freetds AT lists.ibiblio.org
Subject: FreeTDS Development Group
List archive
- From: "Brian Bruns" <camber AT ais.org>
- To: freetds AT franklin.oit.unc.edu
- Subject: Re: [ freetds-Patches-596008 ] trusted tds7 login
- Date: Mon, 19 Aug 2002 11:25:13 -0400
> One thing I haven't done yet (among many) is change the default --prefix to
> be /usr/local and the default --sysconfdir to /etc/freetds. I think there's
> broad consensus that those are OK choices, and that someone using those
> defaults would avoid mucking with ldconfig and friends. And, as Brian said,
> we want to make this change once and only once.
0.60 is full of big changes and so the version bump is well earned.
> As regards the libtds API, I think it's unrealistic to try to freeze it at
> this stage of the game. It's not even a public interface, and I think it
> will stabilize naturally as the client APIs are completed. But if it's an
> issue, perhaps our numbering scheme could reflect libtds changes, such as
> odd-numbered releases don't change it, but even-numbered ones might?
It's not a public interface, no, but I'd like to minimize the number of
releases for which we break it. So, if we break it for .53 to .60 then,
well that's ok, but knowingly leaving api changes out of .60 and
rebreaking it in 0.61 is just causing aggravation for third parties.
We've always said we'd change it if we needed but wouldn't do it
willy-nilly. Better to do a bunch of changes now than incrementally
rebreaking it over the next few releases. I'd reccommend doing 0.x0
releases as API changes but we'll run out of numbers before 1.0 ;-)
> As above, from the project's point of view, complete datatype handling is a
> big feature. Of run-time problems that appear on this list, I bet half of
> them are datatype related.
>
> Still, I agree with Brian's reasoning about domain logins. Many people have
> asked about it, and it would make FreeTDS useful to a larger audience. When
> you got it working, my reaction was, "woohoo!" As long as it's off by
> default, no one will trip over it by accident (well, most people won't).
> More is more. It doesn't interfere with datatype conversions and bcp,
> right? Then let's put it in.
Beyond login it has absolutely no impact. We were close before and the
code was in there not doing much, so Freddy's patch is filling in the
blanks more than structural items.
>
> --jkl
Have a nice vacation?
Brian
-
[ freetds-Patches-596008 ] trusted tds7 login,
noreply, 08/16/2002
- <Possible follow-up(s)>
- Re: [ freetds-Patches-596008 ] trusted tds7 login, Brian Bruns, 08/16/2002
- Re: [ freetds-Patches-596008 ] trusted tds7 login, Frediano Ziglio, 08/16/2002
- Re: [ freetds-Patches-596008 ] trusted tds7 login, Brian Bruns, 08/16/2002
- Re: [ freetds-Patches-596008 ] trusted tds7 login, Varley, David(CBorn at Alcoa), 08/16/2002
- Re: [ freetds-Patches-596008 ] trusted tds7 login, Frediano Ziglio, 08/17/2002
- Re: [ freetds-Patches-596008 ] trusted tds7 login, Varley, David(CBorn at Alcoa), 08/17/2002
- Re: [ freetds-Patches-596008 ] trusted tds7 login, Lowden, James K, 08/19/2002
- Re: [ freetds-Patches-596008 ] trusted tds7 login, Brian Bruns, 08/19/2002
- Re: [ freetds-Patches-596008 ] trusted tds7 login, ZIGLIO Frediano, 08/19/2002
Archive powered by MHonArc 2.6.24.