Re: [Evolution-hackers] libecal soname change

[Matthew Barnes <mbarnes redhat com>]

On Thu, 2014-01-02 at 11:30 +0100, Patrick Ohly wrote:
> I just noticed that libecal bumped its soname to libecal-1.2.so.16 in
> EDS 3.10. The corresponding commit is:
>
> commit f30ae26320b359666b345c92405bf87f3f43250a
> Author: Matthew Barnes <mbarnes redhat com>
> Date:   Tue Jul 16 06:34:02 2013 -0400
>
>     Bump libecal / libedata-cal soname for API breaks.
>
> Matthew, can you elaborate what that break is?

The libdata-cal bump was to cover the preceding commits:

commit f9ee1477151f2747745f2e90e1a7d34bc992b931
Author: Milan Crha <mcrha redhat com>
Date:   Tue Jul 16 12:05:30 2013 +0200

    Insufficient return values from e_cal_backend_get_object/_list()
    
    These also returned ECalComponent, but it cannot hold a vCalendar,
    which is returned when the component has detached instances.

commit 9da18738a150915500a44e0e7c54d9ef1eb4f3c4
Author: Milan Crha <mcrha redhat com>
Date:   Tue Jul 16 09:01:55 2013 +0200

    Bug #704220 - Incorrect runtime check in e_data_cal_respond_send_objects()

The libecal bump looks like it may have been a mistake on my part.  I
must have been under the impression the client-side API changed as well.

Sorry about that.

Matthew Barnes