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

[James Michael DuPont]

--- Norbert Bollow <address@hidden> wrote:
> And that "system
> integration" arena is what mdupont has been referring to in his posts
> on the arch lists (I believe he wants "arch" with an XMLRPC interface
> for a project of his which is not part of GNU or DotGNU).

Thats right Norbert, thanks for clarifiying it. 

As you might have noticed on the bottom of my signature, I am working
with a focus on  the introspector project. 

It needs to have an interface into an Revision Control System, to be
able to extract metadata out of it, to be able to intercept calls into
it, and to be able remotly access the backend and agents of a revision
control system programatically.

The only reason I am looking into Arch at all is because RMS suggested
that you might need help in fighing bitkeeper. I really think arch is a
good system , and would be usefull for dotgnu and many free software
projects as it is. 

I am an *user*, *supporter*, *tester* and *advocate* or dotgnu and
freesoftware, but at _NO_TIME_ am I speaking for them. Nor am I
speaking for arch. 

Please If you feel that any of my statements could have been viewed as
speaking for anyone else at all, point them out to me and I will
publically retract them.

Again, I hope that some of the misunderstandings about dotgnu have been
cleared up. Really, I dont want to bring dotgnu into the forground on
the arch discussion, but the DGEE represents an great environment for
working on system intgeration components, via the dgee you have
multiple backend engines, that could even include an hackerlib based
engine or the arch server itself, and via the pnet engine,you have the
great way to pinvoke existing c/c++ code. Via pnet/c you could even
compile arch directly for pnet. 

My focus was on discussion of an simple prototype of an remote api and
the semantics of arch. After repeated (and in my eyes irrelavant)
criticizm, it got dragged down into a discussion of protocols and
implementation. 

My personal conclusion is that arch team is maybe not prepared to
cooporate on the creation of an API that exposes the services of arch
in a machine processeable manner, that allows many people to implement
arch front end and arch back ends based on a well defined
specification.

I offer to spend 200$ of my personal money for arch consultancy for a
formal machine readable description (IDL,WSDL) of the arch concepts and
functions in arch, providing that is describes the functions of the
front and the back end in a useable manner. It should also reference
the existing functions and command names. That would be need to
continue on this path.

My next step is to work with cvs, which is well documented and for
which there many wrapper libs to continue with this effort. Arch will
have to wait untill the arch team makes some effort to document thier
interfaces and decide what they want to do.

Peace!
mike

=====
James Michael DuPont
http://introspector.sourceforge.net/

__________________________________
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.yahoo.com