Re: The future of the W3C?

Aw crap, wrote a giant email again. Every year whether they want it or not…

On Thu, 22 Dec 2022 at 09:28, Chaals Nevile <chaals@fastmail.fm> wrote:

> To the extent that Nick Doty is a former W3C staff member, current AC rep
> and participant and candidate for the next Advisory Board, sure it's an
> outsider view. (To be fair, that describes me, too).
>

That W3Cish people who are not “on the Team” can be considered to be
outsiders even after decades of work is one of the smaller problems  here).

When I got involved - and met you - 1997  onwards, two things stick in my
memory as relevant to the ongoing “what even is W3C” agonising and
navel-gazing.

One is that there was back then a hint of a different conception of “W3C”,
in that I remember W3C issuing W3C branded business cards to at least some
chairs. In the years you and I were on the Team, the “Team or Not Team”
distinction was the most significant important, and the strongest barrier
for information flow. I’d say it was even a bigger deal than Member vs
Public, so this hint of an alternative path where W3C Team was a different
broader thing, always stayed with me. Being on the Team as a W3C Fellow
working for a Member was a weird experience- you have two roles and little
transparency between them.

The other thing was, more or less, this mailing list. Well technically
there have been 3 in a series; RDF-DEV, which we migrated into
www-rdf-interest@ in 1999 when I joined The Team, and which we somewhat
grudgingly renamed semantic-web@ when it became clear just how unpopular
RDF was in an XML-centric era.

RDFS was expected to go to REC in May 1999 but it got pulled at the last
minute because the XML Schema WG had heard this and objected strenuously
lest they be forced to define their language somehow using RDF. RDF didn’t
share that big Membership-funding backing and so we kind of got kicked into
the weeds.

At one Advisory Committee meeting in that limbo period when W3C RDF work
was stopped, RDF wasn’t even mentioned during the T&S update, so great was
W3C Team fear of the XML-loving Membership’s wrath. The thing that pulled
us out of limbo and from being abandoned by the Consortium was a
combination of two things: external funding and public participation. Both
of which continue to sit awkwardly within W3C Process even 20+ years later.

Firstly, the W3C staff’s RDF/SemWeb team (concentrated in Boston, i.e. MIT)
sought external funding from DARPA (hello Jim Hendler!) to continue to
progress the planned RDF work that the W3C Members (and associated Process)
were quite vocally  rejecting. For this not to be a disastrous conflict
with officially chartered W3C work it had to be reframed in more researchy,
futuristic terms - hence all the talk of logical reasoning mixed with
digitally signed data and formal proofs.

https://www.w3.org/2000/01/sw/

Proposal doc is
https://www.w3.org/2000/01/sw/DevelopmentProposal

This all rankled with me as a European, both for taking US military funds
and associated duties, and for doing it as an MIT activity rather than as a
W3C Activity, part of a more integrated program of work. Over the years W3C
has been involved in various externally funded efforts, but never quite
found a way for them to exist fully in the system. If they aren’t
Member-funded, why should members get any bigger say than other Web users,
since the Web is “for everyone”? Also W3C as an independent entity didn’t
exist as such, and so without W3C Inc., projects were proposed, staffed and
managed via ERCIM, MIT and the other hosts. With the associated
transparency and coordination costs you’d expect.

All these factors made W3C hard to help. There was a constant pressure
(especially after dot com crash) to keep the Members involved, and happy to
pay their fees. Which meant an emphasis on stuff the Members wanted, even
if it turned out to be bloated snakeoil (ahem, Web Services). And yet
structurally other forms of support were hard to grow: Members naturally
objected when Team members spent more time on eg writing EU project reports
than on replying to their emails about Process-visible work. And efforts to
bring a much wider range of people and organisations into W3C’s orbit
chafed against the ever present fear of alienating the fee paying
Membership.

It was in the middle of all this mess that the original RDF Interest Group
thrived. We were permitted to do a public WG I think in large part because
RDF was seen as weird and failing; doing RDF in public wasn’t leaking any
valuable “member benefits”, etc. We ran the group more like a messy
opensource project than a traditional W3C Interest Group, which was
historically a Member only thing (eg people flying around to meetings or
making international calls to W3C phone lines in Boston). It probably also
helped that RDF was being reframed by the Team as non-XML-threatening, more
a futuristic prototyping activity.

This mailing list (aka rdf-dev, aka www-rdf-interest) was the place where
the groundwork for the later RDF Core WG was done, especially by Brian
McBride (who later co-chaired but basically led the full Working Group).  I
am proud of what we did here, back then (and of your part in all that,
chaals). Brian and the other RDF hackers basically saved the RDF initiative
from the W3C Membership and Process.

I think we tried to do a few important things in pursuit of a healthier
structure for W3C as an organization. We showed strong benefits for public
participation in W3C, for W3C groups working in public view,  and for those
groups to have to respond to all feedback received, not just from paying
Members. Before RDF Core most WGs operated in Member-only secrecy. We
experimented with other (cheaper, maybe more inclusive) mechanisms for
collaboration than those that come naturally for a pay-to-play Membership
Consortium (IRC with bots/logs to support async use; early wikis, blogs,
rss). It also showed that there can be more to W3C involvement than “come
here and get famous for making yet more specs with the bigshots”. Specs are
a means to an end, not an end in themselves.

In SWAD-Europe -
https://www.w3.org/2001/sw/Europe/ - I also think we did a decent job of
creating an externally funded workplan that supported, complemented and
furthered the common endeavours of the wider W3C rather than slipping into
the two easier failure modes: (a) “we’ll show the membership! this research
project will demonstrate they were wrong about X”; or (b) giving the
mistaken impression that regional funders (US; EU, …) had sponsored a full
W3C WG and bought their way into fast tracking whatever it is they wanted.
When this happens, Members push back more on the idea of diverse funding
streams. Oh, and we also made the best Semantic Web tshirts (although the
b/g colour wasn’t the most inclusive choice for dark skin, we learned too
late).

But anyway…

To be clear, since W3C is relevant to the world, there is a certain
> interest in what is happening. But that should be balanced against the
> right of W3C employees to a certain amount of privacy regarding the
> organisational details of their work because it's part of their own life,
> and the added difficulty that is imposed by having to manage everday
> organisational work with a million onlookers providing commentary and
> suggestions.
>

Absolutely. And at this time in particular it is very stressful and
personal too.

I have kept rather away from discussions of what W3C might evolve into as
W3C Inc., to date, because it is too hard to separate from the careers and
lives of the folk on the Team. I wouldn’t be inflicting this long mail on
you all except for Robin’s recent mail suggesting a W3C-MIT rough consensus
has been met.

https://twitter.com/robinberjon/status/1605693369409187840

This does not completely depersonalize discussions of what W3C could or
should be as W3C Inc.  But it means some folk may have a job next week at
least.


Essentially, as Danbri and Ivan have noted, this is about organisational
> management - who pays the staff, writes their contracts, etc. As Ivan
> noted, the latest news is different each day, as negotiations over what
> people keep and give up in the transition go down to the wire.
>

It took 25 years of agonising to get to a situation where nothing was
agreed until the final 2.5 days. Did Brexit teach us nothing? :(


If you're a W3C member, this is relevant because W3C governance heavily
> involves the members (who for 25 years have been the people who pay to keep
> the whole thing going for the benefit not only of themselves but the rest
> of the world who use the Web).
>

From a perspective of participants in W3C standards development, there's
> little to see - in the end, things like the copyright statement and the
> postal address will change slightly, and if you were in the habit of
> calling W3C on a landline telephone or sending faxes, those numbers will
> likely stop working. For the rest, the Process that governs W3C's actual
> work will still be the Process (it gets updated most years, but overall
> it's recognisable as an evolution of the process first published in 1997).
> Staff turnover is a background constant. But URLs, mailing lists, email
> addresses and so on will follow the principles outlined a quarter-century
> ago in TimBL's "Cool URLs don't change" (an important lesson W3C learned
> from experience).
>
> I suggest that there is very little to see here for this mailing list.
>

It’s a pity this was your shortest paragraph as it kinda leaps out and
comes off as snarky and dismissive, unlike the others. I have tries to say
here what I said more concisely earlier in the thread: we should worry
about the human impact to W3C Team members first, the bigger picture can
evolve gradually.

But is there a concern here for mailing list members? I would say:
absolutely, yes. There ought to be something here for the people on this
mailing list to care about. Over and above the historical connections I
described above, we have a perpetual struggle in W3C circles to articulate
what is “core Web” / “core W3C”, vs what is peripheral and a distraction.
W3C has gone through its conflict with WHATWG, been accused of both
neglecting and over-focussing the web platform / browser work area. And yet
every time something is proposed as a new work item it is hard to see how a
principled decision could be made.

This matters to us in the Semantic Web and RDF world, because the
governance structures most appropriate for developing large complex pieces
of technology (canonically the web platform, but also RDF + RDFa + SPARQL +
JSON-LD + SHACL + RDFS + OWL etc.)… might not be also appropriate for
participation in the creation and curation of schemas/ontologies (or even
datasets that use them). We see this with schema.org and other groups
deploying RDF; one week we might have threads with Martin Hepp about
e-commerce schemas, and the next with Factcheckers and IPTC about media
files and disinformation, or bioinformatics experts about schemas for
proteins and species. It is very varied topically.  If every topic is
potentially in scope (and RDF was designed that way), is everyone in the
world, or their orgs, supposed to join W3C just in case?

 And so to an awkward circularity: what you/we/they focus upon for
standardization ought to be grounded in the goals and aspirations of an
inclusively assembled group of stakeholders. But you’ll pick different
stakeholder and inclusion (and funding) models depending on your topical
focus is. We know W3C isn’t likely to be good at working on covid vaccines
or 5G or undersea cables, but how does that help us collectively figure out
whether it is more important to prioritise maintenance of XML, XSLT, XPath
etc., versus say “Web technology for self driving cars” or some new browser
API pursuing a world in which the Web is the most usable and useful
interface on phones, tablets, TVs?

Semantic Web has long been on the periphery of what was seen as “core W3C
business”, even while beloved by TimBL. But I’d argue that its ideas about
information sharing are core to the notion of a “for everyone” World Wide
Web as a distributed hypertext information system, even if less essential
to the browser UI and rendering engines notion of Web. So… important, but
hopefully not so expensive to maintain and develop as a browser engine,
perhaps?

I personally don’t believe we’ll find a way beyond the neverending
bickering about priorities without realising that “web” is now being used
to mean two things: the machinery humans use to access something, and the
“stuff” that is accessed. The web platform and the World Wide Web itself.
It’s a term very like “TV”  in that regard, and I have no idea where to go
from that observation :)

Cheers,

Dan



cheers
>
> Chaals
>
>
> On Thursday, 22 December 2022 06:36:53 (+01:00),
> ProjectParadigm-ICT-Program wrote:
>
> Maybe some staff members could elaborate on the ongoing internal processes
> at W3C?
>
> The following link provides AN OUTSIDER view:
> https://cdt.org/insights/changing-governance-at-w3c/
>
>
> Milton Ponson
> GSM: +297 747 8280
> PO Box 1154, Oranjestad
> Aruba, Dutch Caribbean
> Project Paradigm: Bringing the ICT tools for sustainable development to
> all stakeholders worldwide through collaborative research on applied
> mathematics, advanced modeling, software and standards development
>
>
> On Wednesday, December 21, 2022 at 11:53:17 AM AST, Ivan Herman <
> ivan@w3.org> wrote:
>
>
> +1 to what Danbri said.
>
> Furthermore, things are evolving extremely fast, with meetings happening
> every day. By the time this mail goes out my SMTP server, it might already
> be outdated. So…
>
> Ivan
>
> On 19 Dec 2022, at 17:33, Dan Brickley <danbri@danbri.org> wrote:
>
>
>
> On Mon, 19 Dec 2022 at 15:40, Martynas Jusevičius <martynas@atomgraph.com>
> wrote:
>
> Hi,
>
> Not sure why this is not all over the @w3.org mailing lists, but it
> seems like the future of the W3C is at stake?
>
> "At this point it looks like we will not have an operational W3C
> nonprofit on Jan 1. Every Director will vote their conscience, but it
> seems likely that the asset transfer will be rejected, leaving MIT
> responsible for its contracts with W3C Members (for which they have
> paid)."
> https://twitter.com/robinberjon/status/1603835007323209729
>
> What consequences does this have for the existing and future
> specifications?
>
>
>
> With respect, I think this is a moment for considering the specific
> situation of the W3C Team at MIT - the staff whose employment is under
> threat from this mess. The specs will broadly be ok, and don’t need
> healthcare or a plan for paying rent/mortgage/heating in the coming weeks.
>
> Dan
>
>
>
>
> Martynas
> atomgraph.com
>
>
>
> ----
> Ivan Herman, W3C
> Home: http://www.w3.org/People/Ivan/
> mobile: +33 6 52 46 00 43
>
>
>
> --
> Chaals Nevile
> Using Fastmail - it's worth it
>