[Qgis-developer] Re: Proposal: Lift feature freeze for labelling

[Tim Sutton]

Hi Martin

2008/10/3 Martin Dobias <wonder.sk at gmail.com>:
> Hi Tim
>
> On Thu, Oct 2, 2008 at 11:45 PM, Tim Sutton <tim at linfiniti.com> wrote:
>> Hi all
>>
>> If you cast your mind back to our feature plans for QGIS 1.0, one
>> sorely needed improvement is a better labelling engine. During the
>> FOSS4G conference, I came across just such a project [1] where they
>> have implemented just such a labelling engine in C++. This api also
>> has a JNI interface and has been used in sextant / gvSIG via an
>> extension (see [2]). Martin I know you had plans to do some work in
>> this area - have you started anything yet?
>
> I must say I'm impressed by this project. Looking at the slides from
> foss4g presentation, it looks they've implemented quite a lot from the
> things I would implement by myself and even in the way I wanted to do
> it too (e.g. they're referencing A. Wolff's combinatorial approach).
>
> I haven't started with labeling implementation, I've just considered
> doing that for my master thesis... but it looks like I'll have to
> search for another topic :-)
>
>> Since 1.0 will promise a stable API, and we have a golden opportunity
>> here to incorporate proper labelling before we finalise the 1.0 API, I
>> would like to propose that we revoke the feature freeze termporarily
>> in order to allow us to incorporate the work of the PAL library.
>> Marco and I will look at the feasibility of incorporating the PAL
>> library into QGIS, and what kind of effect it would have on the API
>> during the FOSS4G code sprint on saturday.
>
> Incorporating PAL would be probably a good step forward. But I can
> imagine that we could do that also as an incremental change - although
> that would mean that we keep the old labeling engine in place.
>

So the idea would be based on our discussions with them (PAL) tomorrow
we could see if any API changes are required. If API changes are
required my proposal would be to do it now before 1.0 release so that
we can live up to our promise of a stable long supported API. However
if it looks like it will be feasible to do it with no API changes, my
feeling would also be to slot it in as a new feature for 1.1.


> One thing I'm affraid of is the license - GPLv3. Any ideas about v2
> and v3 compatibility?

Can anyone else comment on this? I'm not sure what the implications are....

Regards

Tim

>
> Regards
> Martin
>



-- 
Tim Sutton - QGIS Project Steering Committee Member (Release  Manager)
==============================================
Visit http://linfiniti.com to find out about:
 * QGIS programming services
 * Mapserver and PostGIS based hosting plans
 * FOSS Consulting Services
Skype: timlinux Irc: timlinux on #qgis at freenode.net
==============================================