IRC log for #brlcad on 20150408

00:04.56 maths22 Take a look at my latest ctest run http://brlcad.org/CDash/viewBuildError.php?onlydeltap&buildid=44
00:19.17 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
01:26.31 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
01:58.33 Notify 03BRL-CAD:starseeker * 64617 brlcad/trunk/src/libgcv/screened_poisson.cpp: More setup for screened poisson testing. This might actually have some potential when it comes to level-of-detail rendering - a low-point-count fast raytrace that is wrapped in a mesh, stepping up the number of points for more detail...
03:54.47 *** join/#brlcad pujani (~pujani@202.164.53.117)
04:07.35 *** join/#brlcad sofat (~androirc@106.192.152.52)
04:09.18 *** join/#brlcad sofat (~androirc@106.192.152.52)
05:19.05 *** join/#brlcad sofat (~androirc@106.192.152.52)
06:08.58 *** join/#brlcad hiteshsofat (~androirc@106.192.152.52)
06:10.15 *** join/#brlcad dracarys983 (dracarys98@nat/iiit/x-bbmqbnibtotcmixx)
06:30.25 *** join/#brlcad teepee (~teepee@unaffiliated/teepee)
06:42.20 *** join/#brlcad luca79 (~luca@host200-107-dynamic.15-87-r.retail.telecomitalia.it)
07:34.37 *** join/#brlcad merzo (~merzo@user-94-45-58-141.skif.com.ua)
07:38.05 *** join/#brlcad luca79 (~luca@host200-107-dynamic.15-87-r.retail.telecomitalia.it)
08:02.01 *** join/#brlcad sofat (~androirc@106.192.152.52)
08:05.10 *** join/#brlcad teepee-- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
08:09.10 *** join/#brlcad alisha (~quassel@115.184.117.219)
08:15.01 *** join/#brlcad merzo (~merzo@user-94-45-58-141.skif.com.ua)
08:19.36 *** join/#brlcad hiteshsofat (~androirc@106.192.156.26)
08:26.09 *** join/#brlcad andrei_il (~andrei@109.100.128.78)
08:43.32 *** join/#brlcad merzo (~merzo@user-94-45-58-141.skif.com.ua)
08:45.20 *** join/#brlcad sofat (~androirc@183.87.15.45)
09:01.19 *** join/#brlcad alisha (~quassel@115.184.117.219)
09:02.04 *** join/#brlcad ejno_ (~ejno@unaffiliated/kazaik)
09:35.31 *** join/#brlcad sofat (~androirc@223.225.249.12)
10:07.17 *** join/#brlcad sofat (~androirc@223.225.249.12)
10:19.23 *** join/#brlcad sofat (~androirc@223.225.249.12)
11:04.30 *** join/#brlcad alisha (~quassel@115.184.117.219)
11:28.06 *** join/#brlcad unicodesnowman (~unicodesn@wikipedia/unicodesnowman)
11:32.14 *** join/#brlcad alisha (~quassel@101.60.239.155)
12:06.54 *** join/#brlcad sofat (~androirc@223.225.249.12)
12:10.25 *** join/#brlcad luca79 (~luca@host151-108-dynamic.15-87-r.retail.telecomitalia.it)
12:11.26 *** join/#brlcad andrei_il (~andrei@109.100.128.78)
12:18.34 *** join/#brlcad unicodesnowman (~unicodesn@167.160.164.140)
12:39.53 *** join/#brlcad Stragus (~alexis@modemcable090.29-19-135.mc.videotron.ca)
12:48.47 *** join/#brlcad sofat (~androirc@183.87.15.45)
13:39.36 *** join/#brlcad FreezingCold (~FreezingC@135.0.41.14)
13:47.38 *** join/#brlcad sofat (~androirc@183.87.15.45)
13:52.17 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
14:26.10 Notify 03BRL-CAD:carlmoore * 64618 brlcad/trunk/src/libgcv/screened_poisson.cpp: remove trailing blanks/tabs
14:43.05 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
15:04.03 Notify 03BRL-CAD:starseeker * 64619 brlcad/trunk/src/libgcv/screened_poisson.cpp: Make a bit more progress towards running the mesh build
15:12.49 *** join/#brlcad dracarys983 (dracarys98@nat/iiit/x-bwmcpykdyaqeaqxr)
16:12.42 *** join/#brlcad pujani (~pujani@202.164.45.204)
16:17.08 *** join/#brlcad pujani (~pujani@202.164.45.204)
16:23.21 *** join/#brlcad pujani (~pujani@202.164.45.204)
16:23.22 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
16:26.05 *** join/#brlcad pujani (~pujani@202.164.45.204)
16:29.37 *** join/#brlcad sofat (~androirc@223.225.218.165)
17:03.36 *** join/#brlcad pujani (~pujani@202.164.45.212)
17:13.50 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
17:29.58 *** join/#brlcad unicodesnowman (~unicodesn@2602:ffea:1001:1fa::97d2)
17:30.02 *** join/#brlcad sofat (~androirc@223.225.218.165)
17:30.23 *** join/#brlcad alisha (~quassel@101.60.158.242)
17:30.55 *** part/#brlcad unicodesnowman (~unicodesn@2602:ffea:1001:1fa::97d2)
17:46.09 *** join/#brlcad alisha_ (~quassel@101.60.158.242)
17:50.11 *** join/#brlcad luca79 (~luca@adsl-ull-249-43.44-151.net24.it)
18:08.44 *** join/#brlcad sofat (~androirc@223.225.218.165)
18:09.06 sofat starseeker, hello
18:11.50 sofat I am working on mged html tutorial. I am converting these document into xml. I want ask question regarding this. There i found one folder animation. In this folder i see more files that files are part of this tutorial or its different from them ?
18:12.40 sofat It is articles or book ?
18:19.59 sofat brlcad, you know about this?
18:22.52 *** join/#brlcad teepee_ (~teepee@unaffiliated/teepee)
18:28.35 *** join/#brlcad pujani (~pujani@202.164.45.204)
18:33.37 *** join/#brlcad dracarys983 (dracarys98@nat/iiit/x-polzkqkkestrjqce)
18:35.22 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
18:48.23 Notify 03BRL-CAD:starseeker * 64620 (brlcad/trunk/src/libgcv/CMakeLists.txt brlcad/trunk/src/libgcv/gcv_private.h and 2 others): First successful generation of a bot primitive using screened poisson reconstruction. Quite a bit more to do here, but this establishes a proof-of-concept.
18:50.06 *** join/#brlcad albertcoder (caa42dcc@gateway/web/freenode/ip.202.164.45.204)
18:56.43 *** join/#brlcad pujani (~pujani@202.164.45.212)
18:57.48 pujani hi albertcoder,
19:07.05 *** join/#brlcad pujani (~pujani@1.39.34.227)
19:07.37 albertcoder Hi pujani I have accepted your pull request but I think you should work on some serious improvement like improving import / export (XML).
19:09.52 pujani Okay What can I do? Please tell me about
19:13.39 albertcoder Alright. I have been working on improving export of mysql data in JSON and CSV and my recent commit at github should reflect the changes I did. Did you do "git pull"?
19:14.28 pujani Yes
19:16.30 albertcoder Alright. Now see how the code of export works and then try to improve XML export. I guess it has something to improve. Just try export XML and see what it does.
19:17.47 pujani Okay I will try and report
19:19.37 pujani One more thing, images in MenuBar are not visible and source path is set in skins folder. Why is that?
19:20.27 dracarys983 brlcad, I didn't find anything that supports surface normals in DXF export for BoTs. Should I just throw an error then or should I output the normals as comments after each 3DFACE if the option is set?
19:20.42 albertcoder Yeah the path is to be changed there. I set the path in skins folder earlier because I think it is a standard way of doing.
19:22.27 pujani But I think I think images should be in extension folder
19:23.14 albertcoder yes pujani you can do that. It is not a big issue.
19:26.52 pujani albertcoder, Okay I will start on work of XML
19:27.22 albertcoder pujani: (y)
20:10.55 *** join/#brlcad ``Erik (~erik@pool-96-244-200-66.bltmmd.fios.verizon.net)
20:10.58 brlcad dracarys983: stuffing the data into a comment is creative, but I think it'd be ultimately self-defeating
20:11.26 brlcad users would have requested normals and then been surprised when that dxf file imported into something else didn't have normals
20:12.12 brlcad an error is sufficient
20:12.58 dracarys983 brlcad, Yes, that would be an issue. Okay, error it is then :)
20:13.04 brlcad if there were some means to store the normal data in dxf entities, even in raw form, that would something to consider, but don't think we need to dwell on this point too much
20:13.21 brlcad it's more important to have normal support for all the formats that support normals first
20:13.29 maths22 brlcad: did you see the errors I linked to last night?
20:14.30 dracarys983 brlcad, Yeah, we have surface normal support for OBJ, STL and SAT at this point for BoTs. :D
20:15.15 dracarys983 In OBJ we need to request for the normals. STL and SAT output them by default.
20:17.55 brlcad maths22: huh, those are news to me -- looks like clang bugs
20:18.09 brlcad https://llvm.org/bugs/show_bug.cgi?id=17788
20:19.07 brlcad dracarys983: that inconsistency is bothersome
20:19.58 brlcad from an interface perspective, it should output them by default if the model has them calculated, and there should be options for turning them on or off
20:20.12 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
20:20.58 brlcad maths22: huh, apparently it's not a new issue either: http://webcache.googleusercontent.com/search?q=cache:vs1kzTKBXnEJ:lists.freebsd.org/pipermail/svn-src-head/2013-October/052722.html+&cd=3&hl=en&ct=clnk&gl=us
20:21.05 maths22 is the default gcc clang on bz now
20:21.41 maths22 I should say cc
20:21.48 ``Erik default is clang, you can always do "cc -v"
20:22.06 maths22 when did that change?
20:22.14 ``Erik with fbsd 10
20:22.30 ``Erik bz went from 9 to 10 43 days ago
20:23.29 ``Erik gcc is still in the base system (/usr/bin/gcc), um, -DCMAKE_C_COMPILER=/usr/bin/gcc -DCMAKE_CXX_COMPILER=/usr/bin/g++ I think?
20:23.43 *** join/#brlcad pujani (~pujani@202.164.45.212)
20:23.48 dracarys983 brlcad, I guess in STL and SAT we can't turn it OFF. We can turn normals ON/OFF in OBJ.
20:24.38 dracarys983 I can make normals default in OBJ as well, to get consistency in normals being outputted by default, maybe? :)
20:26.10 brlcad maths22: it's definitely possible to get a clang bulid
20:27.10 brlcad can either disable strict, add -Wno-c11-extensions as a build flag, or read up on the math.h header to see if there's some way to get at isinf() without the next _Generic() construct getting involved
20:27.16 brlcad s/next/new/
20:28.28 brlcad dracarys983: not true, stl at least lets you set normal to 0,0,0
20:28.39 brlcad implying that normals have not yet been calculated
20:29.12 brlcad for obj, it's just a matter of not outputting vn lines iirc
20:29.34 maths22 ``Erik: thanks
20:30.08 brlcad ``Erik: you see the 8088 on HN?
20:30.28 dracarys983 brlcad, Ah-ha! Okay, that's good news :)
20:32.19 dracarys983 brlcad, I'll check up on SAT then whether it's possible to turn normals off for it.
20:37.01 maths22 I'm now re-running with gcc
20:37.12 maths22 I'll think about what was going on there later
20:43.02 Stragus FreeBSD followed OSX down the path of clang decadence and decay? Tssk
20:43.50 brlcad dracarys983: you'll also want to look more closely at the obj and other file formats -- some like obj have a variety of normal mechanisms
20:44.45 brlcad Stragus: most of the BSDs embraced the BSD-licensed compiler. shocking? :)
20:45.06 dracarys983 brlcad, You mean I add all those formats that support normals (maybe like OBJ)? :)
20:45.21 brlcad they want a bsd stack as much as debian folks want a gpl stack
20:46.16 maths22 brlcad: The issues seems to be fully on clang's end
20:46.31 brlcad dracarys983: I mean if you're going to be mucking with formats, you should have a specific goal in mind that is applied consistently across all formats so that the usability/interface is the top priority, and that any changes are made while consulting the format specifications
20:47.22 brlcad maths22: sure, but they are minor issues that we are exagerrating in our build
20:47.35 brlcad by turning warnings into errors -- that's basically a new warning
20:47.59 brlcad isinf() and friends are usually implemented as macros
20:48.54 maths22 brlcad: I wonder if -std=c99 would fix it
20:48.59 brlcad someone wrapped those macros with the _Generic() construct, which is a new c11 feature that lets the compiler bind calls to things like isinf() to different functions based on the type of macro argument
20:49.45 maths22 Look at the condition in math.h
20:49.53 brlcad highly powerful, but the problem is that 1) it's reporting a warning on a system header that it probably shouldn't and 2) that c11 construct is being used in C compilation mode which it probably shouldn't
20:50.36 maths22 by default, both gcc and llvm build according to the latest c standard
20:50.54 maths22 see http://clang.llvm.org/compatibility.html#inline
20:51.50 dracarys983 brlcad: Okay. :)
20:52.42 brlcad maths22: this isn't an inline issue
20:52.59 brlcad but is related to the fact that c11 is now default
20:53.41 maths22 Right
20:53.51 maths22 Inline just had the quote
20:54.56 brlcad it's mostly #1 -- it shouldn't be reporting issues in system headers
20:55.13 maths22 Correct
20:55.14 brlcad if I request -std=c89, I shouldn't get system header warnings about using c11 constructs
20:55.46 brlcad the system header should deal with the appropriate switching or the compiler should be aware it's parsing a system header (it usually does)
20:56.29 maths22 Have you looked at the condition in math.h?
20:56.50 brlcad clang is probably blaming the libc/libm guys and the libc/libm guys are probably blaming clang (as both are deficient), or it's simply not gotten much attention yet
20:56.55 brlcad yes
20:57.58 maths22 The issue seems to be that is a macro, so when llvm gets to looking at it, it does not realize it came from a system header
20:58.03 brlcad if STDC is C11 and we're clang ... or has_extension(c_generic_selections) .. so the latter is succeeding
20:59.30 brlcad nah, llvm knows when it's parsing constructs and where they came from -- very good at managing all that
21:00.25 brlcad it's just naive simple logic coupled with a generic warning check
21:00.43 *** join/#brlcad pujani (~pujani@202.164.45.204)
21:01.01 *** join/#brlcad FreezingCold (~FreezingC@135.0.41.14)
21:01.52 brlcad the header literally and simply says "if I have _Generic(), use it" ..
21:01.58 brlcad the compiler says "if I see _Generic, warn about it (-Wc11-extensions) because I was told to compile in c89 mode"
21:02.18 brlcad brl-cad is saying compile in c89 mode and treat warnings as errors
21:02.50 brlcad trifecta resulting in the build error -- fixing any of the three makes the problem go away
21:02.52 maths22 See http://clang.llvm.org/docs/LanguageExtensions.html#feature-checking-macros
21:03.05 maths22 I'm not sure why has_extension would pass
21:04.20 ``Erik brlcad: no... musta missed it, url? (8088 on hn)
21:04.36 maths22 brlcad: we need to pass the -pendantic-errors flag
21:04.38 maths22 that fixes it
21:05.08 brlcad ``Erik: http://trixter.oldskool.org/2015/04/07/8088-mph-we-break-all-your-emulators/
21:05.44 brlcad that would have blown my mind, seeing that many colors on a cga monitor
21:06.37 ``Erik all hail dithering... lots of c64 games looked like that and the video capabilities were ... strange :D
21:07.40 ``Erik cga was 4 color? huh, thought it was 16
21:08.37 brlcad he's doing more than dithering iirc
21:10.32 teepee hey, I saw that on live stream, the sound was a bit funny :)
21:10.32 brlcad 4 colors at 320x200, 16 colors at 160x100
21:10.42 brlcad hardware color-mapping
21:12.01 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
21:12.08 Stragus Was that hardware where you could change the palette between scanlines?
21:13.04 teepee you can on lots of hardware, for atari 2600 that even was the way to generate the image ;)
21:14.03 teepee checks if the video from the opengl shader live coding competition is already online...
21:18.06 brlcad from the same competition, in 64k https://www.youtube.com/watch?v=mEjTnstaMLM
21:20.07 teepee ahh, that one is awesome
21:20.38 teepee I did like the oldskool ones too :)
21:20.57 Stragus I must be too young to fully appreciate these old demos
21:21.30 Stragus I grew up with a 386 66mhz, writing true 32 bits protected code code with DJGPP
21:21.51 brlcad Stragus: that last link was not an old demo...
21:21.52 teepee yeah, I guess it helps having programmed on a system with like 16k or 64k ram
21:22.05 Stragus I was watching the 8088 one
21:22.13 brlcad ah
21:22.39 brlcad 8088 was in the PC category, I think 504 kb or something
21:23.06 brlcad in some ways the 64kb category winner is more impressive (and definitely more visually impressive)
21:23.11 teepee likeley "wild" where any system is allowed with one "normal" medium
21:23.24 brlcad all the artwork, music, and code are included in that 64kb limit
21:23.45 brlcad 64kb binary on disk, run it :)
21:24.13 Stragus I wrote a little game for a 64kb compo many years ago
21:24.18 Stragus 64kb is a ton of memory :p
21:25.03 brlcad for a game sure :)
21:25.03 teepee not so much for full-hd video with textures. I guess that's pretty much all precedural stuff
21:25.51 brlcad for a 5 minute video with custom scene animations (all displaying in real time) with custom artwork and music, much more impressive I think
21:26.12 teepee all video (hopefully soon) at https://www.youtube.com/user/RevisionParty/videos
21:27.00 teepee I found most impressive what some did in the 25 minutes live coding session
21:27.15 brlcad cool
21:27.35 teepee like shader toy, but 2 people live on stage with the huge splitscreen in the background
21:28.27 brlcad I would totally love to attend one of those if one were nearby
21:29.08 teepee seems to be mostly european locations http://www.demoparty.net/
21:29.33 teepee oh, one east and one west coast US
21:30.06 Stragus Here's something I had written for a 64kb x86 asm compo: http://www.rayforce.net/raptor000.png
21:30.21 Stragus Game was infinite and getting harder indefinitely, with bosses and such
21:30.25 teepee looks like los angels and boston
21:30.42 *** join/#brlcad andrei_il (~andrei@109.100.128.78)
21:31.28 brlcad Stragus: that looks fun :) like an old nintendo game
21:31.50 Stragus Yup that was the idea :)
21:31.54 brlcad teepee: yeah, boston .. huh
21:32.11 teepee still too far away?
21:32.50 brlcad 'bout 8 hours
21:32.57 teepee and LA?
21:33.09 brlcad 'bout 8 hours ;)
21:33.19 teepee haha, no demo party then
21:33.28 teepee well, none they list on demoparty.net that is
21:33.32 brlcad ones a car ride, the other a flight
21:33.54 Stragus And work won't reimburse that flight? :) Come on, it's a conference
21:34.04 teepee I guess I could reach all of the 40 EU parties in that time :P
21:34.06 brlcad would be cool to host a cademoparty
21:34.30 teepee and live shader coding would still almost fit
21:34.31 brlcad model something impressive in just a couple days
21:34.39 teepee the software is on github
21:34.55 teepee https://github.com/Gargaj/Bonzomatic
21:36.24 brlcad that's awesome
21:36.32 ``Erik cga was 4 color? huh, thought it was 16
21:36.38 ``Erik woops
21:37.11 brlcad glitch in the matrix
21:37.35 brlcad maths22: that's really odd that switching to pedantic errors fixed it!
21:37.52 Stragus blames clang
21:38.44 maths22 http://clang.llvm.org/docs/LanguageExtensions.html#feature-checking-macros
21:38.57 maths22 See the comment about pendantic-error
21:43.03 *** join/#brlcad andrei__ (050cdda5@gateway/web/freenode/ip.5.12.221.165)
21:45.57 brlcad maths22: yeah, that explains it .. interesting way to handle it
21:47.32 maths22 Its not necessarily a bad flag anyway
21:47.54 brlcad the fact that it fixes this specific case is still kind of an unintended consequence
21:47.55 maths22 what is this code from: src/libgcv/poissonrecon/?
21:48.00 maths22 That is true
21:48.08 maths22 about the unintended consequence
21:48.10 brlcad it's the fact that there's a mismatch with the header and its logic
21:48.34 maths22 the header should be using __has_feature can
21:48.42 maths22 I mean __has_feature
21:48.53 maths22 not __has_extension
21:50.05 brlcad yeah, there needs to be a README in that directory and authorship info
21:50.25 brlcad it's from http://www.cs.jhu.edu/~misha/Code/PoissonRecon/Version6.13/
21:51.07 maths22 Is it permissible to edit that code in any way?
21:51.07 brlcad bsd licensed code
21:51.20 maths22 http://www.cs.jhu.edu/~misha/Code/PoissonRecon/license.txt
21:51.24 maths22 Yep
21:52.00 brlcad can you drop that in the directory?
21:52.22 brlcad surprised starseeker would put that in there without proper docs, but he is working on that stuff now
21:52.59 Notify 03BRL-CAD:maths22 * 64621 brlcad/trunk/src/libgcv/poissonrecon/SparseMatrix.inl: added license file for possionrecon
21:53.00 maths22 Each file includes the license at the top
21:54.37 Notify 03BRL-CAD:maths22 * 64622 (brlcad/trunk/src/libgcv/poissonrecon/BSplineData.h brlcad/trunk/src/libgcv/poissonrecon/Factor.h and 2 others): added newlines to end of possionrecon source
22:00.46 brlcad huh, except the random one that I picked to check, Hash.h
22:01.08 brlcad I meant just adding a license.txt file there so it's explicit
22:01.28 maths22 I did
22:01.42 brlcad cool, thanks
22:02.30 brlcad how about a README citing the source?
22:02.38 *** join/#brlcad teepee (~teepee@unaffiliated/teepee)
22:03.17 maths22 Can do soon
22:03.30 starseeker brlcad: sorry 'bout that - license info is in the src files, forgot to add a separate file
22:03.53 brlcad starseeker: not in the one I randomly picked :)
22:04.03 starseeker figures
22:06.18 starseeker if this approach proves useful, maybe need to do some siginficant rework on the code - right now it needs OpenMP for parallelism, so we're compiling it in serial mode
22:07.35 starseeker not sure how turning on OpenMP would play with our own threading
22:09.29 brlcad ugh
22:11.59 brlcad for a self-contained gcv plugin, maybe
22:12.31 brlcad for a library (including libgcv or librt), I wouldn't be in favor
22:23.33 Notify 03BRL-CAD:maths22 * 64623 brlcad/trunk/src/libgcv/screened_poisson.cpp: removed unnecessary semicolons (pedantic error)
22:31.37 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
22:31.51 maths22 brlcad: Is it OK to commit the following? http://paste.ubuntu.com/10777205/
23:13.28 brlcad maths22: misc/CMake/BRLCAD_CompilerFlags.cmake
23:13.39 brlcad probably belongs there as that's where strictness is enabled
23:13.44 brlcad see pedantic
23:13.47 maths22 Ok.
23:13.52 brlcad probably good to just use both :)
23:14.18 maths22 I'll move it.
23:14.49 brlcad if you can, make sure it compiles with gcc too
23:15.03 brlcad that's going to be almost certain to fail something somewhere
23:20.03 brlcad darn I don't have part 21 handy
23:22.08 maths22 Of what?
23:22.21 maths22 Running with GCC now
23:45.46 brlcad part 21 of the step standard, stepcode bug reported but don't know what the spec says
23:48.33 maths22 http://brlcad.org/CDash/viewBuildError.php?buildid=46

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.