Re: XML protocol comparison

On Sat, 20 May 2000, Dave Winer wrote:

> Laird,
> 
> Here's how I feel about this now -- let's make sure that this doesn't happen
> in the future. I've spoken my piece, you have too, now let's talk about
> technology and publishing, what's done is done.

If you don't want "this" to happen in the future, I would suggest that you
direct comments about the ICE protocol to the ICE Authoring Group rather
than posting broadsides on your web site and public lists. Admittedly,
several members of the ICE AG happened to be on the list, so we've been
able to engage in this educational dialogue, but more direct communication
is probably more productive.

> If there's a specification in progress and there are obvious people who are
> not represented, let's assume that there's a routing failure, as happened in
> this case, and do the proactive thing, make sure that everyone who has
> something to contribute has a chance to make a contribution.

You mean, the way you and Netscape contacted the ICE AG regarding the RSS
effort?

I certainly agree with the idea that standards effors should be widely
communicated and open to all interested parties. The ICE AG has made
contact information as public as it can (embedded in the spec, press
releases, etc., for two years), and had a large number of companies
involved in the process, including major content and technology companies.
I am sorry that our marketing efforts were apparently insufficient to get
you involved in the effort, and that in the two years since the initial
public announcements, and hear and a half since the submission of 1.0 to
the W3C, that we've been unable to inspire you to contact us.

To be honest, we targeted companies that are in business of producing,
syndicating, or consuming content, not (at the time) scripting language
vendors. I would have thought that when you repositioned your company as a
"content management and syndication" company you would have contacted
existing relevant standards bodies.

That being said, the current Authoring Group (in alphabetical
order) consists of:

Sasha Aickin, Plumtree Software, Michael Branch, Vignette Corporation, Jay
Brodsky, Tribune Media Services - Implementation Sub-Committee Chairman,
Phil Gibson, National Semiconductor, Jack Gudenkauf, Microsoft, Martin
Hardee, Sun Microsystems - Standards Outreach Sub-Committee Chairman,
Bruce Hunt, Adobe Systems - Vice Chairman, Brad Husick, Vignette
Corporation - Marketing Sub-Committee Chairman (non-voting), Sami Khoury,
What U Want, Inc., Daniel Koger, Herrick Douglass, Erik Leckner, Seagate
Technology, Inc., Dr. Richard Martin, Active Data Exchange, Inc., Nathan
Pride, Wavo Corporation, Laird Popkin, Sotheby's Holdings - Chairman,
Mondhipa Ratnarathorn, Fresher Information Corp., and Adam Souzis, Kinecta
Corp. (formerly ShiftKey Inc.).

And there are two dozen or so in the ICE Network (which includes
participants who aren't in the Authoring Group), such as Documentum,
ArcadiaOne, Intershop, and Intervu.

I trust that you are familiar enough with the market to pick out which
ones are direct competitors of Vignette's.

Detailed information on the standard and membership in the ICE Network is
at http://www.icestandard.org.

> I wholeheartedly support the approach of including content creators and
> encouraging them to drive the process. The technology developers exist to
> support and enhance their work. But, I would encourage you to broaden your
> reach beyond the corporate content creators and include inviduals who
> publish because they love their subject and are not playing footsy with the
> companies they work for and cover. There's a new journalism developing
> around the economics of the web. I think that RSS better reflects that
> approach, and people who write for the web have a place at this table, as
> well as people who write for big companies.

I am confused about how you see RSS as being better suited to individual
content creators than ICE; from what I can see, the reverse appears to be
true. RSS requires content syndicators to be in the web site business
(operate a web site with a persistent internet connection with enough
bandwidth to transmit content to all subscribers at any time, sell
advertising, etc.). ICE syndicators can create content offline and need
only be online to transmit the content (e.g. to a publisher) and can use
scheduling to manage bandwidth consumption.

This doesn't make RSS a bad protocol, just one that targets a more limited
application space than ICE.

> It's probably just as well that we didn't participate in ICE, because we
> would have told you to scrap the whole thing and pass around links instead
> of stories, and reverse the flow of dollars, pay for delivery of readers
> instead of paying for the delivery of content. That seems to me how the web
> dictates this must work.

As I said before, the "simple" model works fine for web sites who want to
share headlines to drive traffic, and it's a case that ICE supports. I
think that your input could be valuable in making sure that we have kept
that "simple case" as simple as possible, and in communicating that fact
to people who would otherwise misunderstand the ICE standard.

I think that Jonathan Eisenzopf's article on RSS at
http://www.webtechniques.com/archives/2000/02/eisenzopf/ portrays the
relationship between the protocols appropriately: "It's important to note
that while RSS is capable of syndicating content headlines, there are
other XML formats like XMLNews and ICE that are better suited for handling
larger syndication systems." It's an insightful comment, particulary given
that Reuters recently launched their online content syndication service,
based on ICE and an XML news format. The announcement is at
http://www.reuters.com/aboutreuters/newsreleases/2000/XMLnews.html.

Given that ICE does everything that RSS does, and was published (and
implemented) well before the RSS effort started, I would have preferred
coordinating from the start. Let's make up for lost time; I'd like to open
exploration of whether there's a means of unifying the two standards, so
that rather than picking between RSS and ICE, there could be a gradiation
of capabilities of a unified syndication standard that addresses the full
range of content syndication scenarios.

Let's discuss how to proceed. You have 
my contact information. I suspect that the rest of the people on this list
would appreciate it if we continued this discuss offline.

> Dave
> 
> 
> 
> ----- Original Message -----
> From: "Laird A Popkin" <laird@io.com>
> To: <xml-dist-app@w3.org>
> Cc: <dave@userland.com>; <eric@w3.org>; <bernhard.dorninger@scch.at>;
> <ice-ag@egroups.com>
> Sent: Saturday, May 20, 2000 12:19 AM
> Subject: Re: XML protocol comparison
> 
> 
> > Re: XML protocol comparison
> >
> >    RE: XML protocol comparison
> >
> > > Perhaps the perception derives from the fact
> > > that you excluded your competitors from the design process of ICE.
> > >   We tried to get involved, and had the door slammed in our face.
> > Arrogance begets isolation, outages are always routed around.
> >
> > I am the chair of the ICE Authoring Group, and have been since its
> > inception. As such, I cannot allow Dave's continued misrepresentation to
> > go without comment.
> >
> > In terms of Dave's company having "had the door slammed in our face," my
> > contact information has been on every release of the ICE specification,
> > and I have never been contacted by anyone from UserLand about
> > participation in the ICE effort. I certainly would welcome your
> > participation in the ICE Network. My email address is laird@io.com (or
> > laird.popkin@sothebys.com), and my office phone number is 212/774-5338.
> >
> > I don't know who you spoke with at Vignette, but since they don't run the
> > ICE effort, you may have made a "protocol error" -- if you asked a random
> > Microsoft or Adobe employee, for example, they couldn't have been too
> > helpful either. Other the other hand, the people listed as contacts on the
> > standard would certainly have welcomed your contact.
> >
> > The repeated presentation of ICE as being run by Vignette is incorrect.
> > Vignette initiated the ICE effort, but from the beginning the ICE AG
> > members have had equal votes, and funded ICE AG activities equally.
> > Vignette does not control either the AG or its composition; applications
> > for membership are administered by The GCA's IDEAlliance, an independent
> > organization, and are voted on by the ICE AG members, each of whom have a
> > single vote.
> >
> > A critical component of ICE's success thus far, I believe, is that a
> > majority of the AG is and must be content companies, not vendors,
> > specifically to ensure that the ICE standard is driven by market needs
> > rather than vendor goals. Thus, while there are a number of vendors active
> > in the AG, even if they all voted together they could not control the AG.
> >
> > >   Maybe Vignette will lose some of the arrogance and work openly on
> > syndication technologies.
> >
> > Vignette doesn't run the ICE effort, and the ICE AG has always been open
> > to input and participation by any interested parties.  For what it's
> > worth, the Vignette employees participating in the ICE effort have never
> > been arrogant; they were quite open to fundamental changes in their
> > initial proposals based on arguments by other AG members.
> >
> > The ICE effort has been open since the beginning; I will admit that our
> > efforts have been more focused on creating the standard and in working
> > with participating vendors and content companies to get ICE implemented
> > and deployed than in marketing the protocol to standards groups; we felt
> > that was premature until we'd had sufficient experience in production to
> > have more credibility.
> >
> > With ICE 1.0 being in production for over a year, we've produced ICE 1.1
> > and are stepping up our "outreach" efforts this year.
> >
> > To get back to the topic of this mailing list, I can say that I would
> > welcome the development of a standard approach to building protocols in
> > XML. A good percentage of the ICE specification addresses fairly generic
> > protocol issues (e.g. packaging, request/response, header/body, logging,
> > etc.) that were required in order to achieve ICE's goals, but which could
> > be addressed in a common manner across many protocols, allowing the ICE AG
> > to focus on the issues around content exchange by layering over a generic
> > messaging standard, in the same way that we layer over generic transport
> > standards (HTTP, SSL, sockets, mail, etc.), XML, and so on. I would hope
> > that we could participate in the development of such a standard in order
> > to (1) contribute the real-world production experience that ICE has
> > accumulated, and (2) ensure that the generic XML messaging standard would
> > support ICE's needs.
> >
> >    Dave
> >
> > - Laird Popkin, CTO, Sotheby's Holdings and Chair, ICE Authoring Group.
> >   laird.popkin@sot

Received on Saturday, 20 May 2000 08:55:35 UTC