Open GoogleCodeExporter opened 8 years ago
I have recently set almost all files in the svn to be CRLF and also converted
them to CRLF explicitly, assuming that all editors should detect this and svn
diff should also automatically take care of this. The svn checkout should check
out the files with CRLF.
Windows with the least capable tools is the common denominator here and
dictates the line ending, all other platforms should not have problems with
such files.
Original comment by prof7...@gmail.com
on 16 Jan 2011 at 12:33
if i would set the eol-style to native i would have all files with LF in my
working copy (linux) but the windows zip is currently built from the same
working copy that is also mounted as a network share. So I set svn:eol-style to
CRLF to enforce it everywhere.
Original comment by prof7...@gmail.com
on 16 Jan 2011 at 12:36
How do you deal with insert_missing.py? Just run unix2dos on it afterward?
Original comment by schecko...@gmail.com
on 16 Jan 2011 at 12:55
i haven't run it since I converted all to CRLF a few weeks ago ;-)
I guess I have to fix insert_missing.py
Original comment by prof7...@gmail.com
on 16 Jan 2011 at 1:59
it was broken anyways (the backup copies of the changed files). fixed in r471.
Original comment by prof7...@gmail.com
on 16 Jan 2011 at 2:21
once I have reorganized some things and some workflows here and have a separate
dedicated windows machine instead of only a small VM on my laptop I might set
it all back to native in the svn properties and instead use two completely
separate working copies here, one for windows and one for linux but at the
moment this would seriously disrupt the way I am doing some things here or make
them unnecessarily hard or complicated.
Original comment by prof7...@gmail.com
on 16 Jan 2011 at 2:27
I will use this bug as a reminder to change insert_missing.py to detect and use
the proper line endings once I change the whole repository back to eol-style
native. But this has low priority
Original comment by prof7...@gmail.com
on 16 Jan 2011 at 2:38
Original issue reported on code.google.com by
schecko...@gmail.com
on 16 Jan 2011 at 12:11