Danijel,
>I import a directory containing certain text file to CVS repository. When I do the
>checkout of the same directory, I get a copy of that file with different size than
>before it was imported, and when running diff tool on the two files, I get that the
>files differ "totally", with the same text being recognized as different. I guess the
>problem is some text converting,
I'm willing to bet that the file you had problems with somehow is the only one in the
set (I assume from your statement there were more) that has Windows/DOS style CR/LF
line endings.
These must have gotten comitted via the import (and not filtered in any way), but when
you do the checkout / diff, they did get filtered at one stage or another. I have seen
similar things; files not showing differences in some environments or client/server OS
combinations and (like in your case) 100% differences in others.
Try to find the original file you imported and open it in vi in a UNIX environment.
That will clearly show the line endings. (unless you're using Vim in the
too-smart-for-its-own-good setting; you MAY have to read the odd man page :)
I found that the cleanest way to solve these things (even though that goes against
some of CVS's record keeping aspects) is to export the file in question, physically
delete it from the repository (if you don't know how to do that, you had better not
try it without supervision :), clean it in a UNIX editor and re-add it. Any tags can
be removed beforehand and re-applied afterwards.
This obviously doesn't work very well for files that have dozens or more revisions
that you want to hang on to for some reason. In your case, it seems to be just after
an import, so it could be applicable. (caveat emptor)
Hope this helps,
Schmolle
_______________________________________________
Info-cvs mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/info-cvs
