FreeTDS Development Group

["James K. Lowden" <jklowden AT schemamania.org>]

On Wed, 18 Dec 2002 14:41:24 -0500, "Mark J. Lilback" <mark AT lilback.com>
wrote:
> Since Sybase was released for Mac OS X, I've been working closely 
> with Sybase. They've just informed me that they have been using 
> FreeTDS for some internal work, modifying ctlib to:
> 
> a) support some of the newer TDS tokens that have been introduced 
> with ASE 12.5;
> b) fix some datatype handling issues (binary);
> c) update the connection capability exchange to let the server know 
> that this client can handle new tokens possible with ASE 12.5;
> d) optional support for REAL Basic;
> e) a ctlib property to enable tds logging.

What good news this is!  I would be so glad for a channel to communicate
with Sybase's people.  We don't have an up-to-date TDS 5.0 document, and
not many of us have access to a 12.5 server, so protocol updates to TDS
5.0 are difficult to do.  

I've long said I think the best thing Sybase could do -- as a business
plan -- would be to put the TDS protocol definition and its ct-lib
implementation into the public domain.  It would mean that every web
server on the planet would work flawlessly with a Sybase server, and
Sybase could save itself the expense of continued client library
developement.  Not standard databasethink, but how else will they stay in
business?  

Also, ct-lib could really use beefing up....

> They've meant to contribute it back to the project, but have been 
> lacking the time. However, they've sent me a tarball of their 
> changes. The last mod on the included ChangeLog is "Tue Oct 15 
> 09:37:03 EDT 2002", so the code isn't that far behind (I was worried 
> it might be based on 0.53).
> 
> So, who feels like going over their code changes?

I would do it, Mark, but I'm going to France until the new year.  One
problem you might have is that we've applied src/.indent.pro to a lot of
the source code since mid-October.  Frediano's suggestion is to gindent(1)
the patch before applying....

A suggestion, if I may.  I don't know how closely you followed the whole
indent fracas.  Our practice has been this: if you find something you want
to reformat, just go ahead.  Commit the reformatted code with the cvs log
comment "indent(1)".  Then proceed as usual.  That way, when we see
"indent(1)" in the cvs log, we know that revision contains only formatting
changes.  

The real challenge from a Mac OS X standpoint is straight-through UCS-2
processing.  But that's going to take a concerted effort.  

Regards, 

--jkl