FreeTDS Development Group

["Craig A. Berry" <craigberry AT mac.com>]

On Mar 26, 2009, at 8:45 AM, James K. Lowden wrote:

> The *assumption* that MD5 worked on the tree, I never questioned.   
> Argh!

Yipee!  I love it when wild guesses turn out to be right.  It  
partially makes up for all the things I'm supposed to know but  
don't :-).

> The way home is now short.  I think "find freetds -ls" | md5 will work
> fine.  Or, for that matter, "md5 freetds/ChangeLog".

Yes, it had occurred to me that Changelog would do the trick as long  
as it's updated consistently.  If for some reason you really did want  
to scan the entire directory tree, it might be easier with Perl and  
Digest::MD5 than with md5 and shell builtins.

> > Does CVS not have a change number, or a changeset or commit ID that
> > could be compare with a previous one?
>
> No.  Subversion uses a repository-wide version number, though, and
> Frediano and I both want to switch.  That's a future possibility:  
> instead
> of ".dev.{date}" the development version could be ".svn.{revision}".

SourceForge now apparently supports git and Mercurial, among others.   
I would encourage you to look at those before deciding on Subversion.

> > I don't actually see how make dist could work anywhere
> > unless cp means something rather different on Linux than in BSDish
> > places.
>
> You might not be aware the distribution tarballs are generated on a  
> NetBSD
> system.


I think it was mentioned in your snapshot script but I had forgotten  
it when I wrote that. Basically on OS X

% cp foo/file.dat bar/

fails when the bar directory does not exist.  Maybe on NetBSD it  
automatically creates intervening directories or fails silently?

________________________________________
Craig A. Berry
mailto:craigberry AT mac.com

"... getting out of a sonnet is much more
 difficult than getting in."
                 Brad Leithauser