Martin Desruisseaux wrote:
About the paragraph I added at the end of this page:
http://docs.codehaus.org/display/GEOTOOLS/Supporting+Hacks+and+Versions
Thanks, I was going to comment and then saw this message.
Jody Garnett a écrit :
Um I ment an usage example, that article is trying to show to
geotools programmers what the options look like to them.
I tried to add some more details, but I can hardly do much more.
Developers really need to learn annotations if they want to understand
this proposal. It seems to me that trying to explain more is like
trying to explain what annotations are, which is out of my scope.
Right, I was not looking for an explanation of what annotations are,
just an example of what the a developer would need to know to use the
result. It sounds like you are considering a couple ways of taking
advantage of annotations. I personally favour generating out the two
"standard" interfaces, from the one "generic" source. Actually this
would be great at many levels. My main measure of success here is
wanting strongly typed objects that represent the limitations of each
specification (this is an area where geotools falls short), and still
allow room for experimentation (an area where geoapi understandably
falls short).
Except for the inconvenient cited below, I strongly suggest to use
annotations because it keep all doors open: javadoc, interfaces or
(maybe) compile-time warnings (the later need investigation in order
to determine if it is really possible). We can do everything cited in
the "Supporting Hacks and Versions" page automatically with
annotations, at our choice.
Right, and understood, but what I really need is that choice made: to
wit I am going to hack up style today in geotools (I don't feel bad
about it because the current set up is half interfaces, half classes).
And we can look at a strongly typed result.
The inconvenient is that annotations require J2SE 1.5. We can probably
hack the build in order to generate J2SE 1.4 compatible code. But
given the trouble that it would involve, this bring back again the
question: when will we be allowed to target J2SE 1.5? Note that the
public review ballot for J2EE 5.0 (the last step before J2EE adoption
as an official specification) closed on August 2, 2005. I don't know
yet when Sun implementation will be available.
It is a tough call, I would wait until after GeoTools 2.2 (I am shooting
for a 6 month release cycle, although cholmes would like to a point
release early and often with every RnD merge). I would really like to
engage the wider community in a discussion, but plans appear murky for
everyone these days.
I will also not that there is a Java version burn backlash - a subset
common to 1.3, 1.4, 1.5 in-order to get their work done. If Java keeps
chasing C# we may see more of this - Luddites arn't we all.
Jody
Martin.
-------------------------------------------------------
This SF.Net email is sponsored by the JBoss Inc. Get Certified Today
Register for a JBoss Training Course. Free Certification Exam
for All Training Attendees Through End of 2005. For more info visit:
http://ads.osdn.com/?ad_idv28&alloc_id845&op=click
_______________________________________________
Geotools-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-devel
