Mitesh (Nov 30 2017 at 15:35): Mitesh (Nov 30 2017 at 15:35):Sir I am having an issue.
I referred the compilation guides: -
http://brlcad.org/wiki/Compiling
http://svn.code.sf.net/p/brlcad/code/brlcad/trunk/INSTALL ----->(Quick Install)
I was successful till cmake part.
but my "make" command didn't work in both the cases.
I am getting the following error :-
"make: *** No targets specified and no makefile found. Stop."
i am running the commands in "VS015 x86 Native Tools Command Prompt".
can you please help me out.
thank you
Daniel Rossberg (Nov 30 2017 at 15:52):@Mitesh Are you working on Windows with Visual Studio?
Mitesh (Nov 30 2017 at 15:52):Yes sir.
Daniel Rossberg (Nov 30 2017 at 15:59):I don't know somebody with name Sir here but OK: There should be a solution file BRLCAD.sln somewhere. Open it with Visual Studio (double click on it) and build the INSTALL target (or project).
Mitesh (Nov 30 2017 at 16:12):I don't know somebody with name Sir here but OK: There should be a solution file BRLCAD.sln somewhere. Open it with Visual Studio (double click on it) and build the INSTALL target (or project).
Extremely sorry for addressing everyone as sir.
Thanks alot for replying. I will try the solution.
Daniel Rossberg (Nov 30 2017 at 16:17):Extremely sorry for addressing everyone as sir.
Thanks alot for replying. I will try the solution.
No need to be sorry. I only wanted to mention that the usage of "sir" is extremely unusual in the open source community. If you want to address somebody special use the @ as in @Mitesh . Otherwise, simply ask your question and I'm sure that somebody will try to help you.
Mitesh (Nov 30 2017 at 16:18):sure @Daniel Rossberg
Mitesh (Nov 30 2017 at 17:05):I have build the INSTALL project but got 10 errors:-
can you please help me out.
thank you
Daniel Rossberg (Nov 30 2017 at 17:24):These are CMake errors. What do you get when you run CMake alone?
Mitesh (Nov 30 2017 at 17:29):when i ran the command in the process earlier it was something like this:-
task4.png
Daniel Rossberg (Nov 30 2017 at 17:38):This doesn't look so bad.
Did you choose different directories for the source code and the build? The directory where you run cmake is the build directory where you can find the BRLCAD.sln afterwards. The first argument to cmake is the source code root directory.
Mitesh (Nov 30 2017 at 17:42):Yes @Daniel Rossberg i chose different directories for the source code and the build
Mitesh (Nov 30 2017 at 17:45):And I only downloaded zip file from:- https://sourceforge.net/projects/brlcad/files/BRL-CAD%20Source/7.26.4/
not the other files. is that ok?
Daniel Rossberg (Nov 30 2017 at 17:58):Using the files from the subversion repository would be better but the zip file should work too. For me it looks like the Visual Studio can't find CMake. When you started cmake for the first time you choosed your development environment, right?
Mitesh (Nov 30 2017 at 18:04):@Daniel Rossberg i have gui of cmake....and when it was opened first time as far as i remember there was no option to choose anything, or maybe i don't remember
Daniel Rossberg (Nov 30 2017 at 18:15):I'm using the CMake GUI too. When you press "Configure" there for the first time a dialog should pop up which lets you choose between the supported compilers.
Mitesh (Dec 01 2017 at 02:44):I'm using the CMake GUI too. When you press "Configure" there for the first time a dialog should pop up which lets you choose between the supported compilers.
I remember that...
but I ran cmake command in vs15command line, and did not build it through gui.
Mitesh (Dec 01 2017 at 06:02):There were errors while building INSTALL
but when I opened "Debug folder>> Bin>> I was able to run Mged.exe". Is it done?
Sean (Dec 01 2017 at 06:05):if mged runs, then you're probably done -- screenshot?
Mitesh (Dec 01 2017 at 06:11):
Mitesh (Dec 01 2017 at 06:12):But I would like to know how to solve the cmake errors.
Sean (Dec 01 2017 at 06:23):that's definitely looking like a build of 7.26.4
Mitesh (Dec 01 2017 at 06:23):Thank You @Christopher Sean Morrison
Sean (Dec 01 2017 at 06:23):what cmake errors are you referring to?
Sean (Dec 01 2017 at 06:23):not seeing them in the earlier discussion
Mitesh (Dec 01 2017 at 06:32):
Sean (Dec 01 2017 at 06:35):those look like msvc errors, not cmake
Mitesh (Dec 01 2017 at 06:37):okk... but how to solve them?
I tried one way out i.e by changing the cmake and build directories to another drive.
but got the same result
Sean (Dec 01 2017 at 06:37):now did you run cmake?
Sean (Dec 01 2017 at 06:37):s/now/how/
Sean (Dec 01 2017 at 06:38):via gui or command-line?
Mitesh (Dec 01 2017 at 06:38):command line
Sean (Dec 01 2017 at 06:39):it shouldn't matter but try running it from the gui
Sean (Dec 01 2017 at 06:40):you need to specify where the sources are and where your build directory is
Sean (Dec 01 2017 at 06:40):(the first time)
Mitesh (Dec 01 2017 at 06:41):okk... Thank You @Christopher Sean Morrison
I will try it.
Sean (Dec 01 2017 at 06:43):no problem, that's why we're here
Mitesh (Dec 01 2017 at 06:43):Should I add anything in "option toolset to use"?
Sean (Dec 01 2017 at 06:44):er, picture please
Sean (Dec 01 2017 at 06:44):probably not, but I don't recall that option
Mitesh (Dec 01 2017 at 06:45):
Sean (Dec 01 2017 at 06:45):aha, no don't add anything
Sean (Dec 01 2017 at 06:45):what version of studio do you have?
Sean (Dec 01 2017 at 06:46):community edition?
Mitesh (Dec 01 2017 at 06:46):2015 community
Sean (Dec 01 2017 at 06:50):okay, then that looks fine -- did you run it?
Mitesh (Dec 01 2017 at 06:50):yes....its running
Sean (Dec 01 2017 at 06:50):yay!
Mitesh (Dec 01 2017 at 06:53):
Mitesh (Dec 01 2017 at 06:55):I got this:-Capture4.PNG
Mitesh (Dec 01 2017 at 06:56):should I press generate?
Sean (Dec 01 2017 at 06:56):yep, that looks all good so far
Sean (Dec 01 2017 at 06:58):for what it's worth, your compile more than likely completed 99% of the way
Sean (Dec 01 2017 at 06:58):mged is one of the last steps, so that's a good sign
Mitesh (Dec 01 2017 at 07:00):ok ...I have opened the project...and then after I will try to build the install project.
will update once its done.
thank you.
Sean (Dec 01 2017 at 07:00):OH, did you run the install project the first time too?
Mitesh (Dec 01 2017 at 07:01):yup
Sean (Dec 01 2017 at 07:01):that's the problem -- run the ALLBUILD target
Mitesh (Dec 01 2017 at 07:01):only ALLBUILD Target?
Sean (Dec 01 2017 at 07:01):yeah, that will be fine
Mitesh (Dec 01 2017 at 07:02):okk
Mitesh (Dec 01 2017 at 07:03):bingo....build succeeded
Mitesh (Dec 01 2017 at 07:04):
Sean (Dec 01 2017 at 07:04):I don't know the specifics on Windows, but I think the INSTALL target attempts to create a Windows installer, and you would need to have other tools installed for that
Mitesh (Dec 01 2017 at 07:05):ok
Mitesh (Dec 01 2017 at 07:06):whats the next step?
Sean (Dec 01 2017 at 07:06):have you done any of the tutorials yet?
Sean (Dec 01 2017 at 07:07):I suggest you try that next, just to get a little familiar with the tools, then pick a coding task and just give it a try :)
Mitesh (Dec 01 2017 at 07:08):ok sure @Christopher Sean Morrison
thank you so much.
can you please send me the link of tutorials?
Sean (Dec 01 2017 at 07:09):I don't have it handy, but you can find it on the site if you just search tutorial
Sean (Dec 01 2017 at 07:10):there's two, part one is the first 8 lessons, part 2 is the second 8 lessons
Sean (Dec 01 2017 at 07:10):you can do all of them or just the first one if your intention is to write code
Sean (Dec 01 2017 at 07:10):what kinds of tasks sound interesting to you?
Mitesh (Dec 01 2017 at 07:10):ok these tutoials
I already completed the first one
Mitesh (Dec 01 2017 at 07:11):I am on verge of completion of second
Sean (Dec 01 2017 at 07:11):when you say first one, do you mean lessons 1-8 or lesson 1 ?
Mitesh (Dec 01 2017 at 07:12):lesson 1-8
Sean (Dec 01 2017 at 07:12):ah, cool -- so got all your screenshots?
Mitesh (Dec 01 2017 at 07:13):yup sir....this task was already approved yesterday.
Mitesh (Dec 01 2017 at 07:13):i think it was approved you @Christopher Sean Morrison
Sean (Dec 01 2017 at 07:15):yup sir....this task was already approved yesterday.
I appreciate your respect, but please know that it's VERY unusual to refer to people as sir when working in open source
Sean (Dec 01 2017 at 07:15):very much not necessary, distracting even
Mitesh (Dec 01 2017 at 07:17):sorry i forgot about that
Sean (Dec 01 2017 at 07:17):no need to be sorry @Mitesh
Mitesh (Dec 01 2017 at 07:18):@Christopher Sean Morrison i have one doubt.
once the "all_build" file is built, i could only find these files in the debug>>bin folder:-
Capture6.PNG
where are the other files?
Sean (Dec 01 2017 at 07:22):hm, did you check up in the release dir?
Sean (Dec 01 2017 at 07:22):I didn't see if you compiled debug or release
Mitesh (Dec 01 2017 at 07:24):Its the same in release dir. No other files are there.
Sean (Dec 01 2017 at 07:25):huh, was bin the same folder where you found mged earlier?
Mitesh (Dec 01 2017 at 07:25):yes
Sean (Dec 01 2017 at 07:26):that is odd
Sean (Dec 01 2017 at 07:26):and it's the dir you fed to cmake
Sean (Dec 01 2017 at 07:26):what was the msvc output?
Mitesh (Dec 01 2017 at 07:26):there was no output
Sean (Dec 01 2017 at 07:27):there's always output :)
Sean (Dec 01 2017 at 07:27):it just might be hidden
Mitesh (Dec 01 2017 at 07:28):i just go this result....Capture5.PNG
Sean (Dec 01 2017 at 07:29):was msvc still open and running when you were running cmake?
Sean (Dec 01 2017 at 07:29):quit msvc and try re-opening the solution
Daniel Rossberg (Dec 01 2017 at 07:35):Hi @Mitesh , try the following: Start the CMake GUI, look for the CMAKE_INSTALL_PREFIX there, and set it to a directory where you have write access (e.g. H:/brlcad.install). Then build the INSTALL target. The issue with the ALLBUILD target is that it doesn't copy the files in the order of a BRL-CAD installation. This is what INSTALL does. The reason for the errors you experienced could have been that you hadn't write access to the install directory which is something in C:/Program Files/ by default.
Sean (Dec 01 2017 at 07:36):huh, learned something new ... and I've built with the windows solution many many times
Sean (Dec 01 2017 at 07:36):I don't recall ever having a problem with ALL_BUILD -- that's supposed to be the same target logic as make's default
Daniel Rossberg (Dec 01 2017 at 07:40):Some things work in the build directory (Mitesh could successfully start mged) others may not. On Windows INSTALL simply copies everything into the right order, similar to Linux.
Mitesh (Dec 01 2017 at 08:07):@Daniel Rossberg I tried your method....but built failed:-
Capture7.PNG
Sean (Dec 01 2017 at 08:16):mahdi, okay -- we're exhausting options. I suggest you install Subversion and download the latest BRL-CAD source code from a repository checkout
Mitesh (Dec 01 2017 at 08:17):quit msvc and try re-opening the solution
I also tried this but got the same result i.e built succeeded but only few files are present in bin folder
Sean (Dec 01 2017 at 08:17):and see if you still run into a problem -- but basically starting over, and trying in a different location with the latest source code, and we can try some things
Sean (Dec 01 2017 at 08:18):that said, I'm out for a bit -- have fun!
Daniel Rossberg (Dec 01 2017 at 08:21):That is really strange. The install script complains about a missing regex.dll but mged can be started.
Sean (Dec 01 2017 at 08:22):yeah, it's not adding up
Mitesh (Dec 01 2017 at 10:55):i tries with svn....got this errorCapture1.PNG
Daniel Rossberg (Dec 01 2017 at 11:27):I got the same error with today's svn checkout. The reason is that since yesterday the configuration requires a distinct C++ standard conformity but Visual Studio doesn't know about C++ standard conformity levels at all. I'll see what can be done there.
Mitesh (Dec 01 2017 at 11:55):ok @Daniel Rossberg
Daniel Rossberg (Dec 01 2017 at 13:35):@Mitesh You can now update from svn and give it another try.
Naseef (Dec 01 2017 at 15:58):He's probably Indian, They have tendency to use 'Sir' all the time.
Sean (Dec 01 2017 at 16:02):He's probably Indian, They have tendency to use 'Sir' all the time.
we have many Indian developers that have contributed to the project over the years, so we know, but also using that designation is not limited to them -- it's a cultural norm for respect. part of our job introducing students like yourself to open source is to present what the open source cultural norms are as well.
Sean (Dec 01 2017 at 16:03):with open source, it's incredibly rare to use titles and designations -- any title is usually earned through contributions, not through position or education, and they're rarely indicated during conversation
Naseef (Dec 01 2017 at 16:04):You mean 'everyone is treated same in open source'?
Sean (Dec 01 2017 at 16:05):no, not necessarily
Sean (Dec 01 2017 at 16:05):our community, for example, operates as a meritocracy
Sean (Dec 01 2017 at 16:05):the more you contribute and the more you do, the more you get a say in what gets done, what others can/cannot do
Daniel Rossberg (Dec 01 2017 at 16:24):@Mitesh With the latest svn revision I could successfully build the INSTALL target.
Mitesh (Dec 01 2017 at 17:32):ok @Daniel Rossberg i will try it after some time.
Mitesh (Dec 01 2017 at 17:35):Ya...i am from India:)
Mitesh (Dec 02 2017 at 04:43):@Daniel Rossberg Still facing the same problem when i built install in visual studio
Sean (Dec 02 2017 at 15:30):which problem is that?
Sean (Dec 02 2017 at 15:31):sounded like the issue was building a mixture of mingw and msvc when the whole stack should really be msvc only
Sean (Dec 02 2017 at 15:31):if you remain stuck, we can get you set up in a virtual machine where everything is preconfigured and more seamless
Mitesh (Dec 02 2017 at 16:55):@Sean its fine..got a good news.... I just tried again and it worked
Mitesh (Dec 02 2017 at 17:05):
Naseef (Dec 02 2017 at 17:13):Naseef : He's probably Indian, They have tendency to use 'Sir' all the time.
Well, I've met some Indians in IRC, Chats, and Forums and helped them to solve some small problems, Most of them called me 'Sir' at least once within the conversations.
Naseef (Dec 02 2017 at 17:17):I know it's irrelevant, But I had a little anxiety about saying it, So I wanted to clear it up
Naseef (Dec 02 2017 at 17:17)::wink:
Daniel Rossberg (Dec 02 2017 at 17:18):@Mitesh Congratulations! :tada:
Naseef (Dec 02 2017 at 17:20):Should I install BRL CAD on Windows or Linux?
Daniel Rossberg (Dec 02 2017 at 17:22):On whichever system you feel yourself more comfortable. Both should work.
Naseef (Dec 02 2017 at 22:00):I'd go with Linux then.
Naseef (Dec 03 2017 at 15:44):How much storage space I'll need? I have around 600 mb left on Ubuntu partition.
Daniel Rossberg (Dec 03 2017 at 16:30):My build directory has 2.3 GiB. I performed only a single standard build there. Therefore, I would say you are in trouble :(
Sean (Dec 03 2017 at 16:55):@Naseef a complete build+compile usually takes about 2GB-6GB depending on your OS/filesystem
Naseef (Dec 03 2017 at 17:14):Oh, Well. I have around 32 GB unallocated space left on my disk, unfortunately I already have 4 primary partitions.
Naseef (Dec 03 2017 at 17:18):Windows partition has around 30 GB left, But Windows is always in Unsafe State.
Naseef (Dec 03 2017 at 17:20):Files partition have 4.5 GB left, I think I can build it there.
Daniel Rossberg (Dec 03 2017 at 17:32):The souce code, the build, and the installation can be on different partitions.
Naseef (Dec 03 2017 at 17:33):So I can keep the source and build in Files partition and install it to Ubuntu partition?
Sean (Dec 03 2017 at 17:36):sure, and for most tasks, you don't even need to install (you can run from the build tree)
Daniel Rossberg (Dec 03 2017 at 17:37):Yep, set CMAKE_INSTALL_PREFIX to a directory in your Ubuntu partition, that's it.
Naseef (Dec 03 2017 at 17:37):That's good.
Naseef (Dec 03 2017 at 17:38):Is there a Git repo for BRL-CAD?
Naseef (Dec 03 2017 at 17:38):or Just SVN?
Sean (Dec 03 2017 at 17:38):the website and ogv tasks are in git, the main code is svn
Naseef (Dec 03 2017 at 17:56):Does BRL-CAD SVN have both MGED and Archer?
Sean (Dec 03 2017 at 17:57):yes
Sean (Dec 03 2017 at 17:57):and 398 other tools ;)
Naseef (Dec 03 2017 at 17:57)::cold_sweat:
Sean (Dec 03 2017 at 17:59):most are very small tools that do one specific thing
Naseef (Dec 03 2017 at 18:00)::relieved:
Sean (Dec 03 2017 at 18:02):there are only 4 larger gui tools: mged, archer, adrt, and rtwizard
Naseef (Dec 03 2017 at 20:22):Build takes pretty long as I expected.
Naseef (Dec 03 2017 at 20:25):Is this okay?
/mnt/Files/Home/Dev/brlcad/src/librt/primitives/dsp/dsp.c: In function ‘get_obj_data’:
/mnt/Files/Home/Dev/brlcad/src/librt/primitives/dsp/dsp.c:4087:9: warning: format ‘%llu’ expects argument of type ‘long long unsigned int’, but argument 5 has type ‘uint32_t {aka unsigned int}’ [-Wformat=]
bu_log("ERROR: Binary object '%s' has invalid data (expected type %d, f
^
I've seen similar warnings a few times now, Hope there is no problems.
Naseef (Dec 03 2017 at 20:26):Well, It's a conversion, Won't harm right?
Sean (Dec 03 2017 at 20:31):it's fine -- we fix those types of things just to keep things clean, but they aren't problems. we turn on a LOT of things the compiler can warn about, way more than most.
Sean (Dec 03 2017 at 20:31):helps us keep code quality higher than would naturally happen across a 1M line codebase
Sean (Dec 03 2017 at 20:32):if you want, you could work on fixing all the warnings you encounter -- can make a gci task out of it if there's enough work
Sean (Dec 03 2017 at 20:33):let me know how many warnings or post a full log and I can take a look if it's worth it
Naseef (Dec 03 2017 at 20:33):I think there is more thank enough warnings :joy:
Sean (Dec 03 2017 at 20:34):unlikely they amount to more than a few minutes of work
Naseef (Dec 03 2017 at 20:34):I can see more about 50-100 warnings
Sean (Dec 03 2017 at 20:35):it might seem like a lot if you've not worked with large codebases before ;)
Sean (Dec 03 2017 at 20:35):that's probably about an hour's worth of work
Naseef (Dec 03 2017 at 20:35):Of course!
Sean (Dec 03 2017 at 20:36):that's borderline, but if you want to learn and give them a try, let me know and we can create the task for you
Naseef (Dec 03 2017 at 20:36):I've compiled large codebases with visual studio.
Sean (Dec 03 2017 at 20:36):how big? :)
Naseef (Dec 03 2017 at 20:36):Torque3D
Sean (Dec 03 2017 at 20:36):and studio turns nearly everything off by default
Naseef (Dec 03 2017 at 20:37):I've seen a lot of warnings there too!
Sean (Dec 03 2017 at 20:37):i'm not familiar with how big torque is these days
Sean (Dec 03 2017 at 20:37):last used it nearly 10years ago
Naseef (Dec 03 2017 at 20:37):At least, You've used it.
Naseef (Dec 03 2017 at 20:38):Torque3D is what made me learn C++ :D
Sean (Dec 03 2017 at 20:40):if you know c++, then that dsp.c warning should've made sense, no?
Naseef (Dec 03 2017 at 20:41):What?
Naseef (Dec 03 2017 at 20:41):I haven't really coded much C
Sean (Dec 03 2017 at 20:41):it's good to not ignore warnings from compilers, and is something we do as a matter of standard practice (we tell GCI students to turn strict compilation off because most haven't learned how to read or address warnings)
Sean (Dec 03 2017 at 20:41):that warning was not specific to C
Sean (Dec 03 2017 at 20:41):same in C++
Naseef (Dec 03 2017 at 20:42):I can understand what that warning means
Naseef (Dec 03 2017 at 20:43):long long unsigned int a; unsigned int a; b = a;
Naseef (Dec 03 2017 at 20:43):something similar is happening there I assume.
Sean (Dec 03 2017 at 20:44):not exactly, the warning is partially about types, but more about format specifiers
Sean (Dec 03 2017 at 20:45):it's more like you have some function:(unsigned long long a);
Sean (Dec 03 2017 at 20:46):function(unsigned long long a); unsigned int b; function(b);
Sean (Dec 03 2017 at 20:47):except instead of function(), it's a format specifier (%llu) being passed a size_t (unsigned long)
Sean (Dec 03 2017 at 20:48):upconversion from a smaller size to a larger size isn't generally a problem, so the warning is lame -- but valid - do you know what the fix should be?
Naseef (Dec 03 2017 at 20:50):the variable passed to the function should be long long unsigned int
Naseef (Dec 03 2017 at 20:51):Or like this :
Naseef (Dec 03 2017 at 20:53):function((unsigned long long)b); // casting b as unsigned long long
Sean (Dec 03 2017 at 20:58):casting is one solution, think of any others?
Sean (Dec 03 2017 at 20:58):casting might be appropriate if the %llu was significant for some reason
Naseef (Dec 03 2017 at 21:00):Oh well, I kinda understand I think
Sean (Dec 03 2017 at 21:01):in this case, the warning said the type is uint32_t ... which isn't completely accurate
Sean (Dec 03 2017 at 21:01):it's a size_t
Naseef (Dec 03 2017 at 21:01):printf("Here is my long long unsigned number : %llu",longnum);
Naseef (Dec 03 2017 at 21:02):this how %llu is used right?
Sean (Dec 03 2017 at 21:02):yes
Sean (Dec 03 2017 at 21:02):now make longnum a size_t
Sean (Dec 03 2017 at 21:02):if you compile with -Wformat, you should get the warning
Sean (Dec 03 2017 at 21:03):gotta run, but play with that and see if you can find a better fix than casting ;)
Naseef (Dec 03 2017 at 21:03):I'm on it!
Sean (Dec 03 2017 at 21:07):clue, run "svn annotate" on that file and run "svn diff -cREV" on the last change to that line
Naseef (Dec 03 2017 at 21:18):Build successful!
Naseef (Dec 03 2017 at 21:18):
Naseef (Dec 03 2017 at 21:27):svn: E155007: '/mnt/Files/Home/Dev/src/librt/primitives/dsp/dsp.c' is not a working copy
Naseef (Dec 03 2017 at 21:27):Looks like there's something wrong.
Naseef (Dec 03 2017 at 22:08):Finally, Completed the task and submitted for review!
Time for a real coding task now, The PDF file had a funny warning, It's really funny.
Hope I won't need an Electro-Shock therapy to complete this task!
Naseef (Dec 03 2017 at 22:40):Hey! What do we do with brlcad/.build/doc/?
Sean (Dec 04 2017 at 03:16):svn: E155007: '/mnt/Files/Home/Dev/src/librt/primitives/dsp/dsp.c' is not a working copy
That's because you apparently downloaded the 7.26.4 sources instead of pulling a checkout from SVN. You'll need to redo the build with 7.27.0 sources checked out from SVN at some point.
Naseef (Dec 04 2017 at 10:45):Oops.. Yea. ASAP
Naseef (Dec 04 2017 at 17:36):Do I have to re download everything? Or just the ones modified?
Sean (Dec 04 2017 at 18:17):you'll have to re-download everything (using svn checkout, not a file download)
Naseef (Dec 04 2017 at 18:52):well, I mean...
Naseef (Dec 04 2017 at 18:52):I have brlcad in dev folder.
I checkout BRL-CAD from svn to brlcad from dev.
Most of the BRL-CAD source files are already in the brlcad folder.
Naseef (Dec 04 2017 at 18:55):Now, Will it download all files and overwrite everything or download only the modified files and new files?
Naseef (Dec 04 2017 at 19:07):CMake Warning (dev) at CMakeLists.txt:22 (CMAKE_MINIMUM_REQUIRED): Compatibility with CMake < 2.4 is not supported by CMake >= 3.0. This warning is for project developers. Use -Wno-dev to suppress it. -- Could not locate 'brlcad-config' - no BRL-CAD version available -- Found BRL-CAD at /mnt/Files/Home/Dev/brlcad/ -- Configuration for 'ge'... Omitted. -- Configuration for 'coreInterface'... Omitted. CMake Error: The following variables are used in this project, but they are set to NOTFOUND. Please set them or make sure they are set and tested correctly in the CMake files: BRLCAD_OPENNURBS_HEADERS_DIR used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/coreInterface used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/coreInterface used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/coreInterface used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/coreInterface used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/coreInterface used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/libge used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/libge used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/libge used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/libge TCL_INCLUDE_PATH used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/coreInterface used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/coreInterface used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/coreInterface used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/coreInterface used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/coreInterface used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/libge used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/libge used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/libge used as include directory in directory /mnt/Files/Home/Dev/rt^3/tests/libge -- Configuring incomplete, errors occurred! See also "/mnt/Files/Home/Dev/rt^3/build/CMakeFiles/CMakeOutput.log".
Naseef (Dec 04 2017 at 19:08):WhereOPENNURBS_HEADERS_DIR and TCL_INCLUDE_PATH are located.
Naseef (Dec 04 2017 at 19:12):YOU SHOULD INSTALL BRL-CAD BEFORE COMPILING AND INSTALLING RT^3
How?
Daniel Rossberg (Dec 04 2017 at 19:55):make install in the brlcad build directory
Daniel Rossberg (Dec 04 2017 at 19:57):in rt^3 configuration set BRLCAD_BASE_DIR to the installation path
Naseef (Dec 04 2017 at 23:46):Could not locate 'brlcad-config' - no BRL-CAD version available
Naseef (Dec 04 2017 at 23:50):/usr/bin/ld: cannot find -lcoreinterface collect2: error: ld returned 1 exit status tests/coreInterface/CMakeFiles/printTitle.dir/build.make:94: recipe for target 'tests-bin/printTitle' failed make[2]: *** [tests-bin/printTitle] Error 1 CMakeFiles/Makefile2:165: recipe for target 'tests/coreInterface/CMakeFiles/printTitle.dir/all' failed make[1]: *** [tests/coreInterface/CMakeFiles/printTitle.dir/all] Error 2 Makefile:94: recipe for target 'all' failed make: *** [all] Error 2
Daniel Rossberg (Dec 05 2017 at 08:10):@Naseef I've just tested the coreinterface build from scatch, i.e. starting from an empty build directory. And surprisingly enough, it worked without any additional interaction, setting CMake variables, etc.. All I did before was to install BRL-CAD and adding the BRL-CAD installation bin directory to my path environment variable.
Naseef (Dec 05 2017 at 19:03):PATH=$PATH: [BRLCAD_BUILD_BIN] cd rt^3/build cmake ...
Naseef (Dec 05 2017 at 19:04):Output is too long...
Naseef (Dec 05 2017 at 19:05):CMake Error: The following variables are used in this project, but they are set to NOTFOUND. Please set them or make sure they are set and tested correctly in the CMake files: BRLCAD_HEADERS_DIR BRLCAD_OPENNURBS_HEADERS_DIR TCL_INCLUDE_PATH
Naseef (Dec 05 2017 at 19:19):Where is the TCL_INCLUDE_PATH?
Naseef (Dec 06 2017 at 16:16):@Naseef I've just tested the coreinterface build from scatch, i.e. starting from an empty build directory. And surprisingly enough, it worked without any additional interaction, setting CMake variables, etc.. All I did before was to install BRL-CAD and adding the BRL-CAD installation bin directory to my path environment variable.
I am on linux, so, I have some problems.
Naseef (Dec 06 2017 at 16:17):I have only 1 day remaining to complete this task, Any help would be really helpful.
Daniel Rossberg (Dec 06 2017 at 16:57):I did my test on Linux too. (I need to test it on Windows some day.)
Did you installed BRL-CAD with make install?
Naseef (Dec 06 2017 at 17:23):Yea
Naseef (Dec 06 2017 at 17:23):Should I install again?
Naseef (Dec 06 2017 at 17:23):(deleted)
Naseef (Dec 06 2017 at 17:26):ran make install
Naseef (Dec 06 2017 at 17:32):[100%] Built target buildtimedelta Install the project... -- Install configuration: "Release" CMake Error at cmake_install.cmake:36 (file): file cannot create directory: /usr/brlcad/rel-7.26.4/include. Maybe need administrative privileges. Makefile:126: recipe for target 'install' failed make: *** [install] Error 1
should I install as root?
Naseef (Dec 06 2017 at 17:43):ran make install as with sudo, did I just do something wrong? it worked tho.
Naseef (Dec 06 2017 at 18:28):I still can't build rt^3.
Naseef (Dec 06 2017 at 18:29):same errors as before.
Daniel Rossberg (Dec 06 2017 at 19:57):@Naseef You can set the installation directory in CMake with CMAKE_INSTALL_PATH. I'm usually using a folder in my home directory.
Now, that you did the first thing of what I wrote in bold letters, have you done the second to?
Naseef (Dec 06 2017 at 20:40):Ooh yeah! It compiles now!
Naseef (Dec 06 2017 at 20:40):Thanks a LOT!
Naseef (Dec 07 2017 at 18:26):How do I compile rt_shootray example? Does it have to be compiled alongside BRL-CAD?
Sean (Dec 08 2017 at 08:28):How do I compile rt_shootray example? Does it have to be compiled alongside BRL-CAD?
there's a comment in the example that may help for manually compiling it on linux, and it's build as part of compiling brl-cad via cmake
Jeff Sieu (Dec 08 2017 at 08:29):Is there a guide as to how to export the raytrace into PPM or other image formats?
Sean (Dec 08 2017 at 08:54):jeff, ppm is a crazy simple format that can be written out using printf/cout
Sean (Dec 08 2017 at 08:56):@Jeff Sieu check out this discussion thread: https://cboard.cprogramming.com/c-programming/160019-need-help-printing-circle-ppm-image.html
Sean (Dec 08 2017 at 08:57):basically, you print a simple text header (P6 on one line, width height 255 on another) and then your bytes in rgbrgbrgb... form
Jeff Sieu (Dec 08 2017 at 09:03):Okay, I'll have a look at it. I presume that the output image has to be produced from the coordinates given by the points of the raytrace?
Sean (Dec 08 2017 at 09:04):which task are you working?
Jeff Sieu (Dec 08 2017 at 09:04):Trying Appleseed #2
Sean (Dec 08 2017 at 09:06):ah, yes -- the image should reflect rays that hit or miss
Sean (Dec 08 2017 at 09:08):rtexample shows how to set up a hit/miss callback, the task is to make sure you understand the basics of how that works and to get a grid of rays set up
Jeff Sieu (Dec 08 2017 at 09:11):Oh I see, so just to clarify -- a grid of rays refers to multiple individual rays right?
Sean (Dec 08 2017 at 09:18):yes
Sean (Dec 08 2017 at 09:22):lots of examples throughout the code like this one: src/libanalyze/inside.c
Sean (Dec 08 2017 at 09:23):look for rt_shootray or rt_pattern or rt_gen_ for starters
Sean (Dec 08 2017 at 09:24):(for more examples)
Jeff Sieu (Dec 08 2017 at 09:30):Oh cool, I'll check them out thanks
Mahdi (Dec 26 2017 at 11:00):Has anyone here actually compiled the sources using the MSVC compiler?
Mahdi (Dec 26 2017 at 11:37):I was wondering how long build+compilation would take
Mahdi (Dec 26 2017 at 15:16):I've followed the installation instructions and can't seem to compile (successfully) more than 400 or so projects -- among which include mged and archer. I've heard that other people have had similar issues so is there a resolution that hasn't been included in the installation txt file?
Jeff Sieu (Dec 26 2017 at 15:36):Yup this was precisely my problem
Jeff Sieu (Dec 26 2017 at 15:36):Are you doing the compilation task
Jeff Sieu (Dec 26 2017 at 15:38):I think this is an actual problem -- may be something to do with the compiler version im not too sure but @Daniel Rossberg said that it worked on visual c++ 2010 or smth
Daniel Rossberg (Dec 26 2017 at 16:27):The build with Visual Studio 2017 seems to be broken. The declaration of lrint() doesn't get detected during the configuration which leads to an error for most of the source code files.
Lucas Prieels (Dec 26 2017 at 16:29):For the compilation tasks, we just have to download the compiler and run cmake with -gcc or -clang or so right?
Daniel Rossberg (Dec 26 2017 at 16:29):Beside this there is an issue with src/libbu/parallel.c. Renaming every WIN32 to _WIN32 there should solve it.
Daniel Rossberg (Dec 26 2017 at 16:30):On MS Windows only Visual Studio is known to work.
Lucas Prieels (Dec 26 2017 at 16:32):But I have macOS. Actually I've succeeded to run mged -gcc and mged -clang but I'm just wondering if it's enough
Lucas Prieels (Dec 26 2017 at 16:33):with the flags of course
Daniel Rossberg (Dec 26 2017 at 16:35):If you where able to run a self compiled mged it's enough. Submit a screen shot whit the output of the rt command and you are done.
Lucas Prieels (Dec 26 2017 at 16:36):Alright thanks! I'll do it as soon as my current task is approved
Mahdi (Dec 27 2017 at 01:46):Hm, I'll try investigating some more.
Mahdi (Dec 27 2017 at 01:50):@Sean @Daniel Rossberg
Having to work with CMake really motivated me to start using it to build some of my own side projects. Recently I've been trying to use a command line parser for cmake and was wondering if you guys could guide me on the right path. If I wanted to just take simple commands / flags from the compiler and store them in a variable for cmake to use (like set folder directory -> x) how would I go about doing that? Would I need to define some sort of macro or function?
Sean (Dec 27 2017 at 06:06):Please create a new subject for different topics, but the brief answer is that you don't need to do anything. You just set a variable and it'll be available to you (cmake -DMY_SPECIAL_VAR=/some/folder/here ...).
Mitesh (Dec 28 2017 at 15:35):The build with Visual Studio 2017 seems to be broken. The declaration of lrint() doesn't get detected during the configuration which leads to an error for most of the source code files.
even it is not recognised in visual studio community 2015
Mitesh (Dec 28 2017 at 15:37):In my case i first build the whole project and then build only "install"...which gave me mged
Mahdi (Dec 29 2017 at 02:41):Any thoughts as to why it might have been broken with VS 2017?
Mahdi (Dec 29 2017 at 04:44):Would defining LRINT in the common.h for MSVC version > 1900 (2015) help? Has anyone else tried this?
Mitesh (Dec 29 2017 at 04:46):Sorry...I haven't tried that..
Sean (Dec 29 2017 at 16:36):@Mahdi it's because msvc (finally) added support for lrint() very recently
Sean (Dec 29 2017 at 16:38):we did recently add macros and improved the testing, so something may be wrong in the cmake logic -- someone has to look at the CMakeError.log to see why it's thinking that lrint() isn't available or doesn't work. the lrint test happens in the top-level CMakeLists.txt file around line 2698
Sean (Dec 29 2017 at 16:38):if someone wants to debug and fix that, we can add a task for it ;)
Mahdi (Dec 29 2017 at 20:40):I wouldn't mind trying to fix it. It's weird because CMake finds lrint and defines it in the header but still can't seem to compile it. My guess is that it's related to the ucrt libraries since those are the errors that keep popping up.
Mahdi (Dec 29 2017 at 20:42):I found this on the official CMake documentation notes:
The InstallRequiredSystemLibraries module learned a new CMAKE_INSTALL_UCRT_LIBRARIES option to enable app-local deployment of the Windows Universal CRT libraries with Visual Studio 2015.
I don't think it's a coincidence that version above 2015 aren't compiling. There's definitely some issue with the UCRT and CMake.
Mahdi (Dec 29 2017 at 21:36):That's probably more to do with the dlls though
Mahdi (Dec 29 2017 at 22:01):10>C:\Program Files (x86)\Windows Kits\10\Include\10.0.16299.0\um\winnt.h(20074): error C2059: syntax error: '}'
10>C:\Program Files (x86)\Windows Kits\10\Include\10.0.16299.0\um\winnt.h(20075): error C2059: syntax error: '}'
10>C:\Program Files (x86)\Windows Kits\10\Include\10.0.16299.0\um\winnt.h(20076): error C2143: syntax error: missing '{' before '*'
10>C:\Program Files (x86)\Windows Kits\10\Include\10.0.16299.0\um\winnt.h(20084): error C2061: syntax error: identifier 'IMAGE_POLICY_ENTRY'
10>C:\Program Files (x86)\Windows Kits\10\Include\10.0.16299.0\um\winnt.h(20085): error C2059: syntax error: '}'
10>C:\Program Files (x86)\Windows Kits\10\Include\10.0.16299.0\um\winnt.h(20086): error C2143: syntax error: missing '{' before '*'
10>C:\Program Files (x86)\Windows Kits\10\Include\10.0.16299.0\um\processthreadsapi.h(1126): error C2059: syntax error: 'constant'
10>C:\Program Files (x86)\Windows Kits\10\Include\10.0.16299.0\um\processthreadsapi.h(1128): error C2059: syntax error: '}'
10>C:\Program Files (x86)\Windows Kits\10\Include\10.0.16299.0\um\processthreadsapi.h(1250): error C2059: syntax error: 'constant'
10>C:\Program Files (x86)\Windows Kits\10\Include\10.0.16299.0\um\processthreadsapi.h(1252): error C2059: syntax error: '}'
Anyone know what these errors could mean?
Naseef (Dec 31 2017 at 13:58):Where do I find the cmake build log?
Mitesh (Dec 31 2017 at 14:08):in gui?
Naseef (Dec 31 2017 at 14:09):linux terminal
Mitesh (Dec 31 2017 at 14:12):if in gui cmake build log are found in a space under the generate button..Capture3.PNG
Mitesh (Dec 31 2017 at 14:12):sorry no idea for linux terminal...
Mitesh (Dec 31 2017 at 14:13):i think it shhould be same as that of cmd in windows. Check the build folder..
Armin (LordOfBikes) (Dec 31 2017 at 14:53):@Naseef have a look into CMakeFiles folder which is in your build folder, there you should find CMakeOutput.log or CMakeError.log
Naseef (Dec 31 2017 at 17:36):Is it Ok if I built it without GUI? I reinstalled Ubuntu so it built without a GUI. Will rebuild with GUI soon
Armin (LordOfBikes) (Dec 31 2017 at 17:41):Sure, building witout GUI is fine. I've commented on your task how to get the build log.
Daniel Rossberg (Dec 31 2017 at 17:45):See https://brlcad.org/wiki/Compiling for the additional packages needed to build with GUI on Linux.
Siddharth (Dec 31 2017 at 18:00):Hey @Naseef As far as Ik there will be a txt file named build log in the source folder
Naseef (Jan 01 2018 at 12:12):Yeah, Found that. I'll rebuild it with GUI and send the logs. Thanks.
Naseef (Jan 02 2018 at 20:53):Makefile:162: recipe for target 'all' failed
after a half hour build
Naseef (Jan 02 2018 at 20:53):make again with VERBOSE=1
Naseef (Jan 02 2018 at 20:54):Which version of Qt is required for BRL-CAD?
Naseef (Jan 02 2018 at 20:54):well, I know it was optional.
Naseef (Jan 02 2018 at 21:09):/home/naseef/Code/brlcad/src/libdm/dm-ogl.c
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:44:10: fatal error: X11/extensions/XInput.h: No such file or directory
#include <X11/extensions/XInput.h>
I failed to install something?
Naseef (Jan 03 2018 at 18:45):I have to rebuild everything again because I reinstalled Ubuntu.
Naseef (Jan 03 2018 at 18:46):xinput is already the newest version (1.6.2-1build1).
Mahdi (Jan 03 2018 at 21:19):Would it be reasonable to put fmax and lrint into a namespace? I really don't know how else we can resolve the ms build without renaming them. I tried to include cmath but it seems cmake is building the files using a c compiler so it can't work.
Mahdi (Jan 03 2018 at 21:38):the cmath thing was a little silly I'll admit
Naseef (Jan 03 2018 at 21:41):I think you don't have to change source files to get it compile
Naseef (Jan 03 2018 at 21:42):Maybe change cmakelists?
Naseef (Jan 03 2018 at 21:42):include the source dirs
Mahdi (Jan 03 2018 at 21:45):Well, the issue originated because there are multiple definitions of lrint and without the definition of lrint the solution doesn't compile
Mahdi (Jan 03 2018 at 21:48):If I make it only define lrint for msvc less than 2015 then it just bombards me with a bunch of syntax errors
Sean (Jan 04 2018 at 17:31):@Naseef xinput "should" have been optional if you're using a source checkout. if you're not, you maybe needed to rerun cmake ... if you are using a checkout and ran cmake fresh, then I'd love to see your log files since something didn't test right.
Sean (Jan 04 2018 at 17:32):Would it be reasonable to put fmax and lrint into a namespace? I really don't know how else we can resolve the ms build without renaming them. I tried to include cmath but it seems cmake is building the files using a c compiler so it can't work.
it depends ... there's already testing and wrapping going on for them so we must first make sure that those tests are passing/failing correctly and that the problem isn't something more simple like a cmake test that failed when it should have passed
Sean (Jan 04 2018 at 17:34):Well, the issue originated because there are multiple definitions of lrint and without the definition of lrint the solution doesn't compile
this would be because of the other wrapping going on. you can't just define or not define them -- you have to know, for example, if it's already defined to something and what happened during testing
Sean (Jan 04 2018 at 17:34):if it doesn't build for either of you, please post a log of the entire output from cmake on a fresh checkout
Mahdi (Jan 05 2018 at 00:11):@Sean What could cause the CMake test to fail?
Daniel Rossberg (Jan 05 2018 at 00:25):The problem is in the detection of the lrint() declaration (see the main CMakeLists.txt line 2698ff). Interestingly, lrint() seams to be the only function where the DECL_TEST_SRCS is used. And, somehow the test in line 148 of misc/CMake/BRLCAD_CheckFunctions.cmake fails.
If you want to rerun the test you have to remove the lines with LRINT from CMakeCache.txt. Then you'll see that the configuration finds the lrint() function but not its declaration. The test source in lrint_decl_test alone can be compiled with Visual Studio 2017. The next step would be to debug check_c_source_compiles() to find out why it fails.
However, I'm possibly on the way to find a simple work around.
Daniel Rossberg (Jan 05 2018 at 15:21):Compiling BRL-CAD with Visual Studio doesn't work because of https://core.tcl.tk/tk/info/3d34589aa0ae079039e2e2a60a5ed67c0056836d
Daniel Rossberg (Jan 05 2018 at 16:27):The work around for the lrint() issue is:
Sean (Jan 05 2018 at 17:47):what if you install tcl/tk manually (from a binary distribution) before running cmake?
Mahdi (Jan 06 2018 at 00:07):Is the declaration issue associated with the tcl/tk bug? Why would all the "have decl" fail?
Mahdi (Jan 06 2018 at 00:08):One would suppose those two to be unrelated, no?
Daniel Rossberg (Jan 06 2018 at 10:45):what if you install tcl/tk manually (from a binary distribution) before running cmake?
@Sean It depends on if both sides have to use the same C-runtime or not. I'm afraid they have to use the same what means they need to use the same compiler.
Daniel Rossberg (Jan 06 2018 at 10:47):One would suppose those two to be unrelated, no?
@Mahdi Right, the lrint() declaration and the Tk issues are unrelated. I found a work around for the first one and run into the second then.
Sean (Jan 06 2018 at 15:25):@Daniel Rossberg not sure I follow -- you mean there's some C-runtime incompatibility if we install a system tcl/tk (say compiled with msvc2012) and compile our C code with a newer C-runtime (say msvc2017)? I'm aware of issues with the C++ STL runtime across DLL boundaries, but never heard of a C-runtime incompatibility. And even then, it would have to be across a DLL boundary ... where is that happening?
Daniel Rossberg (Jan 06 2018 at 15:38):@Sean I you malloc() memory on the one side of a DLL boundary and free() it on the other side you have to make sure to use the same ledger. This can be done by using the same C-runtime DLL on both sides, e.g. malloc() and free() will always be calls to this mscrt~.dll. But, if they are using different versions of mscrt ...
Sean (Jan 06 2018 at 15:41):I'm still incredibly surprised that A) there's an incompatible call somewhere in the C runtime (rarely changes, much more surprised the malloc/free would be one they made incompatible) and B) that we so that somewhere, e.g., getting memory from Tcl_Alloc without calling Tcl_Free
Sean (Jan 06 2018 at 15:41):I'd think B is worth fixing if we do that anywhere regardless
Daniel Rossberg (Jan 06 2018 at 15:48):That's not an "incompatibility" but a consequence of the global static memory ledger in the C library. In principle it's the same on Linux, but there everybody is using the same libc and that's way it's fine.
Using something like Tcl_Alloc() and Tcl_Free() is the usual work-around.
Sean (Jan 06 2018 at 15:53):I know that in theory it's possible (on Linux or on Windows), I'm just not aware of this being an issue ever for the C runtime in practice, at least not anywhere in the brl-cad code -- have you seen a failure or is this hypothetical? because I recall (years ago) having installed system tcl/tk on windows, compiled against it with what had to be a different msvc .. maybe with all the c11 changes stuff happened?
Daniel Rossberg (Jan 06 2018 at 16:21):It's an old issue, especially of MS Windows programming, not connected to C11. The cause is simple: There is a global structure which keeps record about the memory retrieved from the OS and the peaces of the given to the application logic. If you use one structure in malloc() and another one for free() the second one doesn't know about this part of memory and may panic. And, whenever somebody doesn't look at this a crash is likely. It isn't a hypothetical issue. That's why e.g. the Object class in the C++ core interface has a Destroy() method.
Curiously, I couldn't find a way to not use libc on Linux. Even llvm uses it. (But I hav't tried the Intel compiler there.) BTW, if you try to start on a new system compiled software on an old one, it'll complain about the libc version even if it's only a "Hello World!". (Can one set the libc compatibility as a compiler option?)
BRL-CAD in general is fully compiled with the same compiler, and on Windows the parts (DLLs, EXEs) are using a/the same C-runtime DLL.
In case of TCL/Tk it looks like if Tcl_Alloc() and Tcl_Free() are used the right way than it could work.
Sean (Jan 06 2018 at 16:36):I meant hypothetical in the sense that you've observed it crash brl-cad code somewhere because of this issue -- I know the problem "can" happen. I'm just surprised that it's actually happening because that's usually bad code. You should always be alloc/free'ing on the same side; you should always open/close on the same side, etc. For there to be an incompatibility, it suggests somewhere that is being violated.
Sean (Jan 06 2018 at 16:38):The libc compatibility on linux is related, but usually triggered by the linker's version checking. Even if there's not an issue, it will abort/complain about the version. You can get past it, but it's a little tricky. Had to do that for some HPC assets a while back compiled against old RHEL, running on newer.
Daniel Rossberg (Jan 06 2018 at 20:45):Speaking about BRL-CAD, I hadn't observed such a memory management issue (or I can't remember).
I had a similar appointment of writing a small program for an old Red Hat. I ended up installing an old Fedora Linux in a VM.
Naseef (Jan 08 2018 at 06:39):CMakeOutput.log Is there something wrong with my CMake output? It doesn't compile
Naseef (Jan 08 2018 at 07:55):here is my build.log
Naseef (Jan 08 2018 at 13:33):Can someone help me? Please
Rahil Malik (Jan 08 2018 at 13:34):Yes, i'll try
Rahil Malik (Jan 08 2018 at 13:35):even i have a doubt in Compilation of BRL-Cad source, after the cmake step many things show - failed , so is that ok ?
Mahdi (Jan 08 2018 at 13:39):Yeah, you don't need to see 'success' on each line
Rahil Malik (Jan 08 2018 at 13:40):yea, not even 'found'
Mahdi (Jan 08 2018 at 13:43):If you didn't do anything strange you should be fine
Rahil Malik (Jan 08 2018 at 13:43):So the build is ready , now to compile the documentation says 'make ' command while that is an unrecognizable to cmd
Mahdi (Jan 08 2018 at 13:44):are you in the build directory?
Rahil Malik (Jan 08 2018 at 13:44):yea BRLCAD/brlcad-svn-trunk/.build/
Mahdi (Jan 08 2018 at 13:44):Wait, you're using Windows
Rahil Malik (Jan 08 2018 at 13:44):Yea , :smiley_cat:
Mahdi (Jan 08 2018 at 13:44):Which generator did you use cmake with
Mitesh (Jan 08 2018 at 13:44):Instead of make we can use MS vS studio in Windows
Mahdi (Jan 08 2018 at 13:45):If you used the 15 generator you're out of luck mate
Jeff Sieu (Jan 08 2018 at 13:45):I think you'll have to open the .sln
Jeff Sieu (Jan 08 2018 at 13:45):^ Oh yeah that too
Rahil Malik (Jan 08 2018 at 13:46):I opened the BRLCAD.sln
Mitesh (Jan 08 2018 at 13:47):Run the BRLCAD-SLN>> "INSTALL" in visual studio...and make sure that in cmake you have changed the installation folder to som place other than C: or C:>>program files....
Mitesh (Jan 08 2018 at 13:47):what is 15 generator? @Mahdi
Mahdi (Jan 08 2018 at 13:48):I was assuming he was using the latest visual studio (2017)
Mitesh (Jan 08 2018 at 13:48):Ohk...
Rahil Malik (Jan 08 2018 at 13:48):cmakepredefinedtargets\INSTALL
Rahil Malik (Jan 08 2018 at 13:49):You assumed right . @Mahdi
Mitesh (Jan 08 2018 at 13:49):I did it in visual studio 2015
Mitesh (Jan 08 2018 at 13:49):got some positive results..(it wasn't 100 % successfull)
Rahil Malik (Jan 08 2018 at 13:51):Run the BRLCAD-SLN>> "INSTALL" in visual studio...and make sure that in cmake you have changed the installation folder to som place other than C: or C:>>program files....
Run or Build ?. ALso i have kept nothing in the C:\ drive
Mitesh (Jan 08 2018 at 13:54):Build
Rahil Malik (Jan 08 2018 at 13:55):btw, i can reclaim a task which i ran out of time to complete , right ?
Jeff Sieu (Jan 08 2018 at 13:55):As long as it's still there when you search for it
Rahil Malik (Jan 08 2018 at 13:55):Yea,, it is there :)
Rahil Malik (Jan 08 2018 at 13:56):by the way i already have BRL-CAD (ReadyMade) installed, that would that create any problem ?
Mahdi (Jan 08 2018 at 14:02):No, that wouldn't be the issue
Mahdi (Jan 08 2018 at 14:02):The build is broken for the latest version of visual studio
Mahdi (Jan 08 2018 at 14:03):Daniel is currently trying to find a workaround
Rahil Malik (Jan 08 2018 at 14:05):so,I would have to revert back to my old version ?
Mahdi (Jan 08 2018 at 14:05):I would personally recommend using linux
Rahil Malik (Jan 08 2018 at 14:06):in VMware station would be fine, right ?
Jeff Sieu (Jan 08 2018 at 14:07):Should be fine, but why not just dual-boot
Rahil Malik (Jan 08 2018 at 14:08):Yes, but wouldn't that take more time ?
Mahdi (Jan 08 2018 at 14:08):If you have enough ram for the vm you should be fine
Rahil Malik (Jan 08 2018 at 14:10):Ubuntu is best of all ?
Mahdi (Jan 08 2018 at 14:11):I don't like ubuntu
Mahdi (Jan 08 2018 at 14:11):Have you ever used linux?
Rahil Malik (Jan 08 2018 at 14:12):Nope, once i installed ubuntu but i don't remember when
Mahdi (Jan 08 2018 at 14:14):Yeah Ubuntu would be the best then
Jeff Sieu (Jan 08 2018 at 14:14):Somehow once I went Ubuntu, I've not gone back to Windows
Rahil Malik (Jan 08 2018 at 14:15):Cool
Mahdi (Jan 08 2018 at 14:16):Once u-bunt, u-want.
Mahdi (Jan 08 2018 at 14:16):Okay that was bad but the effort is what counts
Rahil Malik (Jan 08 2018 at 14:17):i didn'teven get the joke but A for Effort. :laughing:
Jeff Sieu (Jan 08 2018 at 14:17):That was so bad
Mahdi (Jan 08 2018 at 14:17):I want a mod to delete that please
Mahdi (Jan 08 2018 at 14:17):I am ashamed.
Rahil Malik (Jan 08 2018 at 14:18):HAha :rolling_on_the_floor_laughing: :rolling_on_the_floor_laughing:
Rahil Malik (Jan 08 2018 at 14:28):my compilation in visual studio 2017 is done
Rahil Malik (Jan 08 2018 at 14:29):2924 errors 638 warnings
Rahil Malik (Jan 08 2018 at 14:29):build Install.vxcproj - failed
Sean (Jan 08 2018 at 14:31):2924 errors 638 warnings
count of errors in visual studio is practically meaningless
Sean (Jan 08 2018 at 14:32):there can be exactly 1 thing wrong, 1 character, in 1 file .. and it will result in thousands of errors (basically everything fails)
Rahil Malik (Jan 08 2018 at 14:32):no, waiti am attaching the error log
Sean (Jan 08 2018 at 14:33):all that matters when compiling is the command that was run and the first error that resulted
Rahil Malik (Jan 08 2018 at 14:33): Sean (Jan 08 2018 at 14:34):so did you read it?
Rahil Malik (Jan 08 2018 at 14:34):Yes, kinda
Sean (Jan 08 2018 at 14:34):it's basically 1 error ... printed thousands of times
Rahil Malik (Jan 08 2018 at 14:34):I can't figure out most of it
Rahil Malik (Jan 08 2018 at 14:35):Yes,
Sean (Jan 08 2018 at 14:35):there may be two errors in there, but again all that matters is the first error
Rahil Malik (Jan 08 2018 at 14:35):it's basically 1 error ... printed thousands of times
btw @Mahdi told that Visual Studio 2017 has problems in building the Code for BRLCAD
Sean (Jan 08 2018 at 14:36):more specifically, Visual Studio 2017 has problems building Tcl/Tk
Sean (Jan 08 2018 at 14:36):which is bundled and being compiled
Rahil Malik (Jan 08 2018 at 14:37):So, what should i do now ?
Sean (Jan 08 2018 at 14:37):Rahil, if you're up for a test, try installing tcl/tk from a binary distribution
Rahil Malik (Jan 08 2018 at 14:37):Rahil, if you're up for a test, try installing tcl/tk from a binary distribution
I'll check it up
Sean (Jan 08 2018 at 14:38):https://www.activestate.com/activetcl/downloads
Rahil Malik (Jan 08 2018 at 14:39):This will resolve the issue ?
Sean (Jan 08 2018 at 14:39):I'd try the 8.5.18.0 installer
Sean (Jan 08 2018 at 14:40):uhm ... "Rahil, if you're up for a test, ..."
Rahil Malik (Jan 08 2018 at 14:40):OHHH, ok i am up :up:
Rahil Malik (Jan 08 2018 at 14:43):Done, i installed the Tcl/tk . now should i build the INSTALL again ??
Sean (Jan 08 2018 at 14:44):yes, but you'll need to pay attention CLOSE attention when you run cmake
Sean (Jan 08 2018 at 14:44):you need to tell it where tcl/tk is installed
Rahil Malik (Jan 08 2018 at 14:44):i have to run cmake again. ? or just from Visual Studio
Sean (Jan 08 2018 at 14:44):and the summary needs to indicate that it will be using that version
Sean (Jan 08 2018 at 14:44):you have to run cmake again
Sean (Jan 08 2018 at 14:44):it's not going to magically find it :P
Sean (Jan 08 2018 at 14:45):you should delete your build directory
Sean (Jan 08 2018 at 14:45):start fresh
Rahil Malik (Jan 08 2018 at 14:45):OK
Rahil Malik (Jan 08 2018 at 14:48):so i run cmake from command prompt so how can i tell cmake to look for the directory ?
Sean (Jan 08 2018 at 14:50):doesn't matter -- you can use cmake gui
Sean (Jan 08 2018 at 14:50):I'd probably use cmake gui, but just running generate probably won't work -- you have to run configure and then it'll display a list of variables
Sean (Jan 08 2018 at 14:51):then you have to find the variables related to tcl/tk and set them, pointing them at where you installed activetcl
Naseef (Jan 09 2018 at 01:18):td;dr can someone tell me what are the dependencies missing? Have anyone successfully compiled brlcad on 17.10?
Sean (Jan 09 2018 at 01:18):what error are you seeing?
Sean (Jan 09 2018 at 01:19):or conversely, what's the cmake summary?
Naseef (Jan 09 2018 at 01:19):I posted the log above, I've been stuck there for about a week now
Naseef (Jan 09 2018 at 01:19):https://brlcad.zulipchat.com/user_uploads/1549/NX9OArlnnQT1_s1c1rTruplU/build.log
https://brlcad.zulipchat.com/user_uploads/1549/8Li6FaPDcxg-5UfgMr9GXe6a/CMakeOutput.log
Sean (Jan 09 2018 at 01:20):brb, looking
Naseef (Jan 09 2018 at 01:20):"Qt support (optional) .................: OFF" - I need GUI
Naseef (Jan 09 2018 at 01:21):okay, thanks...
Sean (Jan 09 2018 at 01:47):@Naseef you don't need qt support for BRL-CAD -- that's an experimental interface
Sean (Jan 09 2018 at 01:48):cmake summary looks good
Sean (Jan 09 2018 at 01:49):@Naseef aha, so did you read the build log? what's the first error you see?
Sean (Jan 09 2018 at 02:00):@Naseef after answering that -- I just applied an update that should get your further. run "svn up" then make again
Sean (Jan 09 2018 at 02:10):@Naseef that was a build system bug, so feel free to claim the bug report task and I'll give you credit for finding and reporting it well -- just upload those two log files again
Naseef (Jan 09 2018 at 02:25):Well, Was it a bug? I'll find it.
Sean (Jan 09 2018 at 03:04):@Naseef what was the error you saw in that log?
Sean (Jan 09 2018 at 03:09):reading compilation logs is a rather valuable skill
Sean (Jan 09 2018 at 03:10):as is reading all the messages written to you -- I answered your question but you didn't answer mine ;)
Naseef (Jan 09 2018 at 03:14):Well, I was trying to fix the bug while watching cat videos on youtube.
Naseef (Jan 09 2018 at 03:14):XInput?
Sean (Jan 09 2018 at 03:15):is that a question or an answer? :) what about XInput?
Naseef (Jan 09 2018 at 03:16):/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:44:10: fatal error: X11/extensions/XInput.h: No such file or directory
#include <X11/extensions/XInput.h>
Naseef (Jan 09 2018 at 03:16):It's missing
Naseef (Jan 09 2018 at 03:17):xinput is already the newest version (1.6.2-1build1).
Sean (Jan 09 2018 at 03:17):so you have xinput (the library) installed
Naseef (Jan 09 2018 at 03:18):yes
Sean (Jan 09 2018 at 03:18):that does not mean you had the headers installed
Sean (Jan 09 2018 at 03:18):which are a separate package
Sean (Jan 09 2018 at 03:18):so you could have installed them and it would have succeeded
Naseef (Jan 09 2018 at 03:18):like xinput-dev?
Sean (Jan 09 2018 at 03:18):something like that BUT I just changed the code so it should just work for you
Sean (Jan 09 2018 at 03:18):try "svn up" then retry your build (make)
Naseef (Jan 09 2018 at 03:18):should I svn update?
Naseef (Jan 09 2018 at 03:18):up = update?
Sean (Jan 09 2018 at 03:19):yep
Naseef (Jan 09 2018 at 03:19):how do I see the changes?
Naseef (Jan 09 2018 at 03:19):changed lines?
Sean (Jan 09 2018 at 03:19):i also went ahead and fixed all those silly warnings
Naseef (Jan 09 2018 at 03:19):That's great!
Naseef (Jan 09 2018 at 03:20):now I'm deleting everything in build directory and starting again
Sean (Jan 09 2018 at 03:20):I cleaned up some other things at the same time, but you can see what changed with: svn diff -c70572
Sean (Jan 09 2018 at 03:20):you don't need to
Naseef (Jan 09 2018 at 03:21):Oops, I just did!
Sean (Jan 09 2018 at 03:21):helps to learn what each step actually does when compiling so you don't waste time :)
Sean (Jan 09 2018 at 03:22):if you'd installed or uninstalled something, you'd need to delete your build or re-run configuration (cmake) without any cached results
Sean (Jan 09 2018 at 03:22):you were past that, you were compiling
Sean (Jan 09 2018 at 03:22):the error was a compile error, the compile error was fixed, so there's nothing needed -- you just compile again
Naseef (Jan 09 2018 at 03:23):Well, I understand. But deleting it won't harm right?
Sean (Jan 09 2018 at 03:23):if you'd finished compiling and were looking at a link error, then you "might" need to "make clean", but even then usually just make again will work fine
Sean (Jan 09 2018 at 03:23):it harms YOU if you get in a habit of just doing that or not doing that without learning when you NEED to do that :)
Naseef (Jan 09 2018 at 03:24):;)
Naseef (Jan 09 2018 at 03:24):I see a lot of 'failed' in tests
Naseef (Jan 09 2018 at 03:26):and some 'not found'
Sean (Jan 09 2018 at 03:27):yes?
Sean (Jan 09 2018 at 03:28):cmake is called a configuration step because it probes your system, to figure out what you have and do not have
Sean (Jan 09 2018 at 03:28):so it knows how to configure the compilation, what features to turn on or off
Naseef (Jan 09 2018 at 03:31):well, when someone is compiling for distribution, which will be running on different systems, how the configurations works there?
Sean (Jan 09 2018 at 04:06):pretty much the same, because you build for a given configuration
Sean (Jan 09 2018 at 04:08):it's testing to figure out whether it's linux, windows, mac, bsd, etc, and what types of each
Naseef (Jan 09 2018 at 04:58):[ 43%] Built target libged-obj
Makefile:162: recipe for target 'all' failed
make: *** [all] Error 2
I think I should run a verbose build
Sean (Jan 09 2018 at 05:18):As stated in an earlier thread, pasting the end of a log or the last error message is rarely informative. Don’t do that. Post the FIRST error and if not simply make or cmake, the command that preceded it.
Sean (Jan 09 2018 at 05:22):The full log or verbose log are the step after, usually a request when someone fails at reading their log or something exceptional has happened and we want to know how it happened.
Sean (Jan 09 2018 at 05:24):Don’t jump to that step. Read your log and find the first error first, then read it, then google it and ask for help understanding it (usually in that order).
Naseef (Jan 09 2018 at 05:34):A few warnings at 34%
Naseef (Jan 09 2018 at 05:35):more warnings here and there starting at 34%
Naseef (Jan 09 2018 at 05:35):first error at /home/naseef/Code/brlcad/src/libdm/dm-ogl.c:650:5
[ 36% ]
Naseef (Jan 09 2018 at 05:36):In function ‘ogl_open’
Naseef (Jan 09 2018 at 05:36):/home/naseef/Code/brlcad/src/libdm/dm-ogl.c: In function ‘ogl_open’:
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:650:5: error: unknown type name ‘XDeviceInfoPtr’
XDeviceInfoPtr olist = NULL, list = NULL;
^~~~~~~~~~~~~~
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:650:28: error: initialization makes integer from pointer without a cast [-Wint-conversion]
XDeviceInfoPtr olist = NULL, list = NULL;
^~~~
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:650:41: error: initialization makes integer from pointer without a cast [-Wint-conversion]
XDeviceInfoPtr olist = NULL, list = NULL;
^~~~
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:651:5: error: unknown type name ‘XDevice’
XDevice *dev = NULL;
^~~~~~~
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:652:5: error: unknown type name ‘XEventClass’; did you mean ‘XEvent’?
XEventClass e_class[15];
^~~~~~~~~~~
XEvent
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:653:5: error: unknown type name ‘XInputClassInfo’
XInputClassInfo *cip;
^~~~~~~~~~~~~~~
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:891:18: error: ‘XDeviceInfoPtr’ undeclared (first use in this function)
if (list == (XDeviceInfoPtr)NULL ||
^~~~~~~~~~~~~~
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:891:18: note: each undeclared identifier is reported only once for each function it appears in
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:892:26: error: expected ‘)’ before numeric constant
list == (XDeviceInfoPtr)1) goto Done;
^
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:895:10: error: invalid type argument of ‘->’ (have ‘int’)
if (list->use == IsXExtensionDevice) {
^~
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:895:19: error: ‘IsXExtensionDevice’ undeclared (first use in this function); did you mean ‘LastExtensionError’?
if (list->use == IsXExtensionDevice) {
^~~~~~~~~~~~~~~~~~
LastExtensionError
In file included from /home/naseef/Code/brlcad/include/raytrace.h:58:0,
from /home/naseef/Code/brlcad/include/dm.h:33,
from /home/naseef/Code/brlcad/src/libdm/dm-ogl.c:78:
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:896:27: error: invalid type argument of ‘->’ (have ‘int’)
if (BU_STR_EQUAL(list->name, "dial+buttons")) {
^
/home/naseef/Code/brlcad/include/bu/str.h:123:42: note: in definition of macro ‘BU_STR_EQUAL’
#define BU_STR_EQUAL(s1, s2) (bu_strcmp((s1), (s2)) == 0)
^~
[ 36%] Building C object src/libged/CMakeFiles/libged-obj.dir/brep.c.o
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:897:14: error: implicit declaration of function ‘XOpenDevice’ [-Wimplicit-function-declaration]
if ((dev = XOpenDevice(pubvars->dpy,
^~~~~~~~~~~
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:898:16: error: invalid type argument of ‘->’ (have ‘int’)
list->id)) == (XDevice *)NULL) {
^~
[ 36%] Building C object src/librt/CMakeFiles/librt-obj.dir/primitives/eto/eto.c.o
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:898:27: error: ‘XDevice’ undeclared (first use in this function); did you mean ‘ndevices’?
list->id)) == (XDevice *)NULL) {
^~~~~~~
ndevices
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:898:36: error: expected expression before ‘)’ token
list->id)) == (XDevice *)NULL) {
^
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:903:17: error: request for member ‘classes’ in something not a structure or union
for (cip = dev->classes, k = 0; k < dev->num_classes;
^~
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:903:26: warning: left-hand operand of comma expression has no effect [-Wunused-value]
for (cip = dev->classes, k = 0; k < dev->num_classes;
^
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:903:42: error: request for member ‘num_classes’ in something not a structure or union
for (cip = dev->classes, k = 0; k < dev->num_classes;
^~
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:905:18: error: request for member ‘input_class’ in something not a structure or union
switch (cip->input_class) {
^~
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:928:3: error: implicit declaration of function ‘XSelectExtensionEvent’; did you mean ‘XFreeExtensionList’? [-Wimplicit-function-declaration]
XSelectExtensionEvent(pubvars->dpy, pubvars->win, e_class, nclass);
^~~~~~~~~~~~~~~~~~~~~
XFreeExtensionList
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:934:5: error: implicit declaration of function ‘XFreeDeviceList’; did you mean ‘XFreeStringList’? [-Wimplicit-function-declaration]
XFreeDeviceList(olist);
^~~~~~~~~~~~~~~
XFreeStringList
/home/naseef/Code/brlcad/src/libdm/dm-ogl.c:649:9: warning: unused variable ‘unused’ [-Wunused-variable]
int unused;
^~~~~~
src/libdm/CMakeFiles/libdm-obj.dir/build.make:86: recipe for target 'src/libdm/CMakeFiles/libdm-obj.dir/dm-ogl.c.o' failed
make[2]: [src/libdm/CMakeFiles/libdm-obj.dir/dm-ogl.c.o] Error 1
CMakeFiles/Makefile2:18871: recipe for target 'src/libdm/CMakeFiles/libdm-obj.dir/all' failed
make[1]: [src/libdm/CMakeFiles/libdm-obj.dir/all] Error 2
make[1]: *** Waiting for unfinished jobs....
Sean (Jan 09 2018 at 05:36):ah, there we go
Sean (Jan 09 2018 at 05:37):need one more protection
Naseef (Jan 09 2018 at 05:38):what do you mean?
Sean (Jan 09 2018 at 05:39):so the underlying problem is that the code was written to use the XInput API in just a couple places
Sean (Jan 09 2018 at 05:39):to use an API, you need libs and you need headers
Naseef (Jan 09 2018 at 05:39):So I failed to follow the build instructions on wiki?
Sean (Jan 09 2018 at 05:39):it detected that you do not have the headers, so the code needs to disable the XInput code
Sean (Jan 09 2018 at 05:39):no....
Sean (Jan 09 2018 at 05:40):code is constantly changing, you are on the latest code which we change throughout the day nearly every day
Naseef (Jan 09 2018 at 05:40):I updated just a few hours ago.
Sean (Jan 09 2018 at 05:41):yes
Sean (Jan 09 2018 at 05:41):because I asked you to, I made a change
Sean (Jan 09 2018 at 05:41):did you read the patch, you asked for it
Naseef (Jan 09 2018 at 05:41):I did
Sean (Jan 09 2018 at 05:42):I don't have your setup, I don't have your errors, so I made a change that turned off some XInput code that wasn't getting turned off -- entirely just by reading your error log
Sean (Jan 09 2018 at 05:42):it's not something I could test. you're testing it now
Naseef (Jan 09 2018 at 05:42):Well, How you would handle such error?
Sean (Jan 09 2018 at 05:43):there's a different error now because while I turned off one bit of XInput code, there was another less obvious bit that was overlooked
Sean (Jan 09 2018 at 05:43):so that bit needs to be turned off too
Naseef (Jan 09 2018 at 05:43):is the XInput of X11 or DirectX?
Sean (Jan 09 2018 at 05:44):i don't understand your question
Naseef (Jan 09 2018 at 05:44):me neither
Sean (Jan 09 2018 at 05:44):heh
Sean (Jan 09 2018 at 05:45):I'm working on an additional change, which again I cannot test, but I can certainly make the code edits and we can see what happens
Naseef (Jan 09 2018 at 05:45):I'll do the testing and reporting.
Sean (Jan 09 2018 at 05:45):you can test it, maybe understand how it pertains to the first change
Naseef (Jan 09 2018 at 05:46):Sure, I'll try to understand
Sean (Jan 09 2018 at 05:50):none of this is mysterious magic.. computers just only do exactly what we tell them to do, so when programming it takes time to get things exact. this is how coding goes. (always)
Sean (Jan 09 2018 at 05:58):okay, committed
Naseef (Jan 09 2018 at 05:58):Yeah, We need A.I. powered chips.
Naseef (Jan 09 2018 at 05:58):I'll update...
Sean (Jan 09 2018 at 05:58):svn diff -c70576 to see the change
Sean (Jan 09 2018 at 05:59):any plus lines are lines added, any minus lines are lines deleted.
Naseef (Jan 09 2018 at 05:59):I see
Sean (Jan 09 2018 at 06:02):so you'lll see some preprocessor blocks, which is a typical way to "turn off" code at compile-time -- the compiler won't compile code between the #ifdef and #endif blocks unless that HAVE_X11_EXTENSIONS_XINPUT_H symbol is declared, which is tested for and set during cmake
Naseef (Jan 09 2018 at 06:25):It's compiled past 43%, Which means build succeeded.
Naseef (Jan 09 2018 at 06:52):Elapsed compilation time: 1 hour 17 minutes 6 seconds
SUCCESS!
Naseef (Jan 09 2018 at 06:52):Thanks @Sean
Naseef (Jan 09 2018 at 12:56):(deleted)
Naseef (Jan 09 2018 at 12:57):the *.g databases are generated by cmake? I want to check that out.
Naseef (Jan 09 2018 at 13:01):what are the *.asc files?
Jeff Sieu (Jan 15 2018 at 15:19):/home/jeff/Documents/brlcad/src/nirt/parse_fmt.c: In function ‘direct_output’:
/home/jeff/Documents/brlcad/src/nirt/parse_fmt.c:655:22: error: ‘popen’ undeclared (first use in this function)
(openfunc == popen) ? "'| " : "'", dest_string);
^
Getting this error when trying to build from latest source checkout.
Sean (Jan 16 2018 at 07:30):Getting this error when trying to build from latest source checkout.
@Jeff Sieu do you still get this error? it should have been fixed yesterday. one of the devs has been working on strictness compilation, which is tricky cross platform
Jeff Sieu (Jan 16 2018 at 09:33):@Jeff Sieu do you still get this error? it should have been fixed yesterday. one of the devs has been working on strictness compilation, which is tricky cross platform
Yep, I still do.
Sean (Jan 17 2018 at 15:17):@Jeff Sieu just applied a few fixes that should fix that one -- would you test again?
Jeff Sieu (Jan 17 2018 at 15:40):@Jeff Sieu just applied a few fixes that should fix that one -- would you test again?
@Sean yup, works now
Sean (Jan 17 2018 at 15:41):woo hoo! thanks jeff
Last updated: Jan 10 2025 at 00:48 UTC