www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.learn - building OpenMesh/D

reply Jilani <email email.com> writes:
I have tried to build OpenMesh/D but in vain. I did (from the dos command line):
dsss build
But the sytem (*), after a while, hangs infinetely with 100% CPU usage. What
am I missing?
Thank you.
Ji

(*)
- Intel P4 2.4 GH with  512 MB of RAM
- Windows 2000 Pro
- dmd 1.021
Sep 26 2007
next sibling parent reply Tomas Lindquist Olsen <tomas famolsen.dk> writes:
Jilani wrote:

 I have tried to build OpenMesh/D but in vain. I did (from the dos command
 line): dsss build
 But the sytem (*), after a while, hangs infinetely with 100% CPU usage.
 What am I missing?
 Thank you.
 Ji
 
 (*)
 - Intel P4 2.4 GH with  512 MB of RAM
 - Windows 2000 Pro
 - dmd 1.021
I recently had a problem similar to this. Upgrading DSSS to the latest version fixed it if I recall correctly.
Sep 26 2007
next sibling parent jilani <kkkk kkk.com> writes:
I have the last one (V-0.72.1).
An other hint?
Thanks.
Jilani
Sep 26 2007
prev sibling parent Bill Baxter <dnewsgroup billbaxter.com> writes:
Tomas Lindquist Olsen wrote:
 Jilani wrote:
 
 I have tried to build OpenMesh/D but in vain. I did (from the dos command
 line): dsss build
 But the sytem (*), after a while, hangs infinetely with 100% CPU usage.
 What am I missing?
 Thank you.
 Ji

 (*)
 - Intel P4 2.4 GH with  512 MB of RAM
 - Windows 2000 Pro
 - dmd 1.021
I recently had a problem similar to this. Upgrading DSSS to the latest version fixed it if I recall correctly.
There is definitely still a hanging bug in .72.1. Gregor said that it's fixed the not-yet-released 0.73. (The bug I'm talking about comes from failing to close or properly balance nested constructs like parentheses.) But anyway those hangs usually happen before any linker line appears, probably during dsss's initial scan to find imports. This seems to be happening after, so I'm not sure what it could be. I just tried building the docs, though (dsss build --docs), and apparently '--docs' makes dsss revert to trying to build a .lib. Not sure if that's from necessity or just a bug. I suspect the latter, because I think dmd generate docs one-file-at-a-time without building a lib. --bb
Sep 26 2007
prev sibling parent reply Bill Baxter <dnewsgroup billbaxter.com> writes:
Jilani wrote:
 I have tried to build OpenMesh/D but in vain. I did (from the dos command
line):
 dsss build
 But the sytem (*), after a while, hangs infinetely with 100% CPU usage. What
 am I missing?
 Thank you.
 Ji
 
 (*)
 - Intel P4 2.4 GH with  512 MB of RAM
 - Windows 2000 Pro
 - dmd 1.021
I think I can confirm that. Did you get the big linker line to print out first? That's what I get, then it seems to hang. My advice would be to not attempt to compile it as a .lib. It's almost completely templates anyway, so there's not much the compiler can know to instantiate. I just check in a change that makes it a dsss sourcelibrary rather than library. So now dsss build will work (but it won't really do much :-) I actually haven't tried to build it as a lib for a while. Thanks for pointing it out. Instead you should try compiling the demo programs in OpenMesh/Apps ... but looks like those have been broken by some recent changes I made. Hang on a bit while I fix those. --bb
Sep 26 2007
parent reply Bill Baxter <dnewsgroup billbaxter.com> writes:
Bill Baxter wrote:
 Jilani wrote:
 I have tried to build OpenMesh/D but in vain. I did (from the dos 
 command line):
 dsss build
 But the sytem (*), after a while, hangs infinetely with 100% CPU 
 usage. What
 am I missing?
 Thank you.
 Ji

 (*)
 - Intel P4 2.4 GH with  512 MB of RAM
 - Windows 2000 Pro
 - dmd 1.021
I think I can confirm that. Did you get the big linker line to print out first? That's what I get, then it seems to hang. My advice would be to not attempt to compile it as a .lib. It's almost completely templates anyway, so there's not much the compiler can know to instantiate. I just check in a change that makes it a dsss sourcelibrary rather than library. So now dsss build will work (but it won't really do much :-) I actually haven't tried to build it as a lib for a while. Thanks for pointing it out. Instead you should try compiling the demo programs in OpenMesh/Apps ... but looks like those have been broken by some recent changes I made. Hang on a bit while I fix those.
Ok, should be fixed now. The OpenMesh/Apps/Decimating should build out of the box. The GLViewer example depends on the dead Helix project for Quaternion and Matrix44 (http://www.dsource.org/helix). I should probably write some Quaternion / Matrix templates that work together nicely with the open mesh Vector class... --bb
Sep 26 2007
parent reply Jilani <aaa aaa.com> writes:
 Ok, should be fixed now.  The OpenMesh/Apps/Decimating should build out 
 of the box.
But it is a 3.2 GB exec file! Why is it so fat? Jilani
Sep 27 2007
parent reply Bill Baxter <dnewsgroup billbaxter.com> writes:
Jilani wrote:
 Ok, should be fixed now.  The OpenMesh/Apps/Decimating should build out 
 of the box.
But it is a 3.2 GB exec file! Why is it so fat? Jilani
I think you mean 3.2 _MB_. Big difference. :-) It's huge because debugging symbols are on. Building it with buildflags=-O -inline -release I get a 381 KB exe. --bb
Sep 27 2007
next sibling parent Jilani <j.khaldi oltrelinux.com> writes:
 I think you mean 3.2 _MB_.  Big difference.  :-)
Yes, yes.
 It's huge because debugging symbols are on.
 Building it with
      buildflags=-O -inline -release
 I get a 381 KB exe.
That's ok now. Thank you :-) Is there any chance to get GLViewer compiled? Jilani
Sep 27 2007
prev sibling parent reply Jilani <j.khaldi oltrelinux.com> writes:
 I think you mean 3.2 _MB_.  Big difference.  :-)
Yes, yes.
 It's huge because debugging symbols are on.
 Building it with
      buildflags=-O -inline -release
 I get a 381 KB exe.
That's ok now. Thank you :-) Is there any chance to get GLViewer compiled? Jilani
Sep 27 2007
parent reply Bill Baxter <dnewsgroup billbaxter.com> writes:
Jilani wrote:
 I think you mean 3.2 _MB_.  Big difference.  :-)
Yes, yes.
 It's huge because debugging symbols are on.
 Building it with
      buildflags=-O -inline -release
 I get a 381 KB exe.
That's ok now. Thank you :-) Is there any chance to get GLViewer compiled? Jilani
Yeh sure, you just need Derelict, which you can get by doing "dsss net deps" in the GLViewer directory, and Helix which you'll have to download from dsource: http://www.dsource.com/projects/helix. Did you do that and it still doesn't compile? --bb
Sep 27 2007
parent reply jilani <j.khaldi oltrelinux.com> writes:
 Yeh sure, you just need Derelict, which you can get by doing "dsss net 
 deps" 
Here it compiles fine.
in the GLViewer directory, and Helix which you'll have to download 
 from dsource: http://www.dsource.com/projects/helix.
http://www.dsource.org/projects/helix
 Did you do that and it still doesn't compile?
No :(( I have got this error message: D:\OpenMeshSrc\trunk\OpenMesh\Apps\GLViewer>dsss build GLViewer.d => GLViewer + c:\dsss\bin\rebuild.exe -Idsss_imports\ -I. -S.\ -Ic:\dsss\include\d -Sc:\dsss \lib\ -Ic:\dsss\include\d -Sc:\dsss\lib -oqdsss_objs -I../../.. -debug -g GLVi ewer.d -ofGLViewer helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead Command c:\dsss\bin\rebuild.exe returned with code 1, aborting. Thank you again for your help. Jilani
Sep 27 2007
parent reply Bill Baxter <dnewsgroup billbaxter.com> writes:
jilani wrote:
 Yeh sure, you just need Derelict, which you can get by doing "dsss net 
 deps" 
Here it compiles fine.
 in the GLViewer directory, and Helix which you'll have to download 
 from dsource: http://www.dsource.com/projects/helix.
http://www.dsource.org/projects/helix
 Did you do that and it still doesn't compile?
No :(( I have got this error message: D:\OpenMeshSrc\trunk\OpenMesh\Apps\GLViewer>dsss build GLViewer.d => GLViewer + c:\dsss\bin\rebuild.exe -Idsss_imports\ -I. -S.\ -Ic:\dsss\include\d -Sc:\dsss \lib\ -Ic:\dsss\include\d -Sc:\dsss\lib -oqdsss_objs -I../../.. -debug -g GLVi ewer.d -ofGLViewer helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead Command c:\dsss\bin\rebuild.exe returned with code 1, aborting. Thank you again for your help. Jilani
Hmm. Ok. I can believe that. I have made modifications to my local copy of helix that I tried to get incorportated upstream, but to no effect. Dead project, as I said. In that case the best thing is probably for me to check my updated version of helix into the OpenMesh repository. I'll do that later today. --bb
Sep 27 2007
parent reply Bill Baxter <dnewsgroup billbaxter.com> writes:
Bill Baxter wrote:
 jilani wrote:
 Yeh sure, you just need Derelict, which you can get by doing "dsss 
 net deps" 
Here it compiles fine.
 in the GLViewer directory, and Helix which you'll have to download 
 from dsource: http://www.dsource.com/projects/helix.
http://www.dsource.org/projects/helix
 Did you do that and it still doesn't compile?
No :(( I have got this error message: D:\OpenMeshSrc\trunk\OpenMesh\Apps\GLViewer>dsss build GLViewer.d => GLViewer + c:\dsss\bin\rebuild.exe -Idsss_imports\ -I. -S.\ -Ic:\dsss\include\d -Sc:\dsss \lib\ -Ic:\dsss\include\d -Sc:\dsss\lib -oqdsss_objs -I../../.. -debug -g GLVi ewer.d -ofGLViewer helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead helix\basic.d(114): 'l' suffix is deprecated, use 'L' instead Command c:\dsss\bin\rebuild.exe returned with code 1, aborting. Thank you again for your help. Jilani
Hmm. Ok. I can believe that. I have made modifications to my local copy of helix that I tried to get incorportated upstream, but to no effect. Dead project, as I said. In that case the best thing is probably for me to check my updated version of helix into the OpenMesh repository. I'll do that later today.
Helix is checked into the OpenMeshD svn now. --bb
Sep 28 2007
parent jilani <j.khaldi oltrelinux.com> writes:
 Helix is checked into the OpenMeshD svn now.
Great! Very interesting. Jilani
Sep 28 2007