>> For users of Open Source software, the choice between Deegree and
>> Geoserver is a hard one. Imagine what could be done if the two projects
>> worked together.
>>
>
> The code bases and the development philosophy are so much different
> that it's difficult to think about a merger, really.
> There is also quite a lot of personal investment and design decisions
> involved, so I don't see the two projects converge anytime soon... sorry.
> Just my personal opinion
I would probably have to agree, that a merge is a ways off. I do think
that it is rather unfortunate that we can't combine them, since having two
java open source WFS/WMS products is not the best, it'd be better if we
were all working on the same code. We were going to try to merge them,
but there really are quite a few differences. Hopefully we will be able
to converge our feature models through the GeoAPI project
(http://geoapi.sourceforge.net), and after that collaboration will
hopefully be easier. I think we both would like to see it happen, but
neither project has the resources and time to completely refactor
everything and agree on the way forward. So for now we will work on small
steps like GeoAPI, and after that can try out making datasource adapters
between the projects, ect.
thanks for your answer. Could you please elaborate a bit on the
differences in code base and development philosophy? For an outsider these
things are hard to investigate/notice ofcourse.
I could try to elaborate the differences, but it would definitely be
biased, since I'm invested in GeoServer/GeoTools, and have only studied
the Deegree code enough to get a handle on it. So I think you should also
ask the Deegree developers, we like them and think they do great work, and
they can correct any misconceptions I might have.
As for the development philosophy, I think one of the most important
things for GeoServer is to make the server a community effort, to involve
other developers as much as possible, to make the project accessible and
extensible. Before 1.0 there was pretty much only one developer really
working on GeoServer, first Rob and then me, but Rob made the decision to
build GeoServer on top of GeoTools so as to have more people involved in
the core components of GeoServer. There was a much wider community in
GeoTools, and we have certainly benefitted by being associated with them,
as Oracle, ArcSDE and Shapefile were all contributed by GeoTools members.
We make an effort to give as much back to GeoTools as possible, rolling
code that may be useful to them back into their code base. The nice thing
about that process is it gets people whose interests are not server
centric thinking about it, leading to a stronger overall feature model,
whic is easily our biggest benefit.
Only recently has GeoServer started getting developers contributing
directly to its codebase, and I'm still working to make it easier for new
developers to get involved. I'm planning on redoing the build pages, and
writing up some better design documents. Jody Garnett has been doing a
great job of bringing about this change, writing solid javadocs and design
documents - I'd like to extend that to the whole GeoServer codebase,
making it very easy for developers to get involved. Until I do that (and
even after I do that) I try to make myself as available as possible to
answer questions. But I would like to move myself out of the main role,
and hopefully have a team of developers that all has solid ownership of
the project. GeoServer/GeoTools strives to be as democratic as possible,
to make the effort truly a community-wide effort, making it easy for
contributors to have a say in the future of the projects.
I think that deegree is a bit more centralized, and can be, since most
all the developers work in the same place. Deegree definitely moves
faster than us, as they're implementing wfs/wms/catalog/wcs and more
already. I think they are a bit less user focused, they don't work so
much on configuration and accessibility, whereas one of the highest
GeoServer priorities of late is an easy to use web configuration tool
(Thanks Jody and Refractions!).
I can't comment so much on the codebase itself, as I don't remember
deegree's well enough except for the fact that they are not easy to
integrate, the two project just took a different approach on how to deal
with things.
Hope this helps,
Chris
(perhaps I should add this to the faq?)
Best regards,
Bart
> Bart van den Eijnden wrote:
>> Hi list,
>>
>> Geoserver has now incorporated a WMS, and has plans for a catalog server
>> if I understand correctly. These two are already part of the Deegree
>> project.
>>
>> Shouldn't there be more consistency/cooperation between the two
>> projects, as the architectures of the two projects resemble each other
>> so much?
>
> The external architecture is similar. Look at the internals and you'll
> see how much they are different (quite a lot)
>
>
> Best regards
> Andrea Aime
>
>
>
-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html
_______________________________________________
Geoserver-devel mailing list
Geoserver-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-devel
--