bug-cvs
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: "Move file away, it is in the way" errors


From: Derek Robert Price
Subject: Re: "Move file away, it is in the way" errors
Date: Wed, 15 Oct 2003 10:50:09 -0400
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030624 Netscape/7.1

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Paul Edwards wrote:

|"Derek Robert Price" <address@hidden> wrote in message
news:address@hidden
|
|>I meant the other move-file-away cases.  I believe you complained some
|
|
|It didn't say "move file away", it said "already added".


Well, yes, but they are basically the same conflict.  One is caused by a
merge from a branch to a branch and one by a merge from the repository
to the workspace is all.  In both cases the file was added at both the
source and the destination.

Hrm.  Excuse me.  I think the other conflict case is the "added
independently by third party" case.  The "in the way" error should
probably be left alone since if the user did not specify the add then
CVS shouldn't assume the merge.  It's possible that the "in the way"
message could be improved to something along the lines of "move the file
or add it".

|>time back that during a merge from a branch to a branch where a file
|>with the same name had been added to both branches, CVS reported a
|>conflict to you that could be accidentally lost track of by touching the
|>file in question because no merge markers were put in the file.
|>
|>Larry and I were discussing the possibility of using diff's if-then-else
|>syntax to specifiy a two-way-merge with merge markers on info-cvs
|>instead of just printing the conflict warning or warning the user to
|>move the file away in the update merge case.  I don't have time to
|>implement this just now but I thought you might have enough interest to
|>generate the patch.
|
|
|I did have enough interest to generate a patch that would at
|least mark the file as a conflict, which in my opinion is a hell
|of a lot better than the current behaviour.
|
|I submitted it with Changelog entry and sanity.sh already,
|months ago.


I think the general concensus was that your original patch wasn't enough
of an enhancement to be worth the trouble.  I simply thought you might
have enough interest to update it.  The change should be fairly
striaghtforward - you should just need to add a call to the the diff
code which calls a two way diff using diff's if-then-else syntax to get
the merge markers into the generated file.  It might be nice to check
the generated file size and if it is too close to the sum of the two
input file sizes then just insert the entire contents of each file
between merge markers in the new file.  You might need to call something
to notify the server that the new file needs to be sent to the client too.

The exact same code should be insertable whereever the "added
independently" message comes from.  And add would likely need a hook
that didn't refuse to add the file just because someone else already
had, but that should just involve removing a few lines of code and maybe
adding a warning.

If you don't have enough interest to generate the patch, don't worry
about it.  I just don't have time to work more on this right now and
thought you might.

Derek

- --
~                *8^)

Email: address@hidden

Get CVS support at <http://ximbiot.com>!
- --
Sometimes it is said that man cannot be trusted with the government of
himself.  Can he, then, be trusted with the government of others?

           - Thomas Jefferson
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)
Comment: Using GnuPG with Netscape - http://enigmail.mozdev.org

iD8DBQE/jV6hLD1OTBfyMaQRAvE7AKDYzgJW0pUnqF8kV9JIyAbCvOadOQCg+L1Q
qErmG5qzNw3w9vYZt3Y39T4=
=05Zz
-----END PGP SIGNATURE-----






reply via email to

[Prev in Thread] Current Thread [Next in Thread]