[vtk-developers] VTK Master git pull --rebase issue.

Shawn Waldon shawn.waldon at kitware.com
Fri Sep 29 14:23:19 EDT 2017


The branch to fix this is here:
https://gitlab.kitware.com/vtk/vtk/merge_requests/3372

Shawn

On Fri, Sep 29, 2017 at 1:26 AM, Andrew Maclean <andrew.amaclean at gmail.com>
wrote:

> Shawn and Brad,
>
> I can confirm that I am seeing the same thing in Linux and Windows.
> I only got alerted to this because the VTK master was suspiciously stable
> :-)
> You probably don't see this in the CTest machines as I think they do a
> delete of the source and then fetch from the master.
>
> Shawn, you may have something!
>
> There seem to be some inconsistences regarding these .gitattributes files
> in the ./Testing directories.
>
> Iin Testing there are these .gitattributes files:
> ./Testing/Data/LSDyna/.gitattributes
> ./Testing/Data/MFIXReader/.gitattributes
> ./Testing/Data/NetCDF/.gitattributes
> ./Testing/Data/SLAC/.gitattributes
> ./Testing/Data/WindBladeReader/.gitattributes
>
> The only ones being deleted when CMake runs are:
>         ./Testing/Data/MFIXReader/.gitattributes
>         ./Testing/Data/NetCDF/.gitattributes
>         ./Testing/Data/WindBladeReader/.gitattributes
> and in these directories these files register as untracked:
>         ./Testing/Data/MFIXReader/.gitattributes.md5
>         ./Testing/Data/NetCDF/.gitattributes.md5
>         ./Testing/Data/WindBladeReader/.gitattributes.md5
>
> In:
> ./Testing/Data/LSDyna/.gitattributes the file has these contents:
> bug_14713/bug_14713.d3plot.md5 vtk-is-large-data
> foam/foam.d3plot.md5 vtk-is-large-data
> foam/foam.d3plot01.md5 vtk-is-large-data
> hemi.draw/hemi_draw.d3plot.md5 vtk-is-large-data
> hemi.draw/hemi_draw.d3plot01.md5 vtk-is-large-data
> hemi.draw/hemi_draw.d3thdt.md5 vtk-is-large-data
> hemi.draw/hemi_draw.glstat.md5 vtk-is-large-data
> hemi.draw/hemi_draw.k.md5 vtk-is-large-data
> impact/d3plot.md5 vtk-is-large-data
> impact/d3plot01.md5 vtk-is-large-data
> the folders referred to in this file contain .md files of that name.
>
> ./Testing/Data/MFIXReader/ there is no .gitattributes file just
> a .gitattributes.md5 along with a bunch of .md files.
>
> Regards
>    Andrew
>
>
> On Fri, Sep 29, 2017 at 5:05 AM, Brad King <brad.king at kitware.com> wrote:
>
>> On 09/28/2017 09:17 AM, Shawn Waldon wrote:
>> > It appears that CMake thinks they are data files and is making MD5s of
>> them.
>>
>> That should only happen if some code is telling CMake that they
>> are data by mentioning them in a `DATA{}` reference somewhere.
>> That syntax does support directories and globbing, so it may be
>> by mistake.
>>
>> -Brad
>>
>
>
>
> --
> ___________________________________________
> Andrew J. P. Maclean
>
> ___________________________________________
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/vtk-developers/attachments/20170929/0cb3c4a9/attachment.html>


More information about the vtk-developers mailing list