Re: [Kolab-devel] [pkg-kolab] Bug#730600: Bug#730600: Bug#730600: libkolab(xml): New upstream version available

Mike Gabriel <[email protected]> Mon, 07 Jul 2014 10:03:37 +0000

Newsgroups	gmane.comp.kde.devel.kolab
Organization	DAS-NETZWERKTEAM
Message-ID	<20140707100337.Horde.6WtAMSFaJ9Lwi11FjiTMZA1__20897.0819869593$1404727436$gmane$org@mail.das-netzwerkteam.de>

Hi Jeroen,

On  Fr 04 Jul 2014 02:35:50 CEST, Jeroen van Meeuwen (Kolab Systems) wrote:

> On 2014-07-02 21:14, Mike Gabriel wrote:
>> On  Mi 02 Jul 2014 14:09:10 CEST, Sune Vuorela wrote:
>>> I want to ensure that neither of us gets to debug weird crashes if both
>>> libraries are loaded into the same application.
>>
>> @Sandro/Kolabsys: to me it feels as if Sune suggestions should be
>> implemented in libcalendaring (first there) by upstream and not by
>> some Debianic patch work. Do you see any chance that any coder at
>> kolabsys could get those namespace changes into libcalendaring?
>>
>
> Please note that libcalendaring's original purpose had been to  
> circumvent needing to provide a (near-)complete KDE stack >= 4.9 to  
> older platforms such as RHEL 5, 6 and UCS (based on Squeeze).
>
> As such, it has always been a very deliberate Frankenstein-baby and  
> we have the intention to burn it at the earliest opportunity.
>
> If the Debian version you are seeking to package this for has KDE >=  
> 4.9 (not unlikely, I reckon), then technically you should have no  
> requirement for libcalendaring / to compile libkolab{,xml} against  
> libcalendaring.
>
> That said, libcalendaring is the "lighter weight" version of what  
> libkolab needs from kdepimlibs. We are not experiencing the same  
> problems with ld / symbols on RPM-based systems where the  
> libcalendaring .so names are .0 and .0.1, while upstream's are .4  
> and .4.$x, and we're compiling libkolab{,-xml} against kdepimlibs.

We have two issues here:

  1. installation of libkolab(xml) pulls in many packages from the KDE
     desktop (tolerable, but awkward and prone to being a FUD target)
  2. Last time I tested, half a plasma-desktop started up when accessing
     my Kolab-prepped roundcube as user www-data

If 2. has been fixed (haven't got around to retest that with recent libkolab),
I guess we should attempt at living with 1.

Mike

-- 

DAS-NETZWERKTEAM
mike gabriel, herweg 7, 24357 fleckeby
fon: +49 (1520) 1976 148

GnuPG Key ID 0x25771B31
mail: [email protected], http://das-netzwerkteam.de

freeBusy:
https://mail.das-netzwerkteam.de/freebusy/m.gabriel%40das-netzwerkteam.de.xfb

_______________________________________________
devel mailing list
[email protected]
https://lists.kolab.org/mailman/listinfo/devel

signature.asc (application/pgp-signature, 819 B)