Menu
▾
▴
gofigure2-developers — List for developers and corresponding questions
You can subscribe to this list here.
| 2010 |
Jan
|
Feb
|
Mar
(10) |
Apr
(10) |
May
(2) |
Jun
(23) |
Jul
(22) |
Aug
(4) |
Sep
(2) |
Oct
(6) |
Nov
(8) |
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2011 |
Jan
|
Feb
|
Mar
(19) |
Apr
(1) |
May
(2) |
Jun
|
Jul
(1) |
Aug
|
Sep
|
Oct
(2) |
Nov
(14) |
Dec
(5) |
| 2012 |
Jan
|
Feb
|
Mar
(6) |
Apr
(1) |
May
(29) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2014 |
Jan
(6) |
Feb
|
Mar
(6) |
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Arnaud G. <arn...@hm...> - 2010-10-28 17:53:24
|
Hi guys, I have tried to compile with using the trunk version of clang (on linux 64 bits), and it worked fine! I will set up a nightly builds and turn on more flags. Best, Arnaud For more details about clang: http://llvm.org/ http://clang.llvm.org/ |
|
From: Arnaud G. <arn...@hm...> - 2010-10-25 22:19:17
|
Hi all,
In the process of migrating to git, we will have to get rid of the first
part of the header (which was not necessarily used).
Indeed these keyword (author, revision, date) don't necessarily make
sense in a git environment...
So from now on, do not use any more the following
/*=========================================================================
Author: $Author$ // Author of last commit
Version: $Rev$ // Revision of last commit
Date: $Date$ // Date of last commit
=========================================================================*/
/*=========================================================================
Authors: The GoFigure Dev. Team.
...
But files should start as follows:
/*=========================================================================
Authors: The GoFigure Dev. Team.
...
Thanks,
Arnaud
|
|
From: Arnaud G. <arn...@hm...> - 2010-10-22 22:21:04
|
Hi Folks,
Once we are finished with the generation of installers for the latest
release GoFigure2, we will migrate to git within a week:
* From November 1st, we will only use git for GoFigure2's development
* 2 (similar) workflows are being discussed among developers:
o
http://nvie.com/posts/a-successful-git-branching-model/
o
http://www.itk.org/Wiki/Git/Workflow/Topic
* Within the week,
o we will update nightly / continuous / experimental
submission procedures
o we will convert the history from svn to a git one.
In order to convert all svn authors to git format, we will
use the following author mapping (in alphabetical order):
+ agouaillard = Alexandre Gouaillard <ago...@gm...>
+ antonin07130 = Antonin Perrot-Audet
<ant...@gm...>
+ arnaudgelas = Arnaud Gelas <arn...@hm...>
+ evanschwab = Evan Schwab <esc...@gm...>
+ krm15 = Kishore Mosaliganti <kis...@gm...>
+ lsouhait = Lydie Souhait <lyd...@hm...>
+ nicolasrannou = Nicolas Rannou
<nic...@hm...>
Best,
Arnaud
|
|
From: Arnaud G. <arn...@hm...> - 2010-10-22 22:13:37
|
Hi Folks, I am pleased to announce that we are about to tag the new release of GoFigure2, i.e. version 0.6.0 !!! We will shortly start generating and uploading installers for each platform. Following a re-factorization period, this new version is supposed to be more stable and more efficient in terms of timing and memory usage. We will shortly update GoFigure2's website to present the list of new features. Thanks, Best, Arnaud |
|
From: Arnaud G. <arn...@hm...> - 2010-10-22 22:00:43
|
Hi Folks, If you use the git version of ITK to build gofigure, you will have to pull now from the master branch of ITK. Indeed ITK Label Map's code now requires to link against ITKReview. Note that if you use one release version 3.20 or an alpha version of ITKv4, you don't have to do anything! Best, Arnaud |
|
From: Arnaud G. <arn...@hm...> - 2010-10-15 14:25:24
|
Hi all,
I have just updated our uncrustify configuration file (taken from the
ITK project), and ran it on the trunk.
To run uncrustify on gofigure, you need
* to install a recent version of uncrustify (>=0.56, available via
git, see http://uncrustify.sourceforge.net/ )
* to proceed as follows
$ cd /path/to/gofigure2/trunk/KWStyle/
$ ./uncrustify.sh
Then the style should be automatically fixed!
Have a nice day,
Arnaud
|
|
From: Arnaud G. <arn...@hm...> - 2010-09-13 18:27:14
|
Hi all, In order to create portfile for gofigure, I opened several tickets on MacPorts' trac system: #26371: Update ITK portfile for ITK3.20 #26372: make CableSwig-ITK a variant #26373: enable parallel builds for ITK #26374: set ITK_USE_REVIEW_STATISTICS to ON #26375: add variant mysql to vtk => FIXED! #26376: add variant qt4-mac to vtk => FIXED! (Note that, I have tried reaching the guy in charge of ITK portfile, but have not get any answer yet...) Once these tickets will be all close, it would be quite easy to write a MacPorts for gofigure. Cheers, Arnaud |
|
From: Arnaud G. <arn...@hm...> - 2010-09-13 18:07:12
|
Hi guys, Since ITK now requires CMake 2.8.2, we also move forward and require CMake 2.8.2 from now on! Note that it has been active since the revision 2057, i.e. since 08/30/2010. Cheers, Arnaud |
|
From: Nicolas R. <Nic...@hm...> - 2010-08-19 21:44:05
|
Hi all,
The FFMPEG video recorder has been updated in GoFigure2 and VTK.
It is now working with all the FFMPEG versions!
Please let us know if you experience any trouble while making videos
in GoFigure2!
Best,
Nicolas
|
|
From: Arnaud G. <arn...@hm...> - 2010-08-06 17:29:21
|
Hi all, I have recently created an organization gofigure2 on github, and move the mirror from my account to the organization's one. So if you were using git and pulling from my gofigure2's account, or use my gofigure2's account as remote, please update in consequence. gofigure2's github organization http://github.com/gofigure2 gofigure2's new repository http://github.com/gofigure2/GoFigure2 Best, Arnaud |
|
From: Arnaud G. <arn...@hm...> - 2010-08-05 22:01:15
|
Hi, I have just tried, I have no problem to access the wiki... Arnaud Note that I have tried with firefox/chrome from my desktop and laptop... On 08/05/2010 05:57 PM, Antonin Perrot-Audet wrote: > Hello, > gofigure2's wiki is going through some trouble, is it due to sourceforge ? > see attached image > |
|
From: Antonin Perrot-A. <ant...@gm...> - 2010-08-05 21:57:17
|
Hello, gofigure2's wiki is going through some trouble, is it due to sourceforge ? see attached image -- Antonin |
|
From: Arnaud G. <arn...@hm...> - 2010-07-29 16:07:36
|
Hi all, Just for the record... When building vtk with FFMPEG, if you get the following error: /usr/include/ffmpeg/libavutil/common.h:154: error: ‘UINT64_C’ was not declared in this scope you need to add -D__STDC_CONSTANT_MACROS to CMAKE_CXX_FLAGS (Tested on Fedora 12 and 13). Then it works fine! Best, Arnaud |
|
From: Arnaud G. <arn...@hm...> - 2010-07-27 20:44:12
|
Hi guys, I have just published the Cell Lineage Viewer's source code and Data on github: http://github.com/gofigure2/CellLineageViewer http://github.com/gofigure2/CellLineageData Cheers, Arnaud |
|
From: Nicolas R. <nic...@hm...> - 2010-07-20 15:17:25
|
Hi all, GoFigure2 now requires ITK to be built with ITK_USE_REVIEW_STATISTICS turned ON (in addition to ITK_USE_REVIEW). ITK_USE_REVIEW = ON ITK_USE_REVIEW_STATISTICS = ON There was a bug in the previous ITK Statistics Framework which occurred when the meshes on which ones we wanted to process statistics (volume, intensity, etc.) were too big. http://old.nabble.com/ScalarImageToHistogramGenerator-frequency-container-overflow-td28596437.html Nicolas |
|
From: Antonin Perrot-A. <ant...@gm...> - 2010-07-12 19:05:44
|
-- Antonin |
|
From: Arnaud G. <arn...@hm...> - 2010-07-09 13:08:21
|
I have added one call to vtkErrorMacro in the vtkLSMReader, it generates an event vtkCommand::ErrorEvent, that can easily be caught at an higher level. Arnaud On 07/09/2010 08:16 AM, Arnaud GELAS wrote: > Hi Alex, > > You're right, it seems lsm files using lzw compression are not so > common. However, the lsm file we use for testing > (Testing/Data/earpax2isl11.lsm) is lzw-compressed. > We should upload another lsm which is not compressed! > > You're right, for the time being an exception is most probably the best > solution to handle this problem! > I shall fix it today... > > Thanks, > > Arnaud > > On 07/08/2010 09:03 PM, Alexandre GOUAILLARD wrote: > >> hi arnaud, >> >> I i should have already removed this file from the project. i.e. the >> project did not include it and did not compile against it. It has >> never been a problem so far, none of the image sean feed me with was >> compressed. >> >> to be able to catch nicely when you otherwise would require it, I >> suggest you put an exception in the vtkLSM reader when it detects that >> the image part of the LSM is compressed. I do not remember the name of >> the method, but there is a method that is called after parsing the >> tags that tries to read the image data. This method is basically a >> switch that dispatch the data to be read to other methods depending on >> the compression and format. >> >> hope this helps. >> >> alex. >> >> >> On Wed, Jul 7, 2010 at 10:16 PM, Arnaud Gelas >> <arn...@hm...> wrote: >> >> >>> Hi guys, >>> >>> During the NAMIC week at Boston, Mathieu Malaterre pointed me that we >>> were using one header from ffmpeg for the lzw decompression. >>> This header had to be removed for license compliancy!!! >>> >>> This code was used when opening one lsm file compressed with lzw. Note >>> that most of current lsm are not compressed, it seems that only old >>> lsm files were using it (yet to be confirmed). >>> So when opening lsm files, you see weird images on the screen, the >>> reason could be that it was lzw compressed... In such a case, keep use >>> inform!!! >>> >>> We need to figure out one way to solve this issue: either by looking >>> at an LZW implementation with right license, either by reimplementing >>> it, either by reimplementing the lsm reader... >>> >>> Cheers, >>> Arnaud >>> >>> ------------------------------------------------------------------------------ >>> This SF.net email is sponsored by Sprint >>> What will you do first with EVO, the first 4G phone? >>> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >>> _______________________________________________ >>> Gofigure2-developers mailing list >>> Gof...@li... >>> https://lists.sourceforge.net/lists/listinfo/gofigure2-developers >>> >>> >>> > > ------------------------------------------------------------------------------ > This SF.net email is sponsored by Sprint > What will you do first with EVO, the first 4G phone? > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first > _______________________________________________ > Gofigure2-developers mailing list > Gof...@li... > https://lists.sourceforge.net/lists/listinfo/gofigure2-developers > |
|
From: Arnaud G. <arn...@hm...> - 2010-07-09 12:20:33
|
Hi Alex, You're right, it seems lsm files using lzw compression are not so common. However, the lsm file we use for testing (Testing/Data/earpax2isl11.lsm) is lzw-compressed. We should upload another lsm which is not compressed! You're right, for the time being an exception is most probably the best solution to handle this problem! I shall fix it today... Thanks, Arnaud On 07/08/2010 09:03 PM, Alexandre GOUAILLARD wrote: > hi arnaud, > > I i should have already removed this file from the project. i.e. the > project did not include it and did not compile against it. It has > never been a problem so far, none of the image sean feed me with was > compressed. > > to be able to catch nicely when you otherwise would require it, I > suggest you put an exception in the vtkLSM reader when it detects that > the image part of the LSM is compressed. I do not remember the name of > the method, but there is a method that is called after parsing the > tags that tries to read the image data. This method is basically a > switch that dispatch the data to be read to other methods depending on > the compression and format. > > hope this helps. > > alex. > > > On Wed, Jul 7, 2010 at 10:16 PM, Arnaud Gelas > <arn...@hm...> wrote: > >> Hi guys, >> >> During the NAMIC week at Boston, Mathieu Malaterre pointed me that we >> were using one header from ffmpeg for the lzw decompression. >> This header had to be removed for license compliancy!!! >> >> This code was used when opening one lsm file compressed with lzw. Note >> that most of current lsm are not compressed, it seems that only old >> lsm files were using it (yet to be confirmed). >> So when opening lsm files, you see weird images on the screen, the >> reason could be that it was lzw compressed... In such a case, keep use >> inform!!! >> >> We need to figure out one way to solve this issue: either by looking >> at an LZW implementation with right license, either by reimplementing >> it, either by reimplementing the lsm reader... >> >> Cheers, >> Arnaud >> >> ------------------------------------------------------------------------------ >> This SF.net email is sponsored by Sprint >> What will you do first with EVO, the first 4G phone? >> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> _______________________________________________ >> Gofigure2-developers mailing list >> Gof...@li... >> https://lists.sourceforge.net/lists/listinfo/gofigure2-developers >> >> |
|
From: Alexandre G. <ago...@gm...> - 2010-07-09 01:04:07
|
hi arnaud, I i should have already removed this file from the project. i.e. the project did not include it and did not compile against it. It has never been a problem so far, none of the image sean feed me with was compressed. to be able to catch nicely when you otherwise would require it, I suggest you put an exception in the vtkLSM reader when it detects that the image part of the LSM is compressed. I do not remember the name of the method, but there is a method that is called after parsing the tags that tries to read the image data. This method is basically a switch that dispatch the data to be read to other methods depending on the compression and format. hope this helps. alex. On Wed, Jul 7, 2010 at 10:16 PM, Arnaud Gelas <arn...@hm...> wrote: > Hi guys, > > During the NAMIC week at Boston, Mathieu Malaterre pointed me that we > were using one header from ffmpeg for the lzw decompression. > This header had to be removed for license compliancy!!! > > This code was used when opening one lsm file compressed with lzw. Note > that most of current lsm are not compressed, it seems that only old > lsm files were using it (yet to be confirmed). > So when opening lsm files, you see weird images on the screen, the > reason could be that it was lzw compressed... In such a case, keep use > inform!!! > > We need to figure out one way to solve this issue: either by looking > at an LZW implementation with right license, either by reimplementing > it, either by reimplementing the lsm reader... > > Cheers, > Arnaud > > ------------------------------------------------------------------------------ > This SF.net email is sponsored by Sprint > What will you do first with EVO, the first 4G phone? > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first > _______________________________________________ > Gofigure2-developers mailing list > Gof...@li... > https://lists.sourceforge.net/lists/listinfo/gofigure2-developers > |
|
From: Souhait, L. <Lyd...@hm...> - 2010-07-08 18:07:19
|
Hi Kishore, Please, could you explain in more details how you delete the mesh ? Is it with the table widget ? Thanks, Lydie -----Original Message----- From: Kishore Mosaliganti [mailto:kis...@gm...] Sent: Thursday, July 08, 2010 2:05 PM To: gof...@li... Subject: [Gofigure2-developers] mesh and track Hi, I noticed a minor problem with the database mesh and track. If we want to delete a mesh, then we first need to delete the tracks associated with that mesh before deleting the mesh. Else, it creates a crash sometimes. Kishore ------------------------------------------------------------------------------ This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first _______________________________________________ Gofigure2-developers mailing list Gof...@li... https://lists.sourceforge.net/lists/listinfo/gofigure2-developers |
|
From: Kishore M. <kis...@gm...> - 2010-07-08 18:05:30
|
Hi, I noticed a minor problem with the database mesh and track. If we want to delete a mesh, then we first need to delete the tracks associated with that mesh before deleting the mesh. Else, it creates a crash sometimes. Kishore |
|
From: Antonin Perrot-A. <ant...@gm...> - 2010-07-08 15:57:09
|
The problem is solved, by modifying permissions on the .subversion
folder : if svn doesn't accept permanently the certificate, ctest get
stuck. That may be an idea for future ctest developpment : don't get
stuck if there is certificate problem : abord update, inquire or abord
execution.
> In your CMakeCache.txt
> SVNCOMMAND = /opt/loca/bin/svn
>
> and in your ctest script file, the svn command is svn? or
> /opt/local/bin/svn?
>
>
> On 07/08/2010 10:36 AM, Antonin Perrot-Audet wrote:
>> Yes, the way macports does it is by to append the opt/local/bin dir to
>> the usr/local/bin . This way you don't blow up your mac os system when
>> installing a macport application.
>> ctest seems to know where to find it as when I do a ccmake . in the
>> duild dierctory of the nightly, the variable for svn are correct :
>> SVNCOMMAND /opt/local/bin/svn
>> Subversion_SVN_EXECUTABLE /opt/local/bin/svn
>>
>>> You mentioned you use macports svn, but do you have an alias or
>>> symbolic links for svn?
>>> Or can you try with specifying the whole path to svn (something like
>>> /opt/local/bin/svn)?
>>>
>>> my 2cts,
>>> Arnaud
>>>
>>> On 07/08/2010 10:23 AM, Antonin Perrot-Audet wrote:
>>>> I got it to almost work by replacing the "svn" command by "svn
>>>> up" in
>>>> the ctest script file. It seems like it simply skips the update,
>>>> because
>>>> the command is wrong... which is definitely what we don't want.
>>>> ctest says :
>>>> Gathering version information (one . per revision):
>>>>
>>>> Update command failed: "svn up" "update" "--non-interactive"
>>>> "-r{2010-07-08 03:00:00 +0000}"
>>>> and goes on instead of getting stuck.
>>>> I use the macport svn (svn, version 1.6.12 (r955767)).
>>>>
>>>>
>>>>
>>>> here is the verbose output :
>>>>
>>>>
>>>> * Extra verbosity turned on
>>>> Reading Script:
>>>> /Users/antonin/CVSROOT/gofigure2/ctest_gofigure_shared_nightly.ctest
>>>>
>>>> SetCTestConfiguration:SourceDirectory:/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>>
>>>>
>>>>
>>>> SetCTestConfiguration:BuildDirectory:/Users/antonin/CVSROOT/gofigure2/Nightly
>>>>
>>>>
>>>> Run dashboard with model Nightly
>>>> Source directory:
>>>> /Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>> Build directory: /Users/antonin/CVSROOT/gofigure2/Nightly
>>>> Reading ctest configuration file:
>>>>
>>>> /Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk/CTestConfig.cmake
>>>>
>>>> SetCTestConfigurationFromCMakeVariable:NightlyStartTime:CTEST_NIGHTLY_START_TIMESetCTestConfiguration:NightlyStartTime:22:00:00
>>>>
>>>>
>>>> EST
>>>>
>>>> SetCTestConfigurationFromCMakeVariable:Site:CTEST_SITESetCTestConfiguration:Site:wahoofish.megason-lab.hms
>>>>
>>>>
>>>>
>>>> SetCTestConfigurationFromCMakeVariable:BuildName:CTEST_BUILD_NAMESetCTestConfiguration:BuildName:i686-apple-darwin10-gcc-4.2.1-shared-dbg
>>>>
>>>>
>>>> Site: wahoofish.megason-lab.hms
>>>> Build name: i686-apple-darwin10-gcc-4.2.1-shared-dbg
>>>> Determine Nightly Start Time
>>>> Specified time: 22:00:00 EST
>>>> Use Nightly tag: 20100708-0300
>>>>
>>>> SetCTestConfiguration:SourceDirectory:/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>>
>>>>
>>>>
>>>> SetCTestConfigurationFromCMakeVariable:UpdateCommand:CTEST_UPDATE_COMMANDSetCTestConfiguration:UpdateCommand:svn
>>>>
>>>>
>>>> up
>>>>
>>>> SetCTestConfiguration:BuildDirectory:/Users/antonin/CVSROOT/gofigure2/Nightly
>>>>
>>>>
>>>>
>>>> SetCTestConfiguration:SourceDirectory:/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>>
>>>>
>>>> Updating the repository:
>>>> /Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>> Use SVN repository type
>>>> Old revision of repository is:
>>>> Determine Nightly Start Time
>>>> Specified time: 22:00:00 EST
>>>> New revision of repository is:
>>>> Gathering version information (one . per revision):
>>>>
>>>> Update command failed: "svn up" "update" "--non-interactive"
>>>> "-r{2010-07-08 03:00:00 +0000}"
>>>>
>>>> SetCTestConfiguration:BuildDirectory:/Users/antonin/CVSROOT/gofigure2/Nightly
>>>>
>>>>
>>>>
>>>> SetCTestConfiguration:SourceDirectory:/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>>
>>>>
>>>> SetCTestConfiguration:ConfigureCommand:"/opt/local/bin/cmake"
>>>> "-GUnix
>>>> Makefiles" "/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk"
>>>>
>>>> SetCTestConfiguration:BuildDirectory:/Users/antonin/CVSROOT/gofigure2/Nightly
>>>>
>>>>
>>>>
>>>> SetCTestConfiguration:SourceDirectory:/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>>
>>>>
>>>> Configure project
>>>> Configure with command: "/opt/local/bin/cmake" "-GUnix Makefiles"
>>>> "/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk"
>>>> Run command: "/opt/local/bin/cmake" "-GUnix Makefiles"
>>>> "/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk"
>>>> Each . represents 1024 bytes of output
>>>> .-- The C compiler identification is GNU
>>>> -- The CXX compiler identification is GNU
>>>>
>>>>
>>>> And so on....
>>>>
>>>>
>>
>> ------------------------------------------------------------------------------
>>
>> This SF.net email is sponsored by Sprint
>> What will you do first with EVO, the first 4G phone?
>> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
>> _______________________________________________
>> Gofigure2-developers mailing list
>> Gof...@li...
>> https://lists.sourceforge.net/lists/listinfo/gofigure2-developers
>
--
Antonin Perrot-Audet
Electrical Engineering& Computer Sciences, INSA Lyon
M2 GEGP Signal& Image Processing, INSA Lyon, UCBL, Centrale Lyon
|
|
From: Arnaud G. <arn...@hm...> - 2010-07-08 15:20:49
|
Can you retry with "permanently accept", please? On 07/08/2010 11:19 AM, Antonin Perrot-Audet wrote: > no, the update command still fail. > that's what the wiki says : > To use SVN, just change the CVSCOMMAND line by SVNCOMMAND and give the > right path to the svn.exe program > and when I do so : > set (CTEST_UPDATE_COMMAND "/opt/local/bin/svn") > I go back to my initial problem : ctest get stuck because of the svn call. > There is something to note tough Mac osx 10.6 has a problem with > subversion : one always has to accept the certificate, the "permanently > accept" doesn't work. It is a well known issue. Maybe it gets stuck > because of that ? > > |
|
From: Antonin Perrot-A. <ant...@gm...> - 2010-07-08 15:19:09
|
no, the update command still fail. that's what the wiki says : To use SVN, just change the CVSCOMMAND line by SVNCOMMAND and give the right path to the svn.exe program and when I do so : set (CTEST_UPDATE_COMMAND "/opt/local/bin/svn") I go back to my initial problem : ctest get stuck because of the svn call. There is something to note tough Mac osx 10.6 has a problem with subversion : one always has to accept the certificate, the "permanently accept" doesn't work. It is a well known issue. Maybe it gets stuck because of that ? -- Antonin Perrot-Audet Electrical Engineering& Computer Sciences, INSA Lyon M2 GEGP Signal& Image Processing, INSA Lyon, UCBL, Centrale Lyon |
|
From: Arnaud G. <arn...@hm...> - 2010-07-08 14:42:09
|
In your CMakeCache.txt
SVNCOMMAND = /opt/loca/bin/svn
and in your ctest script file, the svn command is svn? or
/opt/local/bin/svn?
On 07/08/2010 10:36 AM, Antonin Perrot-Audet wrote:
> Yes, the way macports does it is by to append the opt/local/bin dir to
> the usr/local/bin . This way you don't blow up your mac os system when
> installing a macport application.
> ctest seems to know where to find it as when I do a ccmake . in the
> duild dierctory of the nightly, the variable for svn are correct :
> SVNCOMMAND /opt/local/bin/svn
> Subversion_SVN_EXECUTABLE /opt/local/bin/svn
>
>
>> You mentioned you use macports svn, but do you have an alias or
>> symbolic links for svn?
>> Or can you try with specifying the whole path to svn (something like
>> /opt/local/bin/svn)?
>>
>> my 2cts,
>> Arnaud
>>
>> On 07/08/2010 10:23 AM, Antonin Perrot-Audet wrote:
>>
>>> I got it to almost work by replacing the "svn" command by "svn up" in
>>> the ctest script file. It seems like it simply skips the update, because
>>> the command is wrong... which is definitely what we don't want.
>>> ctest says :
>>> Gathering version information (one . per revision):
>>>
>>> Update command failed: "svn up" "update" "--non-interactive"
>>> "-r{2010-07-08 03:00:00 +0000}"
>>> and goes on instead of getting stuck.
>>> I use the macport svn (svn, version 1.6.12 (r955767)).
>>>
>>>
>>>
>>> here is the verbose output :
>>>
>>>
>>> * Extra verbosity turned on
>>> Reading Script:
>>> /Users/antonin/CVSROOT/gofigure2/ctest_gofigure_shared_nightly.ctest
>>> SetCTestConfiguration:SourceDirectory:/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>
>>> SetCTestConfiguration:BuildDirectory:/Users/antonin/CVSROOT/gofigure2/Nightly
>>>
>>> Run dashboard with model Nightly
>>> Source directory:
>>> /Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>> Build directory: /Users/antonin/CVSROOT/gofigure2/Nightly
>>> Reading ctest configuration file:
>>> /Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk/CTestConfig.cmake
>>> SetCTestConfigurationFromCMakeVariable:NightlyStartTime:CTEST_NIGHTLY_START_TIMESetCTestConfiguration:NightlyStartTime:22:00:00
>>>
>>> EST
>>> SetCTestConfigurationFromCMakeVariable:Site:CTEST_SITESetCTestConfiguration:Site:wahoofish.megason-lab.hms
>>>
>>> SetCTestConfigurationFromCMakeVariable:BuildName:CTEST_BUILD_NAMESetCTestConfiguration:BuildName:i686-apple-darwin10-gcc-4.2.1-shared-dbg
>>>
>>> Site: wahoofish.megason-lab.hms
>>> Build name: i686-apple-darwin10-gcc-4.2.1-shared-dbg
>>> Determine Nightly Start Time
>>> Specified time: 22:00:00 EST
>>> Use Nightly tag: 20100708-0300
>>> SetCTestConfiguration:SourceDirectory:/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>
>>> SetCTestConfigurationFromCMakeVariable:UpdateCommand:CTEST_UPDATE_COMMANDSetCTestConfiguration:UpdateCommand:svn
>>>
>>> up
>>> SetCTestConfiguration:BuildDirectory:/Users/antonin/CVSROOT/gofigure2/Nightly
>>>
>>> SetCTestConfiguration:SourceDirectory:/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>
>>> Updating the repository:
>>> /Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>> Use SVN repository type
>>> Old revision of repository is:
>>> Determine Nightly Start Time
>>> Specified time: 22:00:00 EST
>>> New revision of repository is:
>>> Gathering version information (one . per revision):
>>>
>>> Update command failed: "svn up" "update" "--non-interactive"
>>> "-r{2010-07-08 03:00:00 +0000}"
>>> SetCTestConfiguration:BuildDirectory:/Users/antonin/CVSROOT/gofigure2/Nightly
>>>
>>> SetCTestConfiguration:SourceDirectory:/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>
>>> SetCTestConfiguration:ConfigureCommand:"/opt/local/bin/cmake" "-GUnix
>>> Makefiles" "/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk"
>>> SetCTestConfiguration:BuildDirectory:/Users/antonin/CVSROOT/gofigure2/Nightly
>>>
>>> SetCTestConfiguration:SourceDirectory:/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk
>>>
>>> Configure project
>>> Configure with command: "/opt/local/bin/cmake" "-GUnix Makefiles"
>>> "/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk"
>>> Run command: "/opt/local/bin/cmake" "-GUnix Makefiles"
>>> "/Users/antonin/CVSROOT/gofigure2/SVN/gofigure2/trunk"
>>> Each . represents 1024 bytes of output
>>> .-- The C compiler identification is GNU
>>> -- The CXX compiler identification is GNU
>>>
>>>
>>> And so on....
>>>
>>>
>>>
>>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Gofigure2-developers mailing list
> Gof...@li...
> https://lists.sourceforge.net/lists/listinfo/gofigure2-developers
>
|
