Re: [Geotools-devel] Style components

[Adrian C. <ac...@gm...>]

Hello,

Jody has essentially answered you already but I'll add my take which may 
flesh out the explanation.

The style specification evolved greatly from its earlier incarnation 
which involved a fairly extensive re-design of the system in place. 
During that work the author of the new style system decided to follow 
the lead of another developer who was starting a major cleanup effort of 
the whole library. So the work has been completed but cannot, in its 
current form, be committed back to trunk; it has, among other things, 
been written against Java 1.6 rather than Java5. I believe the author is 
also waiting for an extensive code review to feel the code is of library 
quality and ready for release.

The cleanup work started for many reasons but three main ones were: 
obtaining a code base which could be built completely every night, raise 
the quality of the code to the standard needed for a library, and use 
the benefits of the Java 6 virtual machine and language constructs. For 
the former, the current version of geotools cannot be completely built 
and no one is quite sure why: for example, we need a different compiler 
to generate the javadocs from that used to generate the code. For the 
quality issue, this is something everyone is facing since the geotools 
library has been written over a long time. Other recent efforts have 
aimed to clean up other parts of the code base such as cleaning the 
access systems for data hosted in a database and more recently for the 
file based data access systems. Finally the language issue is one where 
the official project policy regarding language does not actually reflect 
the reality of what the project decided or what the contributors want. 
Our official policy dates from back in the Java 1.4 days---it was 
intended to express a desire to follow the newer Java platform releases 
but with enough delay to ensure that institutional clients would have 
the platform installed or at the very least be able to install it. The 
official policy we adopted was intended to formalize that balance but, 
since it relies on a numbering system Sun no longer follows, our policy 
no longer works. Formally, we are not even allowed to use Java 5 but 
that change was made by consensus for the Geotools 2.5 release. Since 
then, there has been no consensus to move on; the real policy is much 
more focused now on the JEE platform than it was when our formal policy 
was adopted. Anyhow, the developer launching the cleanup effort decided 
that since it would take a while to finish, he would start the rewrite 
on Java 6 and leverage the improvements of that platform. The style work 
followed and is now stuck in limbo.

So the style work will not land for a while yet. The first phase of the 
cleanup work is almost ready but once that is announced there will be 
lots of work remaining before the styling code can land.


The JAXB question is more complex, raising the existence of the 
religious war that so tiringly rages in our midst. There are the two 
factions in Geotools, the eclipse sect and the sun fanatics. Neither can 
tolerate the technology of the other and both are too lazy to really 
learn the strengths and limitations of the technology of the other. SLD 
parsing has been done both ways, with JAXB and by hand parsing. Both 
work, and both could probably co-exist but religious wars tend not to 
produce a good environment for co-existence.


Finally, on the matching of geoapi and SLD, you may very well be right. 
I know as part of the work on the new styling, the Geoapi interfaces 
were modified. That may indeed have aligned them better with the new 
specification. Geoapi in general is improving over time as we understand 
it better, as the specifications evolve and improve, and as we need to 
maintain the project over time. As I understand things, the plan is 
indeed to move towards the GeoAPI approach.

cheers,
--adrian

Francisco Moura wrote:
> Srs.,
>
> I´m working on a project that is being built on top of Geotools.
>
> One of our demands is some advanced styles, and so, I´ve been studing 
> the current 2.5.1 version implmentation and reading the upcoming 
> modifications.
>
> I saw that you are creating a new MarkFactory and some other great 
> stuff, and I have some questions:
>
> - When should this new style version be released?
>
> - Is there any issue for not using a library like JAXB for doing the SLD 
> parser and tranformer?
>
> - Are you going to make geotools style more like the geoapi style? 
> Because geoapi style seems to be closer to the SLD specification.
>
> Thanks in advance,
>
> Francisco Moura
> Tecgraf - PUC-Rio
>
> ------------------------------------------------------------------------------
> SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
> The future of the web can't happen without you.  Join us at MIX09 to help
> pave the way to the Next Web now. Learn more and register at
> http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
> _______________________________________________
> Geotools-devel mailing list
> Geo...@li...
> https://lists.sourceforge.net/lists/listinfo/geotools-devel
>