[Announce] GPGME 0.4.2 released
Thu Jul 31 20:08:06 2003
On Thu, Jul 31, 2003 at 07:39:03PM +0200, Albrecht Dre? wrote:
> Am 30.07.03 16:07 schrieb(en) Marcus Brinkmann:
> >We are pleased to announce version 0.4.2 of GnuPG Made Easy,
> >a library designed to make access to GnuPG easier for applications.
> >It may be found in the file (about 701 KB compressed)
> Do you think the api is now stable for the forseeable future? I would like
> to port the gpg support in the mail client balsa to the new version (now
> it's 0.3.14 or .15) to get rid of some nasty effects, but I would prefer
> to do it only *once* (started with 0.4.0, and then the api changed again
Well, 0.4.0 didn't change much. However, had I known at that point the
amount of changes I'd do in 0.4.1, I'd have put in a word of warning. Sorry
The API should for now be quite stable, and managed by extensions
rather than changes. The only thing that worries me now is that the
structures used for user input, in particular the user data callback and the
io callback structures, are not extensible. However, this can be overcome
by creating new functions taking extended versions of these structs, so
there is no serious issue in this.
There are some exceptions, though:
* Bug fixes are bug fixes. We are willing to stretch the definition of a
bug fix a bit in our favour, but we try to be careful to not break
anything gratuitously. If an interface is not clear or has strange
behaviour, it's better to ask here if it is how it is supposed to be rather
than relying on possibly buggy behaviour of the code.
* All deprecated interfaces might be removed at some point.
* The trust item interface is marked as experimental in the manual, and
will at some point be redesigned. However, for now it is "stable" as
nobody is working on it ;)
* The gpgme_op_edit interface is undocumented on purpose, as it is
originally only intended for use by GPA. However, as it is in fact
actively used, you can inofficially rely on it. But you have been warned ;)
As it is sort of a break of GPGME's original design and purpose, we truly
want to get rid of it, but that requires first to implement proper
interfaces for the things you want to do with it, and I don't see this
happening right now.
`Rhubarb is no Egyptian god.' GNU http://www.gnu.org firstname.lastname@example.org
Marcus Brinkmann The Hurd http://www.gnu.org/software/hurd/