AW: [Qgis-developer] Re: composer redesign branch

[Hugentobler Marco]

Hi Martin, Tim and others

In my opinion, the proposed design changes are very important and I would like to see them in 1.0. I think it is important that proper printing is possible in 1.0. Furthermore, the rendering context is also important for the possibility to stop map rendering for large datasets. As development is taking place in a branch, there should be no risk for 1.0 if it is not ready and tested in time (but I'm quite optimistic that it will be :-) ).

>> b) who is willing to work on this

> I will be happy to participate, don't know so far about others...

I'm also offering my help.

>> and about the fact that
>> Gavin may be off at sea, having set some contractors to work on
>> composer stuff.

>Hm, I didn't know anything about Gavin's plans for the map composer...
>is the coding already taking place or what's the phase?

Good to know Gavin plans to support the print composer development. 
Actually one of my oncoming projects is also about improvements of the printing capabilities. I'm therefore very interested in a clean solution of the map composer's basic problems and of course I'm happy to help with coding and testing. 

>> One important thing is to really try to consider render performance in
>> any redisign. QGIS has been steadily getting slower at rendering
>> vectors over the last few years and it would be good to try to address
>> this before 1.0 goes out.

>Yes I completely agree, I don't want to do any changes that would make
>QGIS slower.  

This is a good point. So far I couldn't see any points in the current discussion that I would expect to have an impact on rendering performance (so no slowlyness by design).

Regards,
Marco

---------------------
Dr. Marco Hugentobler
Institute of Cartography
ETH Zurich
Technical advisor QGIS Project Steering Committee

-----Ursprüngliche Nachricht-----
Von: qgis-developer-bounces at lists.osgeo.org im Auftrag von Martin Dobias
Gesendet: Di 12.02.2008 11:53
An: Tim Sutton
Cc: Steven Bell; qgis-developer at lists.osgeo.org
Betreff: Re: [Qgis-developer] Re: composer redesign branch
 
Hi Tim

On Feb 12, 2008 10:36 AM, Tim Sutton <tim at linfiniti.com> wrote:
> Hi Martin and others
>
> Martin I agree your listed options will be a better design. Im just
> worried about time scales (we need to get 1.0 out this year - and
> ideally before sept in time for FOSS4GEO2008)

Well, since I haven't seen any release plan, I wasn't worried about
time scales... but getting 1.0 out in september doesn't look like a
challenge - depending on what all should be included in it. Tim, as
release manager, would you mind preparing such thing? (e.g. what
releases are planned, when, and possibly what features should be
targeted)

> and about the fact that
> Gavin may be off at sea, having set some contractors to work on
> composer stuff.

Hm, I didn't know anything about Gavin's plans for the map composer...
is the coding already taking place or what's the phase?


> The other nice thing about the design change is that
> raster and vector will work in similar way (both using renderers) if I
> understand things right.

Actually I've been talking mainly about vector renderers, although
rasters should be easy to do too.

> a) how long will it take

I guess the longest part will be the design stage (so that we won't
have to do the changes to it in longer time). If I start with coding I
expect to finish it in two weeks or so. In fact the code changes for
this stuff are not going to be that great as it might look like. The
changes should be done in following order:
1. rendering context - this is just matter of packing the arguments
for rendering to a class
2. renderers - time for this depends on final design
3. map composer changes - consists of separating frontend and backend
and then moving backend to core library


> b) who is willing to work on this

I will be happy to participate, don't know so far about others...

> c) do others agree to trying to get this implemented in 1.0

The question for others...


> One important thing is to really try to consider render performance in
> any redisign. QGIS has been steadily getting slower at rendering
> vectors over the last few years and it would be good to try to address
> this before 1.0 goes out.

Yes I completely agree, I don't want to do any changes that would make
QGIS slower.

Martin
_______________________________________________
Qgis-developer mailing list
Qgis-developer at lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer