Joshua,

thanks for giving me the opportunity to set some things straight.

First, let me introduce myself: my name is Hannes Wallnöfer, I'm an open
source Java developer based in Vienna, Austria, I'm the author of XML-RPC
for Java, and I just decided to see if I can come up with a decent SOAP
implementation based on what I already have and the BDG (and the SOAP spec,
of course).

You express the idea that Dave's agenda may be to influence developers into
thinking that the BDG is the real SOAP. From my point of view this is
somewhere between funny and insulting. I'm not a "follower" of Dave, and I
don't think anybody else who's signed up to BDG is. I actually used to be
very critical of Dave's central role in XML-RPC, although the early spec
freeze obviously contributed to its success. Also, I didn't follow any of
Dave's work between XML-RPC and now in depth, simply because I'm, well, a
very busy developer.

So what is my reason to find the BDG and SOAP interesting? Certainly not the
idea of being able to talk to Userland software, because we already do that
very well with XML-RPC! Of course, the goal is to be compatible with as many
existing SOAP implementations as possible. So, while I will do whatever
necessary to talk to the other implementations on
http://www.xmethods.net/ilab/ when I'm done, I do much prefer if people
point out what's wrong with BDG *now* to make that unnecessary.

Finally, a word towards "incomplete" implementations. I think it is
completely ok to offer a subset of functionality for an "open end" spec like
SOAP, and I think it is also ok for developers to group around such subsets
- as long as they are open about it. The simple RPC mechanism offered by
XML-RPC has been a big success (we have something like 100 downloads per
day, and people are actually using it) with its very limited feature set. I
think that offering something similar for SOAP be an attractive offer to
many developers who don't want to jump into the depths of the SOAP spec. And
we'll have a big link to Apache SOAP for those who want something more
flexible and complete.

I hope I answered some of your concerns,
Hannes

PS: And I *don't* think WSDL is a bad idea!


on 31.03.2001 8:30 Uhr, allenjs@... at allenjs@... wrote:

> Sorry it has taken me awhile to check my facts and make sure that I
> wasn't misinterpreting things. And I think I was, to a large
> degree. Speaking only for myself, but it seems obvious that any
> subset of SOAP that adheres to the SOAP spec will be supported by any
> conforming SOAP "complete" implementation (and by extension,
> Microsoft), and if that subset gets more people on-board quicker, I
> think it is awesome and I am grateful to whoever is driving it.
>
> HOWEVER...
>
> I wouldn't minimize Dan's concerns because he is not stupid and has
> reasons for liking a more complete XSD. As another example, I know
> that I personally would be agitating here if I thought people were
> going to converge on a subset that did not include utf-8 and utf-16,
> because I have been upset by the bias toward western-european
> encodings in various "standards-compliant" products in the past
> (including Microsoft's). Clearly the full spec is the best record we
> have of what all of the various constituencies have been able to
> mutually agree is an acceptable and desirable interop. As someone
> pointed out, the subset which you have proposed tends to favor
> UserLand products, but is undoubtedly ignoring things that other
> people found very important (or they wouldn't have made it into the
> spec).
>
> Or to put it another way, if Israel and the Palestinians negotiate a
> common agrement, and then Israel decides to implement the "subset" of
> the agreement that involves keeping control of Jerusalem but leave
> the rest of the agreement unimplemented, do you think the
> Palestinians would be satisfied that "at least part of the agreement
> got implemented"? Would it be OK for the Israelis to say, "we know
> we agreed, but the part about withdrawing from west bank was really
> hard to implement, so we did the best we could?" And "treaty
> organization" IS the best way to think of this process. A bunch of
> people sit down together and agree to work together because their
> customers are demanding it. In aggregate, the people who penned the
> SOAP spec (and the people who agreed afterward to support, such as
> Sun) represent a pretty large base of customers. Think of customers
> as voters/citizens, who by hiring developers or purchasing products
> elect their "delegates" to represent them in these negotiations.
>
> My point is that a push for a subset without making it absolutely
> clear that full conformance to the spec is the eventual goal could be
> seen as an attempt to end-run the whole process. The fact that you
> have publicly said that you doubt the full spec is implementable (for
> you) and the way that you often paint this divisively to be some sort
> of BigCo Vs. LittleCo thing would make me just a tiny bit suspicious
> that you *never* intend to support the full spec, and you are playing
> politics to get the subset that best suits you. If you *aren't*
> planning on ever supporting the full spec, then I cannot personally
> support the manner in which you are attempting to redefine the
> agreement (certainly if you got Tel Aviv and Haifa together on a
> yahoo group, they could find an "agreeable" subset of the peace
> agreement that allowed keeping of west bank). So to put it most
> clearly, if your motive is to redefine the agreement (by supporting
> only part of it and politically agitating to get others to do the
> same), then I would ask you to think about doing so in a different
> manner. Maybe the political tone of your rhetoric has got me
> muddled, but I don't know what your motives *are*.
>
> I would like to see you very publicly make it clear that your motives
> are to simply take baby steps toward full SOAP compliance and you
> treat this as an ongoing process. People like Don may or may not be
> annoyed with which directions you lead everyone first, second, and
> whatever -- but I personally am really impressed with the progress
> that's been made so far and *somebody* needed to get everyone talking.
>
> If your motive is *not* to support SOAP 1.1, but rather lead as many
> people as you can attract to a subset (heavily influenced by you) and
> then stop; then I would ask that you be very honest and open about
> this with the people who are trusting you to guide them toward SOAP
> conformance, and perhaps stop calling what you are doing SOAP. After
> all, it would then become "The pieces of the SOAP agreement that Dave
> and others around him found to be best for them version 1.0". And I
> am not saying whether that resultant standard would be good or bad (I
> am sure it would be quite cool, and would have the benefit of being
> upward-compatible with SOAP), but it would not be SOAP, and it would
> be a snub at the constituencies who fought hard to get their
> interests represented in SOAP. Furthermore, it would be difficult to
> argue that this resultant spec was somehow more representative of the
> needs of a broad sweep of customers than the original SOAP spec. And
> it would be difficult to argue that the process by which you arrived
> at the spec (telling implementors that you are helping them implement
> the SOAP 1.1 spec, but leading them to just the pieces of it that you
> liked) was a better and more balanced process than the W3C process.
>
> -Joshua (speaking only for myself)
>
>
>
> To unsubscribe from this group, send an email to:
> soapbuilders-unsubscribe@yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>