.164.53.117) | |
| 10:39.10 | *** join/#brlcad harmanpreet (~chatzilla@202.164.53.117) | |
| 10:46.01 | *** join/#brlcad Skriptkid (~Skriptkid@117.202.101.128) | |
| 10:47.08 | Skriptkid | I'm on the LIBBN comments task but didn't quite understand where in the header files the comments need to be moved to. Could anybody please explain it? |
| 11:10.59 | *** join/#brlcad harmanpreet (~chatzilla@202.164.53.117) | |
| 11:44.56 | ``Erik | Skriptkid: comments describing functions/globals should be immediately before the func/global... bn.h should generally be arranged in sections based on files and file comments should be at the beginnings of those sections (inside the @addtogroup @{ stuff ) |
| 11:45.43 | ``Erik | the anim group in bn.h is how it should all be |
| 11:46.31 | *** join/#brlcad toni (tonimitkov@78.90.2.133) | |
| 11:46.45 | ``Erik | bu.h has already had the treatment, if you want a full example |
| 11:47.14 | toni | hi |
| 11:47.30 | toni | hi can i talk with some admin ? |
| 11:47.45 | Skriptkid | So if there's a function xyz() with certain comments above it, I cut the comments and paste them on top of the function declarations in the bn.h header file? |
| 11:47.50 | ``Erik | toni: sure, ask questions and eventually someone will answer |
| 11:48.24 | Skriptkid | declaration with the BN_EXPORT before it, that is(As the description says) |
| 11:48.33 | ``Erik | Skriptkid: if the comment above it describes the function, yes... a little bit of human judgement is required to see if it's appropriate, otherwise it'd be scripted |
| 11:50.35 | Skriptkid | ok |
| 11:52.26 | Skriptkid | do we have to move the comments at the beginning of the file, below the license agreement also? To that .C file's section in header file? |
| 11:53.09 | ``Erik | if the comments describe the purpose of the file/section, they should be duplicated at the beginning of the section in the header |
| 11:54.05 | Skriptkid | ok |
| 11:57.01 | ``Erik | Skriptkid: the purpose is so a codemonkey can quickly find if the library has the desired functionality and what the interface is by reading the header instead of looking at the implementation details, the funky markup is to support doxygen (a c/c++ javadoc type tool) |
| 11:57.32 | Skriptkid | got it. Thanks :) |
| 11:58.02 | ``Erik | np and good luck :) |
| 11:58.45 | *** join/#brlcad harmanpreet (~chatzilla@202.164.53.117) | |
| 12:00.10 | ``Erik | starseeker: I was right about the mirc color thing, it's freenode filtering. they added a +c mode on channels O.o I can haz kulrz nao |
| 12:03.07 | caen23 | i have a question regarding the man to docbook conversion task |
| 12:04.38 | caen23 | i've converted all the .1 files to xml using doclifter, and then i tried viewing the first one (bot_dump.1.xml), and all i've got is text on a single line, no formatting whatsoever. is this how it works, or have i done something wrong? |
| 12:05.27 | ``Erik | caen23: are you viewing with 'notebook' on 'windows'? |
| 12:05.43 | caen23 | no, with chrome and an xml viewer on a mac |
| 12:06.59 | ``Erik | hm, d'no, could always use vim or emacs in a terminal.app window to look at the file and see if it looks like reasonable xml? or blast it through the docbook parser to see if you can output decent ps/pdf or something? |
| 12:07.26 | ``Erik | the guy who knows the most on that topic is probably still asleep, he should be online in a few hours |
| 12:07.53 | caen23 | the file does look like valid xml, but when i try viewing it with chrome (for example), it doesn't look like a man page. i'll try running it through the docbook parser |
| 12:13.47 | caen23 | well the pdf seems fine |
| 12:13.59 | caen23 | thanks for helping |
| 12:39.31 | *** join/#brlcad Al_Da_Best (Al_Da_Best@cpc2-shep12-2-0-cust21.8-3.cable.virginmedia.com) | |
| 13:18.44 | *** join/#brlcad DarkCalf (~DarkCalf@173.231.40.98) | |
| 13:19.55 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) | |
| 13:20.41 | *** join/#brlcad n_reed_ (~molto_cre@BZ.BZFLAG.BZ) | |
| 13:37.42 | *** join/#brlcad moinnadeem (~androirc@233.sub-174-229-203.myvzw.com) | |
| 14:19.23 | Notify | 03BRL-CAD:starseeker * 53830 brlcad/trunk/src/librt/test_subdivision.cpp: Remove unused variable assignment. |
| 14:23.44 | Skriptkid | I finished a task and I need to submit it. So do I just create a patch file as shown on the wiki and upload it on the task page? |
| 14:30.21 | Skriptkid | Or on Sourceforge? |
| 14:49.42 | Notify | 03BRL-CAD:starseeker * 53831 brlcad/trunk/src/librt/test_subdivision.cpp: Not doing the right thing with edges yet, but at least recognize them. |
| 14:53.57 | brlcad | Skriptkid: sourceforge unnecessary |
| 14:58.55 | Skriptkid | ok |
| 15:15.25 | brlcad | starseeker: what do the folks running doclifter need to know/do? anything specific? |
| 15:18.48 | *** join/#brlcad harmanpreet (~chatzilla@124.253.21.57) | |
| 15:34.25 | *** join/#brlcad Skriptkid (~Skriptkid@117.202.101.128) | |
| 15:44.09 | starseeker | brlcad: I don't recall any specific options to doclifter (although it's been quite a while and I see there have been new doclifter releases) |
| 15:44.37 | starseeker | brlcad: the main thing is to be aware that doclifter may not quite get everything right |
| 15:44.49 | starseeker | i.e. you still have to review it and make tweaks as needed |
| 15:47.01 | brlcad | how will they/we know? should make fail if they add to our build? |
| 15:48.53 | *** join/#brlcad caen23 (~cezar@92.81.166.138) | |
| 15:49.34 | caen23 | i tried completing a task that used doclifter. doclifter 2.9 produces docbook 4.4. for some files, i had some small errors (for example, one file had an &tnspace; and it seemed illegal or smth), but i was still able to convert to pdf with apache fop |
| 15:51.44 | brlcad | caen23: how'd you handle the errors? |
| 15:52.16 | caen23 | they didn't seem to break anything, so i just left those ther |
| 16:00.15 | brlcad | how do you know it was an error then? |
| 16:00.53 | caen23 | brlcad: because when i opened the xml with chrome, it displayed the error message before the actual content |
| 16:01.32 | starseeker | Hrm. OK, if doclifter is producing 4.4, that's something that needs to be switched - we're on DocBook 5.0 |
| 16:01.45 | brlcad | ahh, xml errors should be fixed too |
| 16:02.11 | starseeker | man1/en/rt.xml is an example of a doclifter output that's been minimally tweaked to be 5.0 compatible |
| 16:02.49 | brlcad | fop might be okay with it, but not necessarily other tools we'll feed them through |
| 16:02.49 | starseeker | (still got the comment, in fact...) |
| 16:02.49 | starseeker | xmllint is probably helpful for checking docbook xml files |
| 16:03.23 | caen23 | oh, ok then. i'll rework the task |
| 16:03.25 | *** join/#brlcad Skriptkid (~Skriptkid@117.202.101.128) | |
| 16:03.32 | brlcad | the important part being to just run db4-upgrade, yes? |
| 16:03.48 | starseeker | I believe so |
| 16:03.56 | brlcad | looks like that was run instead of doclifter spitting out 5 |
| 16:04.46 | caen23 | i've just noticed that doclifter produces db5 if run with the -x option and 4.4 otherwise |
| 16:05.07 | starseeker | ah, excellent - then we'll want to try the -x output |
| 16:06.17 | starseeker | another xml validation tool is Oracle's Multi-Schema XML Validator - that may have nicer output than xmllint |
| 16:07.14 | starseeker | The ideal way to do this would be to turn on BRLCAD_EXTRADOCS_VALIDATE, which incorporates the xml checking into the build process |
| 16:07.32 | starseeker | problem is lots of our existing man pages in the build need tweaks to pass that test |
| 16:07.57 | brlcad | really? that sounds like a great gci task |
| 16:08.25 | starseeker | that might actually be a logical starting point for caen23, if he's interested in the DocBook tasks |
| 16:08.53 | starseeker | once he figures out the fix for existing docs, that should go pretty quickly - most of them will be very similar |
| 16:09.17 | caen23 | apparently, all the errors xmllint finds are related to   perhaps i should file a bug report? :-?? |
| 16:09.18 | *** join/#brlcad SlightRemorse (slightremo@85.11.180.148) | |
| 16:09.56 | starseeker | caen23: what's the error? |
| 16:10.01 | brlcad | caen23: file a bug report with whom? |
| 16:10.06 | caen23 | esr :D |
| 16:10.17 | caen23 | starseeker: fast4-g.1.xml:120: parser error : Entity 'thinsp' not defined |
| 16:10.25 | brlcad | so that'd be something to fix |
| 16:10.25 | SlightRemorse | I've claimed a task, but it has no link to the files I need to work on |
| 16:10.47 | starseeker | caen23: the first thing is to find out what 'thinsp' is supposed to do |
| 16:10.56 | caen23 | insert a "thin space" |
| 16:11.51 | brlcad | screw that, just make it a space or a nbsp depending on where it's at |
| 16:11.53 | starseeker | caen23: so the "quick fix" there is to try just replacing thinsp with a regular space, and see if the "thin space" is really needed |
| 16:12.02 | starseeker | er, yeah |
| 16:13.15 | starseeker | caen23: do you have a build of BRL-CAD? |
| 16:13.17 | caen23 | star seeker brlcad: i've done that just now. it seems to be the same thing, so i'm going to just remove thinsp altogether. it was only used inside the "example" sections, where the usage of a shell command is exemplified |
| 16:13.23 | caen23 | yes, i do |
| 16:14.08 | SlightRemorse | I've claimed the task "Close MGED when both windows are closed" which needs me to work on the 2 following files "src/mged/mged.c" and "src/tclscripts/mged/openw.c". But there is no link to any repository or anything. Am I missing something? |
| 16:14.40 | starseeker | caen23: OK - if you're interested, I can define a GCI task that will involve fixing existing XML errors in our docs - that will both be extremely useful and pave the way for easier work on new DocBook files |
| 16:14.57 | caen23 | SlightRemorse: http://brlcad.org/wiki/Compiling |
| 16:15.03 | caen23 | starseeker: alright |
| 16:15.41 | starseeker | caen23: give me a couple minutes - apparently a glitch has crept into the DocBook cmake macros... |
| 16:17.00 | SlightRemorse | caen23: thank you |
| 16:20.45 | starseeker | caen23: so here's an example of the types of errors you'll see in our existing pages: http://pastebin.mozilla.org/1963550 |
| 16:21.47 | brlcad | SlightRemorse: also http://brlcad.org/wiki/Deuces#Code, see the details at the top |
| 16:23.09 | Notify | 03BRL-CAD:n_reed * 53832 brlcad/trunk/src/librt/primitives/pipe/pipe.c: clarify bend drawing algorithm |
| 16:23.15 | starseeker | caen23: if you look up funcprototype here: http://docbook.org/tdg5/en/html/docbook.html |
| 16:23.55 | brlcad | so I'm thinking we should just make that a second task if we already have that for existing pages |
| 16:24.10 | starseeker | you see that the "Parent" section of funcprototype lists only funcsynopsis as a possible parent |
| 16:24.22 | brlcad | do the conversions, fix the blatant xml errors .. then we can have another task to go through and clean up the schema errors |
| 16:24.25 | starseeker | so we need an enclosing funcsynopsis tag set |
| 16:24.33 | starseeker | brlcad: agreed |
| 16:25.20 | brlcad | caen23: how long did it take you to submit t2.patch? |
| 16:25.30 | starseeker | brlcad: I was illustrating that one to allow everyone to gauge whether it looked reasonable as a task |
| 16:25.40 | brlcad | i mean how long did you work running doclifter and such |
| 16:25.42 | starseeker | will add it to the Deuces page |
| 16:25.49 | caen23 | brlcad: two minutes at most, i've got shell scripts to do most everything for me |
| 16:25.53 | brlcad | starseeker: definitely a reasonable task :) |
| 16:26.09 | brlcad | starseeker: more a question of how many errors they need to fix for completion |
| 16:26.26 | brlcad | all of them in a particular directory, all errors everywhere, .. |
| 16:27.05 | brlcad | caen23: so I mean total time, time to write those scripts, figure out the tools, run them, review the xml errors, etc |
| 16:27.13 | starseeker | right - that's probably why I didn't think to put it down intitially |
| 16:27.19 | brlcad | i know running doclifter itself is just a few seconds |
| 16:27.49 | starseeker | supposes we could start with something like "20 man pages" and tweak as needed... |
| 16:28.00 | caen23 | brlcad: sorry, i believe around an hour at most. the hardest part was trying to set up a db to pdf converter |
| 16:30.31 | *** join/#brlcad Skriptkid (~Skriptkid@117.202.101.128) | |
| 16:31.57 | caen23 | brlcad: i've got a second patch for teh conv directory, this time with db5 and no errors |
| 16:41.17 | brlcad | awesome |
| 16:41.34 | brlcad | then they're scoped about right |
| 16:41.35 | caen23 | brlcad: not so awesome, i've managed to break something :( fixing it now |
| 16:41.43 | brlcad | :) |
| 16:42.23 | starseeker | brlcad: first cut at Deuce up - see "Fix strict XML errors in 30 BRL-CAD DocBook documents" |
| 16:43.02 | brlcad | looks |
| 16:43.49 | starseeker | fixes minor formatting glitch |
| 16:47.36 | Notify | 03BRL-CAD:starseeker * 53833 brlcad/trunk/misc/CMake/DocBook.cmake: Fix DB_VALIDATE_TARGET calls. |
| 16:47.54 | caen23 | how can i redownload a folder from svn? |
| 16:48.05 | starseeker | caen23: svn up |
| 16:48.33 | starseeker | (from the parent that wasn't deleted) |
| 16:49.15 | starseeker | OK, BRLCAD_EXTRADOCS_VALIDATE is back in business |
| 16:50.49 | caen23 | the folder i need to patch had smth like file1.xml file2.xml when i first downloaded it from the svn. but i've messed things up a bit, and now svn up reverts to a local revision that includes file3.xml and file4.xml that have been added by me, and are not on the remote repo |
| 16:51.40 | starseeker | You can clear those (iirc) by doing svn revert file3.xml file4.xml |
| 16:51.58 | starseeker | then remove them - svn should have undone the adds |
| 16:52.48 | caen23 | starseeker: svn revert file3.xml echoes "Skipped file3.xml" :-?? |
| 16:53.24 | starseeker | what if you just remove file3.xml and file4.xml, then svn up? |
| 16:54.26 | caen23 | now they seem to be gone, but i can't tell which ones were there first, and which ones i added |
| 16:54.44 | starseeker | what does "svn status" say? |
| 16:54.51 | caen23 | at least not without looking at CMakeList |
| 16:55.10 | caen23 | the new files have a ? on their left |
| 16:55.15 | starseeker | right |
| 16:55.38 | starseeker | so what's the difficulty? |
| 16:55.55 | starseeker | you want to completely revert the whole directory? |
| 16:55.59 | caen23 | yes |
| 16:56.33 | starseeker | I usually cheat - I cd into the parent directory, rm -rf the one I want to reset, then svn up from the parent |
| 16:56.53 | starseeker | not the most efficient approach, but it should pull down a clean copy of the directory in question |
| 16:57.03 | caen23 | ooh, ok, so i can't do it in one svn command, then, right? |
| 16:57.20 | starseeker | I know there's a command for that in git, but I don't know about svn |
| 16:58.21 | caen23 | yeah, i've only used git prior |
| 17:00.25 | starseeker | caen23: did removing the directory and updating work? |
| 17:00.36 | caen23 | yes, it did |
| 17:00.41 | starseeker | cool |
| 17:01.18 | starseeker | if you want to try that docbook validation, make sure you have that last commit (r53833) |
| 17:01.47 | caen23 | that's the number i get from svn up :D |
| 17:02.14 | starseeker | ah, but did you run svn up in the top level or just in your subdirectory? |
| 17:02.55 | caen23 | just in the subdir |
| 17:03.21 | starseeker | IIRC, that updated just the subdirectory - you'll need to run it again at the top level |
| 17:03.34 | caen23 | brlcad: i've submitted the final patch. conv.path , 76.3 kb |
| 17:03.47 | caen23 | starseeker: will do, before taking on that assignment |
| 17:11.33 | caen23 | starseeker: i think there's something wrong with the docbook and the entities. i'm trying to do the same thing for src/fb, and I get errors for "Entity 'nbsp' not defined" :-? maybe the xml namespace has something to do with this? |
| 17:14.01 | caen23 | starseeker: i think i now understand why the default for doclifter is 4.4. there seems to be no namespce or dtd or schema for docbook 5. http://docbook.org/ns/docbook |
| 17:28.18 | *** part/#brlcad caen23 (~cezar@92.81.166.138) | |
| 17:32.04 | *** join/#brlcad caen23 (~cezar@92.81.166.138) | |
| 17:38.01 | Notify | 03BRL-CAD:n_reed * 53834 brlcad/trunk/src/librt/primitives/pipe/pipe.c: generalize bend drawing to an arbitrary number of connecting lines |
| 18:04.14 | *** join/#brlcad elf` (~elf@p5.eregie.pub.ro) | |
| 18:47.06 | *** join/#brlcad Al_Da_Best (Al_Da_Best@cpc2-shep12-2-0-cust21.8-3.cable.virginmedia.com) | |
| 18:47.47 | Notify | 03BRL-CAD:n_reed * 53835 brlcad/trunk/src/librt/primitives/pipe/pipe.c: calculate bend normal locally instead of passing it as an argument |
| 18:59.13 | *** join/#brlcad Alex______ (450f6b03@gateway/web/freenode/ip.69.15.107.3) | |
| 19:00.39 | brlcad | ``Erik: what's the notify e-mail addr? |
| 19:08.38 | ``Erik | notify@elfga.com |
| 19:09.35 | brlcad | sprinkles a few "svn up brlcad.svn" references on the wiki |
| 19:11.03 | ``Erik | I've updated all 3 'NeedReview' to 'NeedsWork' :/ 2 for the svn up issue, one for doing *.1 instead of *.1 */*.1 |
| 19:11.23 | brlcad | awesome |
| 19:12.05 | brlcad | and I just pointed the cianotify extension so it'll be getting xml-rpc messages via e-mail ;) |
| 19:12.14 | brlcad | implement support for that and you'll be cia-compatible :) |
| 19:12.36 | ``Erik | heh, why not use the actual xmlrpc stuff? (it's defined as using http for transport) |
| 19:12.48 | brlcad | doing that via php is a pita |
| 19:13.32 | ``Erik | heh, "php is a pita, so let's make EVERYONE suffer!" :> *duck* |
| 19:14.23 | brlcad | if you already support it, just strip out the xml and feed it internally, no? |
| 19:14.25 | ``Erik | I thought the php to do that kinda stuff was simple, just shove the xml in an http request and it goes |
| 19:15.04 | ``Erik | once I see an email, I'll look at how to detect if it's xml and add a logic fork there, I guess O.o |
| 19:15.21 | brlcad | I would have thought so too, but most docs mention recompiling php with an extension or opening a raw socket yourself |
| 19:15.55 | brlcad | netiher of which would work from behind a managed service |
| 19:16.13 | brlcad | it should have already sent one |
| 19:17.53 | brlcad | aaand another |
| 19:18.21 | ``Erik | ok, now I have the procmail rules in place O.o heh |
| 19:20.18 | *** join/#brlcad iwanmanew (~iwanmanew@94.236.208.130) | |
| 19:21.36 | ``Erik | heh, typo in my procmailrc match line, that's fixed |
| 19:23.55 | *** join/#brlcad caen23 (~cezar@92.81.166.138) | |
| 19:36.47 | *** join/#brlcad maths22 (ada798d1@gateway/web/freenode/ip.173.167.152.209) | |
| 19:42.36 | ``Erik | https://github.com/erikg/cl-cia pheer it |
| 19:53.00 | starseeker | caen23: if doclifter's db5 output isn't workable, you may be able to use the db4 output with the converter: http://mwop.net/blog/264-Converting-DocBook4-to-DocBook5.html |
| 19:55.25 | caen23 | starseeker: all the errors are related to entities, which seem to be considered invalid. other than that, everything works fine. i believe there is something wrong with db5 lacking a namespace |
| 19:56.02 | starseeker | caen23: can you post one of the errors you are seeing? |
| 19:56.36 | caen23 | starseeker: proe-g.1.xml:86: parser error : Entity 'ldquo' not defined |
| 19:56.49 | caen23 | all are the same, but the entities may vary |
| 19:56.58 | starseeker | caen23: actually, if you look at that mwop blog article, the first section is on handling entities |
| 19:57.54 | starseeker | caen23: what if you use the db4 output and then the db4-upgrade.xsl translation? |
| 19:59.41 | caen23 | starseeker: i've just read the article; as far as i can tell, db4-upgrade removes all the entities without replacing them |
| 20:01.11 | caen23 | as far as i can tell, he basically does the same thing to fix the problem :-? |
| 20:02.25 | caen23 | db4 to db5: normally, &hypen; is removed, but we don't want that, so he replaces ‐ with [hyphen]. this way, conversion takes place without removing entities. then, after he's got the db5 format, he simply replaces [hyphen] with the actual "-" :-? |
| 20:02.29 | starseeker | caen23: so essentially, doclifter's problem is that it leaves the entities in place? |
| 20:04.35 | caen23 | starseeker: i don't think it's doclifter's problem :-? i'm not at all familiar with xml, but as far as i can tell, you need to define what "‐" means inside a file on the internet. but db5 lacks those files, so xmllint doesn't recognise "‐" and treats it as an error |
| 20:05.00 | starseeker | nods - replacing it with the literal '-' is probably fine |
| 20:06.04 | caen23 | starseeker: so basically, xml works like C headers |
| 20:07.09 | Notify | 03BRL-CAD:carlmoore * 53836 brlcad/trunk/doc/BRL-CAD.bib: fix spelling |
| 20:07.28 | caen23 | for now, i'll just replace each entity with the literal character, assuming db5 is broken. if anyone knows any better, feel free to weigh in |
| 20:09.45 | starseeker | caen23: that should be fine |
| 20:10.36 | caen23 | alright |
| 20:11.10 | starseeker | IIRC, the only situation where a literal character is a problem is the less than/greater than characters - I think < and > work for those |
| 20:11.29 | brlcad | 2410 errors |
| 20:11.49 | starseeker | I saw 1021 after piping through uniq... |
| 20:12.56 | iwanmanew | Hello, I've the following task "Write a tutorial on compiling BRL-CAD with Eclipse on Linux", is there any place where I can see how it's going on ? |
| 20:13.00 | *** join/#brlcad M0J0E (~johannes@p4FDFB793.dip.t-dialin.net) | |
| 20:13.28 | brlcad | 812 if I strip out the make lines |
| 20:13.54 | starseeker | hmm - OK, you're probably doing a better filtering than I am |
| 20:15.53 | starseeker | caen23: this may be helpful: http://www.sagehill.net/docbookxsl/Db5Entities.html#CharEntities |
| 20:16.08 | Notify | 03BRL-CAD Wiki:Sean * 4670 /wiki/Compiling: bold, not italic |
| 20:16.11 | Notify | 03BRL-CAD Wiki:Sean * 4671 /wiki/Deuces: bold it |
| 20:16.14 | Notify | 03BRL-CAD Wiki:Sean * 4672 /wiki/Compiling: add quotes |
| 20:16.17 | Notify | 03BRL-CAD Wiki:128.63.32.74 * 4669 /wiki/Compiling: mention svn up |
| 20:16.18 | starseeker | woot |
| 20:16.20 | Notify | 03BRL-CAD Wiki:128.63.32.74 * 4668 /wiki/Deuces: mention svn up |
| 20:16.51 | *** join/#brlcad GrantMercer015 (GrantMerce@ip72-193-246-126.lv.lv.cox.net) | |
| 20:17.24 | GrantMercer015 | does anyone know when mentors are usually online? |
| 20:17.50 | starseeker | caen23: if we end up needing some characters, that may be a way to go |
| 20:19.37 | caen23 | starseeker: hmm… they seem to suggest i should find/create a file that contains those entities in the format xml understands. i should look into it, but in the mean time, i've written some simple scripts that simply replace entities with literals :-? it's a poor method, indeed, but i'll look into xml later |
| 20:20.28 | starseeker | caen23: I'm not too worried about the literals, for the moment - it may be a separate task later to investigate the special characters, in fact |
| 20:21.01 | Notify | 03BRL-CAD Wiki:Sean * 4670 /wiki/Compiling: bold, not italic |
| 20:21.04 | Notify | 03BRL-CAD Wiki:Sean * 4671 /wiki/Deuces: bold it |
| 20:21.07 | Notify | 03BRL-CAD Wiki:Sean * 4672 /wiki/Compiling: add quotes |
| 20:21.10 | Notify | 03BRL-CAD Wiki:128.63.32.74 * 4669 /wiki/Compiling: mention svn up |
| 20:21.18 | caen23 | starseeker: indeed |
| 20:23.21 | ``Erik | #!~@!~@! damn your destructive side effects, #'sort |
| 20:23.53 | ``Erik | ok, xml email seems to be working and we have wiki noise coming to channel now |
| 20:25.21 | starseeker | caen23: a quick grep for & characters shows <, > (which are both necessary to allow for < and > in XML) and a few other characters scattered in various files - those characters may actually constitute some of the XML strict failures, in fact... |
| 20:26.38 | caen23 | starseeker: xmllint does not seem to complain about those |
| 20:27.08 | starseeker | nods - thats a Good Thing :-) |
| 20:29.49 | Notify | 03BRL-CAD:n_reed * 53837 brlcad/trunk/src/librt/primitives/pipe/pipe.c: pull handling of inner and outer diameters into separate functions |
| 20:30.46 | brlcad | caen23: http://www.google-melange.com/gci/task/view/google/gci2012/7950213 |
| 20:31.10 | ``Erik | http://paste.lisp.org/display/133935 some day, I'll have to fix that... |
| 20:32.37 | iwanmanew | Is the mentor here ? |
| 20:33.24 | brlcad | ~ask |
| 20:33.24 | ibot | Questions in the channel should be specific, informative, complete, concise, and on-topic. Don't ask if you can ask a question first. Don't ask if a person is there; just ask what you intended to ask them. Better questions more frequently yield better answers. We are all here voluntarily or against our will. |
| 20:33.34 | brlcad | iwanmanew: what's your real question :) |
| 20:34.07 | caen23 | brlcad: i will try to fix those too, after i am done with the current task. i didn't notice the subdirs at first, lol :( |
| 20:34.41 | brlcad | ah, the beauty of the "find" command |
| 20:34.47 | brlcad | find . -name \*.1 |
| 20:34.56 | iwanmanew | i have the task "Write a tutorial on compiling BRL-CAD with Eclipse on Linux", what actually i have to do ? |
| 20:35.02 | brlcad | find . -name \*.1 -exec doclifter -x {} \; |
| 20:35.33 | iwanmanew | just to set Eclipse to work with Makefiles, Microsoft Visual Studio build files, XCode project files and document that ? |
| 20:35.53 | brlcad | iwanmanew: you have to actually write a tutorial on compiling brl-cad with eclipse... |
| 20:36.18 | brlcad | your write-up doesn't involve msvc, xcode, or anything else |
| 20:36.22 | ``Erik | steals with gleeful abandon |
| 20:36.24 | ``Erik | !notify ask |
| 20:36.25 | Notify | Questions in the channel should be specific, informative, complete, concise, and on-topic. Don't ask if you can ask a question first. Don't ask if a person is there; just ask what you intended to ask them. Better questions more frequently yield better answers. We are all here voluntarily or against our will. |
| 20:37.00 | brlcad | iwanmanew: have you used eclipse before? |
| 20:37.09 | iwanmanew | yes, and ? |
| 20:37.14 | brlcad | have you used cmake before? |
| 20:37.32 | iwanmanew | no |
| 20:37.41 | brlcad | so cmake is a build system generator |
| 20:38.00 | brlcad | it makes build system files (like makefiles, msvc project files, xcode projects, etc) |
| 20:38.10 | brlcad | brl-cad's build system uses cmake |
| 20:38.34 | brlcad | but instead of letting it generate makefiles (the default), you're going to tell it to create an eclipse project |
| 20:39.09 | brlcad | then you're going to compile within eclipse |
| 20:39.15 | brlcad | and you're going to write this all down with pretty screenshots telling others exactly how they can do what you just did too |
| 20:39.45 | iwanmanew | ok, thank you :) |
| 20:39.50 | brlcad | should be quick n' easy, but you'll have to read the reference links in the task to know how |
| 20:41.19 | caen23 | crap, i really need to up my cli-fu |
| 20:42.30 | brlcad | yes, yes you do |
| 20:42.32 | brlcad | everyone does |
| 20:43.04 | brlcad | behold the power of the cli |
| 20:43.06 | brlcad | t |
| 20:57.23 | Notify | 03BRL-CAD:n_reed * 53838 (brlcad/trunk/src/librt/primitives/pipe/pipe.c brlcad/trunk/src/librt/primitives/table.c): stub pipe lod plot routine |
| 20:59.38 | *** join/#brlcad GrantMercer015 (~GrantMerc@unlv-nat-131-216-14-33.dhcp.unlv.edu) | |
| 20:59.50 | GrantMercer015 | Anyone know when the mentors are usually online? |
| 21:00.35 | brlcad | GrantMercer015: they're online nearly 24/7, just not sitting here neceesarily waiting for you ;) |
| 21:00.49 | brlcad | what's you're real question? |
| 21:00.52 | starseeker | GrantMercer015: most of us read backlogs too |
| 21:01.26 | GrantMercer015 | wanted to discuss a bit about my task for removing the system calls in tables.c |
| 21:01.42 | ``Erik | almost looked like a reference to jay and silent bob strike back... cli commander? |
| 21:06.29 | brlcad | GrantMercer015: and... what's your question? :) |
| 21:07.22 | *** join/#brlcad maths22 (ada798d1@gateway/web/freenode/ip.173.167.152.209) | |
| 21:07.53 | maths22 | I am participating in the google code-in, and I need a login to update the Gallery part of the website |
| 21:07.54 | caen23 | brlcad: i've now updated my work on the task to include subdirectories. how should i go about cmake? simply update the one insinde man1/en with lines such as asc/g2asc.xml, or should i create a cmake file for each subdir? |
| 21:08.13 | brlcad | maths22: i'm your guy, will talk in PM |
| 21:09.40 | starseeker | caen23: traditionally, what we've been doing in cases like this where existing man pages are converted to Docbook is to move the xml page to the appropriate docbook subdirectory, and add it there |
| 21:10.15 | starseeker | caen23: so for the converters, the appropriate directory would be system/man1/en |
| 21:10.21 | GrantMercer015 | Well i need to replace a system call with a quicksort function, but im confused on what exactly the function bu_vls_addr(register const struct bu_vls *vp) actually returns. It returns a char* but is that the actual data? |
| 21:10.43 | GrantMercer015 | All of this takes place in stc/libged/tables.c lines 428-442 |
| 21:11.13 | caen23 | starseeker: that's what i did, :D but now, i have system/man1/en/asc/g2asc.xml :D so a new subdir. does it require a new cmake file, or do i simply update the one inside system/man1/en to include "asc/g2asc.xml"? |
| 21:11.19 | starseeker | GrantMercer015: bu_vls_addr is what you use when you need to interact with a libbn variable length string the same way you would a char* |
| 21:12.03 | starseeker | caen23: since it's a man page, you don't need the subdirectory - the conv subdirectories are mainly to organize source code files |
| 21:12.12 | brlcad | GrantMercer015: so a "struct bu_vls *" is a class that provides a variable length string (similar to a std::string in c++) |
| 21:12.21 | caen23 | starseeker: thanks! |
| 21:12.41 | brlcad | bu_vls_addr() gives you an actual char * so you can treat that struct just like you would a regular C string |
| 21:12.46 | starseeker | s/libbn variable length string/libbu variable length string |
| 21:12.54 | brlcad | oops, yeah, |
| 21:13.49 | GrantMercer015 | so i would sort the char* member of the struct with a quicksort as opposed to passing it into system? |
| 21:14.07 | brlcad | GrantMercer015: you could write a program like this and it'd do what you think it should: |
| 21:14.39 | brlcad | struct bu_vls v = BU_VLS_INIT_ZERO; bu_vls_printf(&v, "hello world\n"); printf("%", bu_vls_addr(&v)); |
| 21:15.07 | brlcad | does quicksort take a char* |
| 21:15.21 | GrantMercer015 | yes, ill grab the linux function offline |
| 21:15.47 | brlcad | rather, qsort |
| 21:15.53 | GrantMercer015 | void qsort(void *base, size_t nmemb, size_t size, |
| 21:15.53 | GrantMercer015 | int(*compar)(const void *, const void *)); |
| 21:16.15 | GrantMercer015 | actually, it seems to take null pointers |
| 21:16.17 | brlcad | void * != char * |
| 21:16.39 | brlcad | it's wanting an array of things in memory that are equal 'size' |
| 21:17.21 | GrantMercer015 | so a null pointer to the size of the vls data? |
| 21:17.41 | brlcad | that makes no sense :) |
| 21:17.54 | brlcad | what's a null pointer going to do for you? |
| 21:17.56 | GrantMercer015 | ok, let me take a step back |
| 21:19.29 | GrantMercer015 | the qsort function just confuses me, sorry. It's boggling my mind |
| 21:19.51 | brlcad | tons of tutorials out on the web, I suggest writing a little test program to understand it |
| 21:20.00 | caen23 | why not use stl::sort? |
| 21:20.25 | brlcad | I think we might even have a libbu function that'll turn that char* into an array |
| 21:20.31 | brlcad | caen23: because it's not a c++ file |
| 21:20.37 | caen23 | oh, ok |
| 21:22.32 | GrantMercer015 | might seem dumb, but isnt a char* an array of chars already? |
| 21:23.59 | brlcad | GrantMercer015: yes, but you need an array of strings |
| 21:24.00 | starseeker | GrantMercer015: http://www.anyexample.com/programming/c/qsort__sorting_array_of_strings__integers_and_structs.xml |
| 21:24.13 | brlcad | since they all must be the same "size", you're going to have an array of pointers |
| 21:24.15 | starseeker | GrantMercer015: that may help a bit - example of qsort |
| 21:24.25 | GrantMercer015 | thanks starseeker! |
| 21:24.27 | caen23 | brlcad: i've uploaded what i hope to be the final version of my patch |
| 21:24.37 | GrantMercer015 | im starting to understand the qsort logic a little more |
| 21:26.06 | GrantMercer015 | what would i use as my compar function? does qsort give a default arg for that parameter or would i create my own comparison function |
| 21:27.42 | GrantMercer015 | nevermind, i asked to soon before i scrolled down starseekers site |
| 21:27.51 | GrantMercer015 | that sites explain a lot, appreciate it |
| 21:27.56 | *** part/#brlcad iwanmanew (~iwanmanew@94.236.208.130) | |
| 21:29.57 | Notify | 03BRL-CAD:r_weiss * 53839 brlcad/trunk/src/libbn/plane.c: Updated test function "bn_distsq_pt3_lseg3_v2" in file "plane.c". Added test for when dot product is undefined. |
| 21:30.26 | GrantMercer015 | alright, almost done bugging you guys. I looked up the sort command for unix but i couldnt seem to find what the numbers +1 and -2 mean in the command "sort -n +1 -2 -o /tmp/ord_id " |
| 21:30.43 | GrantMercer015 | are these the elements to start and end at? |
| 21:32.23 | elf` | http://unixhelp.ed.ac.uk/CGI/man-cgi?sort unix sort command |
| 21:34.13 | starseeker | elf`: that doesn't seem to document the numerical arguments |
| 21:36.21 | starseeker | +1 means to skip the first field |
| 21:38.35 | *** join/#brlcad moinnadeem (409521fc@gateway/web/freenode/ip.64.149.33.252) | |
| 21:40.24 | starseeker | brlcad: is the "+1 -2" bit shell scripting? |
| 21:42.06 | caen23 | apparently, i cannot add asc/g2asc.xml inside CMakeLists because of the dir sep. so how should i proceede? simply ignore subdirs, or have separate CMakeLists inside each subdir? :) |
| 21:42.11 | GrantMercer015 | in the case of my task, would argv[1] be the file that needs to be sorted? |
| 21:42.53 | starseeker | caen23: I thought you were moving g2asc.xml to doc/docbook/system/man1/en ? |
| 21:43.33 | caen23 | starseeker: oh, i'm sorry. i'm dumb. now i understand what you meant :D i'll do that |
| 21:44.39 | starseeker | GrantMercer015: my sort foo isn't strong enough to be of great help here - what I'd suggest as a intermediate step until someone more knowledgable chimes in would be to look at the behavior of the command in MGED |
| 21:46.41 | *** join/#brlcad maths22 (~maths22@173-167-152-209-illinois.hfc.comcastbusiness.net) | |
| 21:47.45 | maths22 | brlcad: when you are back online, can we chat again? |
| 21:48.06 | n_reed | GrantMercer015: the "+1 -2" seems to be old syntax for specify sort keys |
| 21:48.25 | n_reed | see http://www.gnu.org/software/coreutils/manual/html_node/sort-invocation.html |
| 21:49.11 | GrantMercer015 | thanks, so the file pretty much says sort one line |
| 21:49.47 | elf` | here what sort does |
| 21:49.48 | elf` | http://paste.ubuntu.com/1392886/ |
| 21:50.20 | elf` | from what I could gather playing with it |
| 21:52.25 | GrantMercer015 | thanks elf, so from what it seems it sorts downwards |
| 21:52.57 | elf` | columns |
| 21:52.59 | elf` | yeah |
| 21:53.56 | Notify | 03BRL-CAD:carlmoore * 53840 (brlcad/trunk/src/conv/iges/iges.c brlcad/trunk/src/libged/rfarb.c and 4 others): fix spellings and insert some missing blanks |
| 21:55.48 | *** join/#brlcad minichrispy (b82f3004@gateway/web/freenode/ip.184.47.48.4) | |
| 21:56.18 | minichrispy | hey I was wondering if anyone could help me with compiling brlcad with xcode |
| 21:56.33 | starseeker | minichrispy: what's the trouble? |
| 21:58.13 | minichrispy | Well, I'm doing this for the google code-in thing and Im supposed to write a guide on how to go through the process. I kind of just downloaded some things and tried to do some combining. This is so far what I have with my guide, and I feel like there must be some erros because it wasnt exactly working: http://pastebin.com/cR89SLy3 |
| 22:00.41 | starseeker | minichrispy: could you repost that on the mozilla pastebin? |
| 22:00.51 | starseeker | http://pastebin.mozilla.org/ |
| 22:01.13 | caen23 | minichrispy: do you need help with xcode specifically, or do you want to compile it on os x? :-? |
| 22:01.34 | minichrispy | http://pastebin.mozilla.org/1964174 |
| 22:01.42 | minichrispy | It has to be done through xcode |
| 22:02.33 | starseeker | minichrispy: what were the actual CMake errors? |
| 22:03.13 | minichrispy | It that the cmake cache txt file? |
| 22:03.44 | minichrispy | http://pastebin.mozilla.org/1964181 |
| 22:04.29 | starseeker | minichrispy: there's a file that records the process...one sec... |
| 22:06.05 | starseeker | CMakeFiles/CMakeOutput.log |
| 22:06.12 | starseeker | does that file exist? |
| 22:06.20 | moinnadeem | brlcad: check PMs. |
| 22:07.16 | minichrispy | i dont see it |
| 22:07.26 | starseeker | in the build directory? |
| 22:07.39 | minichrispy | nvm got it |
| 22:09.03 | minichrispy | http://pastebin.mozilla.org/1964182 |
| 22:11.45 | caen23 | regarding the "fix doc xml errors" gci task ( http://www.google-melange.com/gci/task/view/google/gci2012/7950213 ). as far as i can see, it seems to be related to the xml namespace (i.e. to docbook) more than it is related to brlcad. am i mistaken? |
| 22:12.09 | starseeker | caen23: yeah, they're docbook errors |
| 22:12.38 | caen23 | my solution would be to just create a namespace :-? is it valid? |
| 22:12.59 | starseeker | caen23: if you like, you could pick a particular error and we can discuss it... |
| 22:13.11 | starseeker | pastebin.mozilla.org ftw |
| 22:13.23 | caen23 | sure, let me just paste it to mozilla |
| 22:13.25 | caen23 | :) |
| 22:14.04 | caen23 | starseeker: http://pastebin.mozilla.org/1964193 |
| 22:14.11 | starseeker | minichrispy: is that all of it? |
| 22:14.32 | starseeker | minichrispy: try to find the subset of that file that contains the errors you were seeing... |
| 22:15.32 | minichrispy | oh, yea it didnt seem to copy all of it. ill look through it... sec |
| 22:17.10 | minichrispy | i got a good amount of things that were "not found" |
| 22:18.34 | starseeker | minichrispy: not found isn't necessarily bad - it's the errors we're after |
| 22:19.13 | starseeker | caen23: ok, so the first error relates to orgname. If you look at the docbook 5 documentation, you will see that address isn't a valid parent for orgname |
| 22:19.28 | starseeker | affiliation, however, *is* a valid parent |
| 22:19.51 | minichrispy | Ah, i see. i guess I just assumed those would cause errors. Im going to look through it again, but i dont know if i actually saw a real error |
| 22:20.12 | starseeker | and if you look at TEMPLATE.xml, you see that the address block is inside an affiliation block. So, in this case, we can move orgname outside of the address block but keep it within the affiliation block |
| 22:20.59 | minichrispy | Are there any specific words i could search for other than "error"? |
| 22:21.21 | starseeker | the next problem, the inlinegraphic failure, is different - that tag isn't part of db5 |
| 22:21.46 | starseeker | so, looking at the list of candidates, inlinemediaobject may be a suitable replacement to try |
| 22:21.56 | starseeker | minichrispy: error isn't present? |
| 22:22.09 | starseeker | minichrispy: what happens if you just re-run CMake? |
| 22:22.22 | caen23 | starseeker: thank you. so all i need to do is read the error, find the appropriate place in the code, and try to fix it using the documentation, right? |
| 22:23.13 | starseeker | right. this link should let you find out about various individual tags and their associated rules: http://docbook.org/tdg5/en/html/docbook.html |
| 22:23.21 | minichrispy | the word error is, but not exactly in the context of an actual error if that makes sense. I'll try re-running cmake |
| 22:24.29 | starseeker | caen23: feel free to ask if you aren't sure how to handle a particular situation |
| 22:24.38 | caen23 | ok :) |
| 22:24.56 | starseeker | that website will sometimes have helpful examples too |
| 22:25.27 | starseeker | inlinemediaobject, for example, has a worked example at the bottom: http://docbook.org/tdg5/en/html/inlinemediaobject.html |
| 22:25.47 | minichrispy | when I select the path to the source, this source I downloaded is the correct one, right? http://sourceforge.net/projects/brlcad/files/BRL-CAD%20Source/ (ver. 7.22.0) |
| 22:26.02 | starseeker | that should be fine |
| 22:26.16 | starseeker | (for instructions for non-developers) |
| 22:28.45 | starseeker | caen23: the inlinemediaobject example in turn uses imageobject, which also has an example which in turn uses imagedata (which has slots similar to those in inlinegraphic) |
| 22:30.03 | starseeker | caen23: that's probably one of the more elaborate rewriting instances - they'll vary between that and the orgname example in difficulty |
| 22:32.28 | caen23 | starseeker: thanks. i'll see what i can do. i hope to find an optimal way of dealing with this complexity after i improve a few files |
| 22:33.50 | caen23 | starseeker: one more question: when a file fails validation, do i receive all the errors for that file? so for example, if i make these three changes to template.xml, is the file completed? |
| 22:34.33 | starseeker | yes, provided new errors aren't introduced |
| 22:35.19 | starseeker | well, I think so at least... |
| 22:35.22 | starseeker | hmm |
| 22:36.02 | starseeker | caen23: some things might be automatible - remember though, that some of them may require judging context |
| 22:36.44 | caen23 | alright |
| 22:41.20 | minichrispy | ok i just went through the cmake process again. i dont see anything that really stands out in terms of the console, im just not exactly sure what im supposed to do from here (im basically guessing at this point) |
| 22:43.04 | starseeker | caen23: this table is handy for converting ulink entries: http://www.sagehill.net/docbookxsl/Db5Tools.html#Db5LinkExamples |
| 22:59.00 | caen23 | starseekr: hmm… if i try to replace the <ulink>s in template.xml with <link xlink:href>, i get a namespace error |
| 23:03.49 | starseeker | caen23: give me a second - I'm fixing up TEMPLATE.xml as a worked example of a 'before and after' conversion, but I'm getting a problem with the build |
| 23:13.03 | *** join/#brlcad bhlegm (~bhlegm@d108-180-27-19.bchsia.telus.net) | |
| 23:14.18 | Notify | 03BRL-CAD:starseeker * 53841 brlcad/trunk/misc/CMake/DocBook.cmake: Grr - ordering issues with script generation in DocBook targets. |
| 23:20.19 | brlcad | caen23: we haven't fixed errors in a while so keep track of how much time it takes you on the first file and then again on the second file (separately) |
| 23:20.37 | brlcad | and we can see how many files should be identified per task, whether it's 1, 5, 10, 50, 1000, etc |
| 23:22.37 | Notify | 03BRL-CAD:starseeker * 53842 brlcad/trunk/doc/docbook/articles/en/TEMPLATE.xml: Make TEMPLATE.xml pass DocBook5 xmllint validation. |
| 23:22.44 | caen23 | brlcad: hmm… i understand what you're trying to achieve, but i think it varies a lot. for example, i have no previous knowledge of xml, and i believe this is a huge impediment :-? |
| 23:22.59 | brlcad | that's why you're timing two files |
| 23:23.00 | starseeker | caen23: OK. Update your BRL_CAD trunk to r53842 |
| 23:23.11 | brlcad | presumably by the second file you'll have overcome most of the learning curve |
| 23:24.32 | caen23 | starseeker: there is a conflict. which do i choose? mc or tc? |
| 23:24.59 | starseeker | ah - you were working on TEMPLATE.xml? |
| 23:25.04 | caen23 | yes |
| 23:25.17 | starseeker | whoops |
| 23:25.27 | starseeker | go with mc |
| 23:25.47 | caen23 | ok, i am now on 53842 |
| 23:26.01 | starseeker | what does svn diff tell you on TEMPLATE.xml? |
| 23:26.32 | starseeker | minichrispy: what are the build failures? |
| 23:27.11 | starseeker | caen23: to see the changes I made, take a look at http://brlcad.svn.sourceforge.net/viewvc/brlcad/brlcad/trunk/doc/docbook/articles/en/TEMPLATE.xml?r1=52571&r2=53842 |
| 23:27.35 | starseeker | they may not be ideal for formatting, but the important part is they pass validation |
| 23:27.43 | caen23 | starseeker: hmm… <simpara> was removed, and i'm using <para> instead inside <listitem> |
| 23:27.50 | brlcad | caen23: how far along are you? |
| 23:28.07 | brlcad | and was your previous task reviewed? |
| 23:28.09 | starseeker | caen23: ok, that's not a bit deal |
| 23:28.13 | starseeker | big deal rather |
| 23:28.39 | caen23 | brlcad: i'm still trying to fix the first file. my previous task is yet to be reviewed |
| 23:28.54 | caen23 | starseeker: i'm not looking over the diff |
| 23:29.24 | brlcad | caen23: first file being which? |
| 23:29.36 | caen23 | starseeker: i see you've added an xmlns |
| 23:29.51 | starseeker | right - that activates the xlink syntax |
| 23:29.52 | caen23 | brlcad: doc/docbook/articles/en/TEMPLATE.xml |
| 23:30.37 | brlcad | someone already claimed the task we created for doc/docbook/articles, but we'll see if they follow through |
| 23:30.41 | brlcad | I'll tell them to skip that file |
| 23:30.54 | starseeker | winces |
| 23:31.19 | brlcad | it's a feeding frenzy ;) |
| 23:31.20 | starseeker | caen23: well, it's a good example anyway |
| 23:31.32 | caen23 | brlcad: yeah, i noticed. the more the merrier, i guess :) |
| 23:31.34 | brlcad | caen23: we'll put TEMPLATE.xml along with the next set |
| 23:31.47 | brlcad | so not lost effort, just means we need to create another task already ;) |
| 23:31.51 | *** join/#brlcad maths22 (~maths22@173-167-152-209-illinois.hfc.comcastbusiness.net) | |
| 23:32.01 | brlcad | which set do you want to tackle after template? |
| 23:32.30 | caen23 | preferably finish en/ ? |
| 23:32.34 | starseeker | brlcad: sorry about that - figured TEMPLATE.xml would be a good "before and after" example... |
| 23:33.23 | starseeker | caen23: somebody else called the articles subdirectory - there's lessons, man1, mann (lots of files but probably very repetetive...) |
| 23:34.32 | caen23 | ooh, ok. i'll try to go for man1 |
| 23:36.23 | brlcad | caen23: yeah, there are LOTS of docs, no worries there |
| 23:36.48 | brlcad | counted 800+ errors, only about 10 of them were in the doc/docbook/articles hierarchy |
| 23:38.04 | *** join/#brlcad GrantMercer015 (~GrantMerc@ip72-193-246-126.lv.lv.cox.net) | |
| 23:40.32 | starseeker | caen23: do you have any other questions you'd like to ask on the DocBook stuff before tackling it? |
| 23:41.28 | caen23 | starseeker: nothing at the moment. but i should improve my knowledge of xml, and soon |
| 23:42.36 | starseeker | caen23: don't worry too much - I figured enough of it out doing exactly what you're doing now |
| 23:42.54 | starseeker | caen23: that hyperlinked DocBook website is the key |
| 23:44.03 | caen23 | so far i've been using it to correct the hierarchy mistakes |
| 23:44.31 | starseeker | and patterns will develop - for example, anywhere you see a hyperlink with ulink, you know now to add that xlink header and alter the ulink lines to the new format |
| 23:44.44 | caen23 | yep |
| 23:44.51 | *** join/#brlcad GrantMercer015 (GrantMerce@ip72-193-246-126.lv.lv.cox.net) | |
| 23:45.10 | GrantMercer015 | hello everyone, anyone available to assist me? |
| 23:45.31 | starseeker | caen23: alrightie - I'll be back online in a few hours |
| 23:45.58 | starseeker | GrantMercer015: what did you need? |
| 23:45.59 | caen23 | starseeker: alright. thanks for your help so far |
| 23:46.32 | GrantMercer015 | i could use a little help with the solution im preparing for my task, it has some holes in it |
| 23:46.33 | starseeker | caen23: my pleasure - thank you for working on correcting these errors! |
| 23:46.49 | starseeker | GrantMercer015: letsee - you're doing the qsort conversion? |
| 23:47.29 | *** join/#brlcad adam_ (~adam@ppp-70-130-38-225.dsl.wchtks.swbell.net) | |
| 23:47.40 | adam_ | hi everyone! where do you keep the code? |
| 23:49.15 | GrantMercer015 | i need to remove calls to system in src/libged/tables.c that sort a file, and im thinking of using qsort with argv[1] as such: qsort(argv + 1, argc - 1, sizeof(argc),Sort) until i noticed that argv is const in my file |
| 23:49.18 | starseeker | adam_: http://brlcad.org/wiki/Compiling |
| 23:49.57 | GrantMercer015 | which sets me back, as i need someway to sort argv[1] |
| 23:50.27 | starseeker | GrantMercer015: then you need to make a local copy of argv and work on that, if you need to change it |
| 23:50.56 | GrantMercer015 | that sounds easy enough, what typename would i use to create a copy of argv[1]? |
| 23:51.44 | adam_ | starseeker, thanks |
| 23:51.54 | GrantMercer015 | would it by chance by a string into a filename? |
| 23:51.58 | GrantMercer015 | be a* |
| 23:52.11 | starseeker | GrantMercer015: um. If you take a look at src/mged/utility1.c you can see examples of copying argv |
| 23:52.32 | starseeker | f_edmater in particular is probably relevant here |
| 23:52.56 | starseeker | the difference in this case is you wouldn't be consting your copy |
| 23:53.27 | GrantMercer015 | ill take a look into that, thanks! just one more thing boggling my head(hopefully the last!) |
| 23:54.11 | GrantMercer015 | the sort command specifies specific fields to be sorted (--key=2,2), how would i re-create that in a qsort function? |
| 23:55.06 | starseeker | remember the example functions from that qsort example? |
| 23:55.28 | GrantMercer015 | yea |
| 23:55.51 | starseeker | your cstring_cmp function will have to do something different than just an all up strcmp |
| 23:57.17 | GrantMercer015 | ohh, so ill modify the comparison function to sort the array according to field two instead of straight up str1 < str2? |
| 23:58.14 | starseeker | depending on the sort criteria, you may be comparing portions of the strings, for example |
| 23:58.21 | starseeker | yeah, basically |
| 23:58.42 | starseeker | C has a number of functions which may be helpful: http://faculty.edcc.edu/paul.bladek/c_string_functions.htm |
| 23:59.06 | starseeker | http://en.wikipedia.org/wiki/C_string_handling too |
| 23:59.52 | starseeker | strncmp, starting from an offset, is one fairly likely possibility... |