On Thu, 11 Feb 2021 at 11:24, Dima Pasechnik <dimp...@gmail.com> wrote:
>
> It could also be that core.autocrlf is true by default for you.
> (this is the option that does automatic convesion to LF from CRLF)
> Try
>
> git config core.autocrlf false
I think it is false by default (someone on stackexchange actually
looked at git source code to determine this), at least on unix-like
systems. As far as I can tell the only people who might want it set
to anything else are windows users collaborating with non-windows
users in a repository.
ANyway, Frederic has fixed the offending files in #31379.
John
>
> and see if it helps
>
> On Thu, Feb 11, 2021 at 9:21 AM John Cremona <john.crem...@gmail.com> wrote:
> >
> > On Wed, 10 Feb 2021 at 18:07, Dima Pasechnik <dimp...@gmail.com> wrote:
> > >
> > > On Wed, Feb 10, 2021 at 5:49 PM John Cremona <john.crem...@gmail.com>
> > > wrote:
> > > >
> > > > On Wed, 10 Feb 2021 at 16:56, Dima Pasechnik <dimp...@gmail.com> wrote:
> > > > >
> > > > > On Wed, Feb 10, 2021 at 2:58 PM John Cremona <john.crem...@gmail.com>
> > > > > wrote:
> > > > > >
> > > > > > Has anyone else been seeing the following problem, which has been
> > > > > > plaguing me for a week or two. Here's a simple example. All
> > > > > > computers mentioned here are running ubuntu. On a machine I had not
> > > > > > used for a while I had a sage build of the develop branch at an old
> > > > > > version (pre 9.0). There were no modified files (git status showed
> > > > > > nothing). Then I did "git pull trac develop", after which one file
> > > > > > is
> > > > > > marked as having changed:
> > > > > >
> > > > > > modified: src/sage/misc/element_with_label.py
> > > > > >
> > > > > > It is always this file, on several machines where I have gone
> > > > > > through
> > > > > > similar steps. file shows this:
> > > > > > src/sage/misc/element_with_label.py: Python script, ASCII text
> > > > > > executable, with CRLF line terminators
> > > > >
> > > > > does
> > > > >
> > > > > git config core.autolf
> > > > >
> > > > > show 'true'?
> > > >
> > > > No.
> > > >
> > > > >
> > > > > Set it to false, IMHO this should fix this issue.
> > > >
> > > > It is not set at all. I saw that option after googling for help, but
> > > > the issue it exists to solve is not one which has ever hit me (in sage
> > > > anyway, of course I know about using dos2unix sometimes when a windows
> > > > users sends a file).
> > >
> > > did you try setting this option to false, and see if it helps?
> >
> > Not yet systematically. After setting it to false and doing nothing
> > else, git status still shows
> > src/sage/misc/element_with_label.py as modified, even after 'git
> > checkout --'. Trying to get back to sanity a different way, 'git
> > stash' now outputs
> >
> >
> > warning: CRLF will be replaced by LF in src/sage/misc/element_with_label.py.
> > The file will have its original line endings in your working directory.
> > warning: CRLF will be replaced by LF in src/sage/misc/element_with_label.py.
> > The file will have its original line endings in your working directory.
> > Saved working directory and index state WIP on develop: 8453ffb
> > Updated SageMath version to 9.3.beta7
> > HEAD is now at 8453ffb Updated SageMath version to 9.3.beta7
> >
> > but git status still shows that file as modified. Now I cannot just
> > checkout a different existing branch (since there are apparently
> > modified files), but I can checkout a new branch (g = alias for git)
> >
> > $ g co -b dud
> > M src/sage/misc/element_with_label.py
> > Switched to a new branch 'dud'
> >
> > the delete the 'develop' branch (which was the same as upstream anyway):
> >
> > $ g branch -d develop
> > Deleted branch develop (was 8453ffb).
> >
> > and now recreate the develop branch
> >
> > $ g remote update trac
> > (...)
> > $ g co -b develop trac/develop
> > M src/sage/misc/element_with_label.py
> > Branch develop set up to track remote branch develop from trac.
> > Switched to a new branch 'develop'
> >
> > and we are back to where we started. All of the above was with
> > core.autolf set to false.
> >
> > On the same machine, in a completely new clone made and built
> > yesterday, git status shows
> >
> > $ g st
> > On branch develop
> > Your branch is up-to-date with 'origin/develop'.
> > Changes not staged for commit:
> > (use "git add <file>..." to update what will be committed)
> > (use "git checkout -- <file>..." to discard changes in working directory)
> >
> > modified: src/sage/rings/invariants/__init__.py
> >
> > no changes added to commit (use "git add" and/or "git commit -a")
> >
> > So the same problem with a new file -- which only has 2 bytes in it!
> > AFter dos2unix-ing it (so it now has 1 byte) it is still showing up in
> > git status. AFter git checkout -- it goes back to having 2 bytes and
> > still shows up.
> >
> > It seems to be impossible to do any sage development here. I am now
> > trying yet another git clone -- so the global config of core.autolf is
> > in force as the clone is made -- but then after the clone from
> > git://github.com/sagemath/sage.git that same file is shown as
> > modified. For some reason I have been making the cones from github
> > and then adding trac as a remote. As a variant I made a fresh clone
> > from trac -- but see exactly the same.
> >
> > A similar problem was reported on stackexchange *over 9 years ago* by
> > a Mac users whose collaborators were all ubuntu users. There, it is
> > suggest to look at the .gitattributes file, and *in this fresh clone*
> > that file contains
> >
> > # Force LF normalization
> > * text=auto eol=lf
> > # except for Windows batch files
> > *.{cmd,[cC][mM][dD]} text eol=crlf
> > *.{bat,[bB][aA][tT]} text eol=crlf
> > *.diff_bin binary
> >
> > This is commit 5328570 which is 9.2, master branch, but the develop
> > branch has the same.
> >
> > To me it seems (from reading all that stackexchange page) that this is
> > not just a problem caused by something in my local configuration, but
> > something which has happened to the repository itself, which needs to
> > be fixed. But I am no expert.
> >
> > Just to add to the fun, on a second machine with the same ubuntu
> > version as the one above (16.04.7 LTS) and same git version 2.7.4, a
> > clone I made yesterday shows no problems at all, while and older clone
> > on which I had done development for a long time has some CRLF file
> > issues. The only differences in the global .gitconfig between the two
> > machines is the core.autolf which I just added to the first.
> >
> > I do have some clones without these problems so can keep on working,
> > but I also have branches I wanted to do more work on in clones which
> > are unusable.
> >
> > John
> >
> >
> > >
> > > >
> > > > So, I did not change anything (as far as I can tell -- except get
> > > > older) but my git state changed. These files stay changed even after
> > > > trying to reset them (though *I* did not change them) using git
> > > > checkout --.
> > >
> > > It's a new version of git that got you, I suppose.
> >
> > Yes, that is likely, since I did do package updates across the board
> > before this started to happen. I am seeing it on machines running:
> > ubuntu 16.04.7 LTS and git 2.7.4
> > ubuntu 18.04.5 LTS and git 2.17.1
> > ubuntu 20.04.2 LTS and git 2.25.1
> >
> > In no cases was the whole OS updated, just packages so possibly git,
> > but I don't know what git version any had before this.
> > >
> > > We've had some "fun" with this lately, as there are Sage spkgs which
> > > need patches for files
> > > with CRLF endings, and so these patches had to be dealt with somehow.
> > > It was not pleasant.
> > > See https://trac.sagemath.org/ticket/30403
> > >
> > >
> > > >
> > > > John
> > > >
> > > > >
> > > > >
> > > > > >
> > > > > > git diff shows this:
> > > > > >
> > > > > > $ git diff -w src/sage/misc/element_with_label.py
> > > > > > warning: CRLF will be replaced by LF in
> > > > > > src/sage/misc/element_with_label.py.
> > > > > > The file will have its original line endings in your working
> > > > > > directory.
> > > > > >
> > > > > > I cannot get this to go away. git stash does nothing, and I cannot
> > > > > > change branches since git thinks there is an uncommitted change. I
> > > > > > have been reduced to deleting the entire build and starting again
> > > > > > from
> > > > > > scratch.
> > > > > >
> > > > > > No, I have not knowingly changed any configuration on these
> > > > > > computers,
> > > > > > though I have done some normal package upgrades using apt.
> > > > > >
> > > > > > I just went through the same steps in another directory on the same
> > > > > > machine where I had an old build (8.7) and after "git pull trac
> > > > > > develop" there are now 5 modified files:
> > > > > >
> > > > > > modified:
> > > > > > build/pkgs/sage_sws2rst/src/test/Adding_Pictures_and_screenshots.sws
> > > > > > modified: src/doc/en/reference/combinat/media/k-rim.JPG
> > > > > > modified:
> > > > > > src/sage/ext_data/doctest/rich_output/example.png
> > > > > > modified:
> > > > > > src/sage/ext_data/notebook-ipython/logo-64x64.png
> > > > > > modified: src/sage/misc/element_with_label.py
> > > > > >
> > > > > > -- the same one as before and 3 others which are not even normal
> > > > > > text
> > > > > > files. I don't want to trash all these since I am trying to
> > > > > > resurrect
> > > > > > and complete work done on some old trac tickets.
> > > > > >
> > > > > > John
> > > > > >
> > > > > > --
> > > > > > You received this message because you are subscribed to the Google
> > > > > > Groups "sage-devel" group.
> > > > > > To unsubscribe from this group and stop receiving emails from it,
> > > > > > send an email to sage-devel+unsubscr...@googlegroups.com.
> > > > > > To view this discussion on the web visit
> > > > > > https://groups.google.com/d/msgid/sage-devel/CAD0p0K7KrVUvvkmqt6Nwhe1Tt9ef_w3ZWAqU5GK6Y6%3Dysqq3Uw%40mail.gmail.com.
> > > > >
> > > > > --
> > > > > You received this message because you are subscribed to the Google
> > > > > Groups "sage-devel" group.
> > > > > To unsubscribe from this group and stop receiving emails from it,
> > > > > send an email to sage-devel+unsubscr...@googlegroups.com.
> > > > > To view this discussion on the web visit
> > > > > https://groups.google.com/d/msgid/sage-devel/CAAWYfq10WrFPb-2wQfLVyhSSRRGdWr_eEQ7gyanQ9Pbgm6o6qQ%40mail.gmail.com.
> > > >
> > > > --
> > > > You received this message because you are subscribed to the Google
> > > > Groups "sage-devel" group.
> > > > To unsubscribe from this group and stop receiving emails from it, send
> > > > an email to sage-devel+unsubscr...@googlegroups.com.
> > > > To view this discussion on the web visit
> > > > https://groups.google.com/d/msgid/sage-devel/CAD0p0K41axYDsedDj4-Dc7TkANM7T3CmVyMmS%3DDHhEXQQZqn4A%40mail.gmail.com.
> > >
> > > --
> > > You received this message because you are subscribed to the Google Groups
> > > "sage-devel" group.
> > > To unsubscribe from this group and stop receiving emails from it, send an
> > > email to sage-devel+unsubscr...@googlegroups.com.
> > > To view this discussion on the web visit
> > > https://groups.google.com/d/msgid/sage-devel/CAAWYfq19RqXJ1oNgiTo3hX0yV06dXtMEenu-yuXjoM5S9XLhfg%40mail.gmail.com.
> >
> > --
> > You received this message because you are subscribed to the Google Groups
> > "sage-devel" group.
> > To unsubscribe from this group and stop receiving emails from it, send an
> > email to sage-devel+unsubscr...@googlegroups.com.
> > To view this discussion on the web visit
> > https://groups.google.com/d/msgid/sage-devel/CAD0p0K4PgJG0PsV1cL8agta4262O5PU%2B7QqeESdSot%3D7-x1dpw%40mail.gmail.com.
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-devel/CAAWYfq0sk-rHs_v2b-%2BaWiVb2%2B4vOAPUN9wo-qpe5QJM8Njmaw%40mail.gmail.com.
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sage-devel+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/sage-devel/CAD0p0K45MfjFfqDpD0m1X-QYGcqwuFKbSUhYDgk-Hc3FKYWhwg%40mail.gmail.com.
