info-cvs
[Top][All Lists]
Advanced

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

Case insensitivity ad nauseum


From: Derek Robert Price
Subject: Case insensitivity ad nauseum
Date: Tue, 04 Nov 2003 13:57:25 -0500
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

Hey all,

I was investigating case sensitivity, thinking that it was somehow
necessary for the support of Windows clients, when it occurred to me
that the _only_ value-added feature this provides is that Windows users
don't need to specify the case of files and paths correctly in remote
commands.

That's right folks, only in remote commands.  Checkout.  rlog, rdiff,
rtag, etc.  Clients keep track of the server case for filenames for
files that exist in the local workspace and return it to the server when
users specify local files, so there isn't a problem there.  This
behavior is actually required by the CVS client/server protocol
specification.  I just confirmed that the entirety of the sanity.sh test
suite passes from a case insensitive client to a case sensitive server
with case insensitivity support disabled once the minor tweaks to enable
the server-to-server test via SSH and a shared TESTDIR on an NFS share
are put in place.

So anyway, why _don't_ we remove the case-insensitivity support?  It
seems to me that it has been causing an awful lot of headaches without a
very good reason for supporting it.  So what if Windows users have to
specify project case correctly during the initial checkout?  They had to
anyhow when conflicting cases existed in the repository.

I've got a little more cleanup and documentation to write before I check
in the test suite modifications to enable server-to-server testing, but
if anyone would like to verify what I've said, I'll be happy to send a
patch of things in their current state.

I'm also going to try and write a few tests to test conflicting case
issues in particular since these weren't tested for when there was no
way to run the suite with a case insensitive client and a case sensitive
server.  I'll likely concentrate on the recent case-insensitivity bugs I
have fixed in the source.  If anyone has any thoughts on test cases I
might otherwise miss, I'll be happy to listen to those too.

Cheers,

Derek

- --
~                *8^)

Email: address@hidden

Get CVS support at <http://ximbiot.com>!
- --
You can't fool me, sonny!  It's turtles all the way down!
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)
Comment: Using GnuPG with Netscape - http://enigmail.mozdev.org

iD8DBQE/p/aULD1OTBfyMaQRAh9sAKCL/R9rDy7dcMQVP7ZQKS31V8Lo4wCgjM8G
10JaGTCp4zc3hhrtLzST8v4=
=PLY5
-----END PGP SIGNATURE-----






reply via email to

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