Re: [DotGNU]Re: [arch-users] In defense of (the idea of) having an XML-RPC api

[Seth Johnson]

Norbert Bollow wrote:
> 
> > --- Karel Gardas <address@hidden> wrote:
> > > Whole WebServices idea is overhyped, reinvent a wheel, idea commint
> > > from M$ just to `invent' yet another middleware to attack their
> > > competitors.
> 
> I don't like MS any more than you do, and much of what they
> "invented" is supported in DotGNU only for the sake of portability.
> 
> However, I'd like to defend XMLRPC from these attacks.  Whether or
> not it was invented by MS is irrelevant (it wasn't), what is relevant
> is that the concept of having a simple standard protocol for remote
> procedure calls is very useful for many kinds of system integration
> tasks.  For example I have just been involved in a little consulting
> project in which we integrated some custom website stuff (written in
> PHP) with Request Tracker which is written in Perl.  In order to
> integrate them we added XMLRPC interfaces to both, when are then used
> from the other program.  In this case we're both the PHP system and
> the Perl-based system are running on the same computer and XMLRPC is
> used to let them communicate with each other via the loopback interface.
> 
> While XMLRPC isn't adequate for everything, for many "ask a remote
> computer to do something" purposes it gives everything that you want.
> Adding XMLRPC interfaces to Free Software programs is a good idea IMO
> whenever remote users or remote programs might want to do something
> with them.  In my eyes, when making the choice between using cvs or
> arch or one of the other alternatives (a choice that I'll have to make
> again soon for a new project) whether or not the system has a good
> XMLRPC interface is an important consideration.


Thought I'd add a comment, supportive of nearly everyone's position (because
that's possible here!).

Arch is a community working at a different level from the XML/web services
world.  They're naturally inclined to see the XML stuff as overhyped
syntactic sugar.  All XML is, is an agreement about how to talk about data
structure.  A lot of the hype is really just "the stuff that's possible now
that we agree to talk this language."

Just FYI, to put things in perspective with respect to Microsoft's latest
middleware strategy:  DotGNU is doing an extremely important thing -- as
quickly as possible (and their work has been amazing in this respect)
creating an alternative to .NET.  The key reason for this is the fact that
.NET/web services/application servers are very dangerous to the idea of free
code -- they create an infrastructure for making web services that are
proprietary, in a form in relation to which copyleft appears to become
irrelevant.  DotGNU isn't so concerned with making something supposedly "as
cool as" .NET, but with making sure that we're in that arena with something
built from the standpoint of providing decentralized and freedom-loving
versions of the web services stuff.  It's a best defense against .NET, that
might just save all of the rest of the free software community.

XMLRPC is a reflection of the XML/web services world.  It's not a bad thing,
in its sphere.  I think appearing to encourage XML as an integral part of
arch, might be a bit of a hard sell, though.  I think the frame of mind of
arch is to build greatly powerful algorithms that may persuade in the
respect of their newness and power.  There's not going to be a lot of
interest in possibly foreshortening the prospect of developing something
incisive, by giving too much shrift to some of the presuppositions and
motivations of the web services world.  I think that's likely a source of
the reaction to James' posting of his XMLRPC interface.  But providing that
interface isn't really a problem in itself.

Seth Johnson

-- 

DRM is Theft!  We are the Stakeholders!

New Yorkers for Fair Use
http://www.nyfairuse.org

[CC] Counter-copyright: http://www.boson2x.org/article.php3?id_article=21

I reserve no rights restricting copying, modification or distribution of
this incidentally recorded communication.  Original authorship should be
attributed reasonably, but only so far as such an expectation might hold for
usual practice in ordinary social discourse to which one holds no claim of
exclusive rights.