> Hi Ed,
>
> this all seems a bit wacky. Normally if you see the below I'd just do:
>
> ./update-modules
>
> and if that refused and I knew I had no local modifications in the
> submodule that I wanted to keep, I'd do:
>
> ./update-modules -f
>
> Usually you would only see these submodules as modified if they were out
> of date, but sometimes you can get locally modified files, usually
> generated ones, which the second case will fix. Maybe you have something
> setup that is causing modified files to be generated when normally this
> wouldn't happen. This could be caused by inconsistent file timestamps.
>
> If you see this again send the full output from ./update-modules which
> should show what has been modified.
>
> One final thought if you have a long lived branch then you will need to
> rebase that against master and run update-modules to get back into sync
> with changes to the expected submodule SHA. If you don't do that you'll
> constantly find the submodules look out of sync.
I just did a git pull --rebase (and I had done before):
$ git pull --rebase
Current branch master is up to date.
Also tried update-modules as before... no problem
$ ./update-modules
Updating local repository submodules:
... applications/extractor/extract/src (v2.5.0-168-g9fc5e42)
... applications/extractor/sextractor (v2.5.0-100-gc921cee)
... applications/smurf/libacsis/specwrite (subversion-head-15-g7af6b1c)
... buildsupport/autobuild (v1.0.2-33-g1c36ebe)
... thirdparty/apache/xerces-c/xerces-c (v3.0.1-4-g65e349f)
... thirdparty/eso/skycat/skycat (v3.0.2-517-g0626978)
... thirdparty/fsf/autoconf (v2.59-62-gdcd4f21)
... thirdparty/fsf/automake (v1.9.6-160-gddc7d5e)
... thirdparty/fsf/gsl/gsl (v1.12-4-g3e67eb0)
... thirdparty/fsf/libtool (v1.5.18-53-g5d7cbaf)
... thirdparty/fsf/m4 (v1.4.12-50-gb0ddff5)
... thirdparty/heasarc/cfitsio/cfitsio (v3.13-17-g744c584)
... thirdparty/ijg/jpeg/jpeg (v6b-4-g0d05c5b)
... thirdparty/kitware/vtk/VTK (v5.4.2-15-g2799983)
... thirdparty/kitware/vtk/cmake (v2.6.4-9-gcf9e384)
... thirdparty/latex2html/latex2html/latex2html (v2002.2-13-geee1edc)
... thirdparty/mesa3d/mesa/mesa (v7.0.1-3-g4b4ea04)
... thirdparty/mit/fftw/fftw (v3.2.1-4-g79b3f65)
... thirdparty/norman/dvi2bitmap/dvi2bitmap (v18012008-1-gd858c73)
... thirdparty/perlsys/perl/perl (perl-5.10.1)
... thirdparty/rit/match/match (v0.10-6-g4e60352)
... thirdparty/tclsys/blt/blt (v3.0a-56-g3628f25)
... thirdparty/tclsys/itcl/itcl (v3.4-33-g7b714ad)
... thirdparty/tclsys/itcl/itk (v3.4-33-g6649f82)
... thirdparty/tclsys/itcl/iwidgets (v4.0.1-10-g291b8e2)
... thirdparty/tclsys/tcl (v8.5.7-61-gd4efcbf)
... thirdparty/tclsys/tclx/tclx (v8.4.0-45-g9ff2b9b)
... thirdparty/tclsys/tdom/tdom (v0.8.3-5-g108d99d)
... thirdparty/tclsys/tk (v8.5.7-67-gb17bbfb)
... thirdparty/tclsys/tkimg/tkimg (v1.3-21-gb021c86)
The output of ./update-modules -f is identical.
Then I check status
$ git status
# On branch master
# Changed but not updated:
# (use "git add <file>..." to update what will be committed)
# (use "git checkout -- <file>..." to discard changes in working directory)
#
# modified: thirdparty/apache/xerces-c/xerces-c
# modified: thirdparty/heasarc/cfitsio/cfitsio
# modified: thirdparty/latex2html/latex2html/latex2html
# modified: thirdparty/rit/match/match
# modified: thirdparty/tclsys/blt/blt
# modified: thirdparty/tclsys/itcl/iwidgets
# modified: thirdparty/tclsys/tclx/tclx
# modified: thirdparty/tclsys/tkimg/tkimg
#
no changes added to commit (use "git add" and/or "git commit -a")
And I see these same "dirty" messaged in gitk with the SHA1 as before.
Is there any way that something in my built $STARLINK tree hanging around
from before could cause this? It's the only think I can think of since I
completely wiped thirdparty and started over (literally sudo rm -r
thirdparty, then git checkout thirdparty, update-modules...).
Bizarre.
thanks,
Ed
|