[Ubuntu] Fwd: [Live-demo] Packaging and project re-organization

[Forrest h]

> So for me this boils down to deciding what release
>> comes most closely to your final goal, start from there and try to keep
>> the difference to this platform as small as possible 

Should we see Trusty as the most desirable release to support, besides
the previous LTS release? Who is working on building versions of
ubuntugis software for trusty?

Forrest Hawes
Techpraxic Systems



On 01/02/2014 02:15 PM, Andreas Tille wrote:
> Hi Ivan,
> 
> On Thu, Jan 02, 2014 at 04:28:23PM +0100, Ivan Mincik wrote:
>> Hi all, I am very happy about this discussion a I hope that it could
>> bring us closer to our common goal.
> 
> :-)
>  
>>> this, upload to Debian and than the automatic cycle:  Debian
>>> unstable -> Ubuntu -> BioLinux starts working without additional
>>> work for any backports.  I'd imagine to implement this workflow
>>> also into the GIS world.
>>>
>>
>> I also think that this is the best work flow in general. But I guess
>> that in multi-software projects like OSGeo-Live is, there still would
>> be some packages which are very problematic to package in Debian. Also
>> it could be very high hurdle for all new software in OSGeo-Live to be
>> accepted in Debian first.
> 
> I was not talking about the next OSGeo-Live release and probably also
> next+1 will not be designed according to this workflow.  However, if we
> are doing things step by step and can try to migrate more and more
> packages to Debian also the workload to create OSGeo-Live will decrease
> step by step and thus this strategy is in the interest of all involved
> parties.
> 
>> If Live wants to have all software packaged
>> in Debs it will need some help of FPM.
> 
> Well, I have no means to stop you using FPM but you will not make me
> liking the FPM approach. ;-)
> 
>> It seems, thanks to all Debian and Ubuntu GIS maintainers, we could
>> have very good Ubuntu Trusty release and packaging projects like
>> UbuntuGIS could serve only as some kind of backports solution or
>> solution for bug fix updates which are not allowed in stable Debian
>> release. Anyway, UbuntuGIS need some redesign.
> 
> I'm too less involved in UbuntuGIS to know its current design and thus I
> can not comment whether it needs a redesign.  If you mean that it should
> you want ro profit more from the work done in Debian I think I can agree
> with this plan.
> 
>> Also, I am not sure if it is always possible to align Debian
>> transitions with Ubuntu LTS releases so well as it happened now.
> 
> What I fail to understand in this whole discussion:  If you want the
> latest and greatest upstream software relying on Ubuntu LTS (or Debian
> stable) is probably not the most reasonable platform to derive from
> since it will age to much.  From my vague knowledge about UbuntuGIS the
> latest Ubuntu release might have a reasonably recent software pool which
> might get some backports of the software where you *really* need the
> latest upstream release because your target user audience is urgently
> asking for this.  So for me this boils down to deciding what release
> comes most closely to your final goal, start from there and try to keep
> the difference to this platform as small as possible by fixing problems
> at the root == inside Debian (via Debian GIS team).
> 
> Everything else means more work for an infinite time.
> 
> Kind regards
> 
>        Andreas.
> 
>