[Biojava-dev] Webservices and choise of SOAP software

Patrick McConnell MCCon012@mc.duke.edu
Fri, 27 Sep 2002 09:09:05 -0400


I agree that an open source solution would be ideal.  However, Axis still
is in beta testing (last time I checked) and is pretty unreliable.

I think the best alternative is to build the webservices module independent
of a web services engine.  That is, the web service implementations should
be pluggable to virtually any engine.  This may prove to be too difficult
in practice, but in theory it is sound.

What is your thoughts on this?  Does anyone have any experience building
web services that will 'support' multiple service engines?

I have made some holistic changes to the webservices module that I should
update.  These include refactoring the packages and adding an
implementation of MSAML and ClustalW.

-Patrick





Kalle Näslund <kalle.naslund@genpat.uu.se>@biojava.org on 09/27/2002
11:04:30 AM

Sent by:    biojava-dev-admin@biojava.org


To:    biojava-dev@biojava.org
cc:

Subject:    [Biojava-dev] Webservices and choise of SOAP software


<HI!><

I have been looking a bit at the webservices module in biojava, and it
seems
to depend on GLUE ( www.themindelectric.com ).  From a quick look it seems
GLUE is available in two versions, one standard and one pro.
The standard version ( witch is what the biojava webservice code relies
on as i understand it )
Is free, as in it doesnt cost any money to use, but it comes under its
own licence,  and its
provided as binaries only.

As BioJava is an opensource project, wouldnt it be better if we relied on
some opensource alternative to GLUE ?

There are several reasons why sticking to opensource alternatives are
preferable such as

1) By going with binary only solutions you are loosing one of the big
wins with opensource,
    the possibility to immediately fix bugs, and the possibility to
actualy change the code so
    it suits your needs.

    If people want things GLUE standard doesnt support then there is no
other alternative
    but to hope that themindelectric implements it in GLUE pro, and then
pay for it.

2) Keeping track of several different licences, that limit what you can
and can not do makes
    it hard for people that want to use biojava in commercial settings
to figure out what they
    can and cant do.

3) Long time support and future, will themindelectric provide GLUE
standard under the
    same licence for a long time ? what happens if they dont ?

4) The old saying of sticking the money where the mouth is. If we claim
biojava is opensource
    and being opensource is good, then it doesnt look that good if we
rely on lots of closed source
    binary only things.

There are other altenatives to GLUE, personaly i would think AXIS (
http://xml.apache.org/axis )
might be a good choice. But, if there are other alternatives aswell.

Mind, this is just my thoughts, what do YOU think ?

mvh ( kind regards or something ) Kalle






_______________________________________________
biojava-dev mailing list
biojava-dev@biojava.org
 http://biojava.org/mailman/listinfo/biojava-dev