I am jumping back into the discussion (sorry I have been off sick for a
bit). Andrea has already furnished you with a reply which I have not read
yet; so please view two replies as the start of an interesting discussion -
we are after all interested in working with you.
On Tue, Feb 17, 2009 at 12:42 AM, Milton Jonathan <[email protected]
> wrote:
> Hello people
>
> Sorry for the fast coding! We just wanted to see something working to give
> us a better idea. But it's definitely not meant to be kept as it is, just to
> start the process for us (especially since we are no SLD-experts).
>
> OK, so let's see if I gather everything here:
> - Ideally, we should move to 2.6 soon. I have two questions about that: is
> it as stable as 2.5.x right now?
You would be the main reason for the instability :-) Working on the style
interfaces is the only technical direction for the 2.6.x branch. You will
find a lot of fixes have been done to the SLDParser addressing many problems
around the handling of Expressions for example.
> And is there much to change to make the move?
The use of geoapi interfaces for style are the only changes; and the geoapi
interfaces were started by copying the geotools ones and removing the set
methods so it should not be too bad. I still want to see a real user (ie
you) try out the geoapi StyleFactory before we decide it is great. You can
see why I am enthusiastic - much of what is needed here is feedback from a
real team.
> I ask this because we want to make a first release of our projects using
> GeoTools in about two months time, so.. Either we do it now, or later on,
> around May-June. And a side-effect question is: does anyone want support for
> this in 2.5.x?
2.5.x is closed for new development (people are working on additional
formats and stuff but the API is fixed). So in a word - no.
Given a couple days warning I am willing to release a 2.6.x milestone
release for you guys; as you approach your deadline and have something you
are happy with.
>
> - Nice thing about the XML bindings! We were actually looking for that, and
> quite puzzled with the whole SLDParser and SLDTransformer thing (why reading
> and writing was done by separate classes, why it didn't use the xsd
> definitions, etc, etc). But I understand it's not in use right now, and
> there is a lot missing, right? Or is that just for 2.5.x? Are things more
> developped in 2.6?
The SLDParser / SLDTransformer has been fixed up for 2.6.x (when we started
using functions for line width or color definitions we ran into trouble and
had to supply bug fixes). These are used by uDig 1.2 right now. The XML
binding stuff is cool; but have not had a chance to work with it much yet -
it is far more capable (and can handle literals expressed in GML3 for
example).
>
>
> - I really like the idea of working with actual Units. Sorry I didn't
> notice that before: it is certainly the right way to do it, be it 2.5.x or
> 2.6, SLDParser or SLD bindings.
Yeah the JSR-275 Units stuff is great; I am happy working with either the a
copy of the SLDParser (copy it and call it SEParser) or constructing SE
bindings. If we construct SE bindings we may want to warn Justin and ask him
for a couple hours help getting started on the right foot (and be sure to go
through the tutorials in that document 1st).
>
>
> So, from my point of view, our next moves back here should be:
> 1. Determine if we can move right away to 2.6. If you have any hints on how
> tough that could be (and how we could make it easier), please tell me. Truth
> is, we sticked to 2.5.x since the beginning since it was labelled as the
> "stable release".
Yep; I really should label the reason the 2.6 release is unstable - it would
of helped you with your planning. I will update the home page now to
indicate that 2.6 is unstable because we are working on the style
interfaces.
>
>
> 2. Determine if we can move to the XML bindings framework. That depends on
> how much work we need to do to start to get things going. I guess it also
> depends on the state of the bindings for Filter.
All of that is fine; and the test cases pass; the bindings for Filter and
SLD work (the question is how much work will it be to make bindings for SE).
>
> In any case, when we decide which way to go, we will reimplement things
> using the GeoAPI interface. What do you guys think?
It will be a pleasure to work with you on this.
Jody
------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H
_______________________________________________
Geotools-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-devel
