W3C home > Mailing lists > Public > public-webpayments@w3.org > April 2016

Re: Update on Web Payments Working Group [The Web Browser API Incubation Anti-Pattern]

From: Roger Bass <roger@traxiant.com>
Date: Tue, 5 Apr 2016 18:29:28 -0700
Message-ID: <CA+nC-XvKAjVfMbAbn7SktK+CkaDCRMuUi6j7TirvsC8V+kxNww@mail.gmail.com>
To: Melvin Carvalho <melvincarvalho@gmail.com>
Cc: Anders Rundgren <anders.rundgren.net@gmail.com>, Gregg Kellogg <gregg@greggkellogg.net>, Web Payments CG <public-webpayments@w3.org>
Ok Melvin, my apologies. I completely misconstrued your response. Despite
answering almost the opposite of your real point, I think we do mostly
agree. The trick, I believe, is to have an extensible technical foundation
that can support the emergence of a big, global ecosystem - while focusing
initial implementations on use cases that can gain early traction in
niches, before the larger-scale network effects kick in. Mapping existing
standards or frameworks that already have scale can also play a key role.

There's been some more specific discussion of those use cases (B2B
Payments) on the Interledger CG list (where I know you also participate).
>From the perspective of this list, it seems to me there may be two
particular areas of interest as regards standards efforts. Both potentially
have touch points for the narrower, early use cases. One is on the identity
/ credentials side - I'll save further comments on that for later. The
other, possibly, relates to converging or mapping existing B2B XML efforts
into the Semantic Web framework. (There may be a role for Linked Data there
somewhere too, though it would require a closer look - from me, or more
likely someone more technically capable).

Googling around, I found that a particular academic has done some work on
precisely this issue. I've copied below an abstract of his paper. I
particularly appreciated his observation that "However, this is a difficult
jump as there is already a huge XML-based B2B framework and ontology-based
approaches lack momentum." Although he talks about mapping various elements
of the B2B XML stack onto OWL/RDF, my hunch is that the CPPA piece (just
now being worked on at OASIS, in fact) may be a key one, as regards
expressing B2B Metadata as service endpoints.

Here's a Slideshare link
<http://www.slideshare.net/rogargon/facilitating-busines-interoperability-from-the-semantic-web>
(or as PPT download here
<http://rhizomik.net/html/~roberto/papers/rgrg-bis07.ppt>). Full paper
accessible here
<http://www.academia.edu/355600/Facilitating_Business_Interoperability_From_the_Semantic_Web>,
with abstract below.

Facilitating Business Interoperability from the Semantic Web
Roberto García, Rosa Gil Universitat de Lleida Jaume II 69, E-25001 Lleida,
Spain rgarcia@diei.udl.es rgil@diei.udl.es

Abstract. Most approaches to B2B interoperability are based on language
syntax standardisation, usually by XML Schemas. However, due to XML
expressivity limitations, they are difficult to put into practice because
language semantics are not available for computerised means. Therefore,
there are many attempts to use formal semantics for B2B based on
ontologies. However, this is a difficult jump as there is already a huge
XML-based B2B framework and ontology-based approaches lack momentum. Our
approach to solve this impasse is based on a direct and transparent
transfer of existing XML Schemas and XML data to the semantic world. This
process is based on a XML Schema to web ontology mapping combined with an
XML data to semantic web data one. Once in the semantic space, it is easier
to integrate different business standards using ontology alignment tools
and to develop business information systems thanks to semantics-aware
tools.





On Tue, Apr 5, 2016 at 3:42 PM, Melvin Carvalho <melvincarvalho@gmail.com>
wrote:

>
>
> On 6 April 2016 at 00:19, Roger Bass <roger@traxiant.com> wrote:
>
>> Melvin, the sarcasm drips from your reply.
>>
>
> Sorry, it wasnt intended to be sarcastic at all!  :)
>
>
>>
>> We certainly agree that an extremely narrow focus is necessary to get any
>> new network effect kick started. In my last email, I perhaps described this
>> in overly broad terms. But there actually is a much narrower segment of
>> this very large payments market where I believe the use cases are clearer,
>> the problems are more tractable, and business interests aligned to make
>> something happen. In driving forward pilots of tactical solutions in those
>> areas, however, it becomes important to be able to describe how such
>> narrow, tactical solutions can become steps towards a larger,
>> standards-based industry (and indeed Web) platform.
>>
>
> So all these years I was an advocate of narrow focus.  But it just hasnt
> worked for us.  Now after looking back what I think is maybe we need a
> holistic approach.  I mean one team doesnt have to build it all, but maybe
> we need strategic presences in all the key areas that have been successful
> on the web so far.
>
>
>>
>> ISO 20022 is the standards effort that has most traction at the data /
>> message content layer - and the effort I'm working on potentially feeds in
>> to that. There are use cases that relate to legacy payment methods and
>> networks - but potentially also a use case that applies in the Interledger
>> context.
>>
>> In bootstrapping a connection between systems, however, starting from an
>> email regarding payment, say, describing service endpoints is an issue that
>> arises. Very pragmatic, tactical approaches would likely make sense early
>> on. I am, however, pointing out that there's potentially a path there to
>> generalizing the formal language for describing those endpoints, and
>> configuring a channel. The OASIS work I referred to is much more closely
>> focused on this type of problem. At this point, it certainly does seem as
>> if the gap is not small between those standards, and theoretically related
>> W3C ones (Linked Data, Semantic Web). I'm just pointing out that
>> conceptually at least, that gap could be closed. And if efforts focused on
>> some such narrow use cases were happening in a W3C CG context, it might
>> even make sense to leverage those Web Standards to do that.
>>
>
> So I do like the 'do one thing and do it well' philosophy (popularized by
> unix).  But more and more I think we need to put it all together into a
> complete user experience.  Why?  Because payment workflows become
> transformed with network economics.  In a sense payments can be a glue
> between different human activities (goods and services) -- but without a
> good range of those goods and services, we are perhaps not bringing digital
> economies to their fully potential.
>
> I have no problem at all with good projects with a narrow scope done
> well.  But I think the competitive advantage might just emerge when we
> start thinking bigger and creating a big economic network of apps.  These
> is the lines Im hopefully going to try and work towards with a focus on
> integration with others doing the same ...
>
>
>>
>> Roger
>>
>>
>>
>>
>>
>>
>> On Tue, Apr 5, 2016 at 2:53 PM, Melvin Carvalho <melvincarvalho@gmail.com
>> > wrote:
>>
>>>
>>>
>>> On 5 April 2016 at 23:25, Roger Bass <roger@traxiant.com> wrote:
>>>
>>>> On 2016-04-05 21:52, Gregg Kellogg wrote:
>>>> "In particular, if the linked data stack is involved, it's clearly a
>>>> web standard, but of course, that is not yet clear."
>>>>
>>>> Any payments scenario focused on interoperability of existing systems
>>>> needs to take one of two approaches, it seems to me. Either scope is
>>>> limited to a (very narrow) intersection of those systems' capabilities. Or
>>>> alternatively, scope could include a fuller, formal description of those
>>>> capabilities and service endpoints, along with some ability to negotiate a
>>>> "handshake" between them (i.e. finding the relative intersection on a
>>>> pairwise basis). Given the vast scope and variability inherent in $700
>>>> trillion a year of business payments, the latter approach seems more
>>>> extensible and scalable.
>>>>
>>>
>>> Yes I've come to the conclusion you need something really massive in
>>> scale to get a bone fide web economy kick started.  Something like a
>>> complete replacement of the whole of web 2.0 (from social, to multi media,
>>> to search, to comms, to browsing, to commerce) with linked data plus crypto
>>> currencies, and integrated payment workflows.  Instead of thinking small
>>> why not try and capture a chunk of a multi trillion dollar system and then
>>> allow an economic paradigm shift from bricks and mortar to digital.  Much
>>> like how in 2000 they talked about 'old economy' vs 'new economy' -- well
>>> that didnt happen, but with the tools we have today, particularly web
>>> standards, I think the web can become really exciting again.
>>>
>>>
>>>>
>>>> Linked Data and Semantic Web standards seem like a potential fit for
>>>> this type of requirement. That said, there is also work going on from a
>>>> more nuts-and-bolts, B2B implementation perspective in OASIS (building on
>>>> the ebXML history - the CPPA work in particular, in case anyone is
>>>> interested). I wonder (and in fact, have been asking the authors of those
>>>> specs) if any thought has been given on that side to converging with this
>>>> W3C standards stack. If anyone here has comments or perspectives, I'd love
>>>> to hear them.
>>>>
>>>> If I hear back any interesting answers to that question, I'll pass them
>>>> on here.
>>>>
>>>> Roger
>>>>
>>>>
>>>>
>>>> On Tue, Apr 5, 2016 at 1:37 PM, Anders Rundgren <
>>>> anders.rundgren.net@gmail.com> wrote:
>>>>
>>>>> On 2016-04-05 21:52, Gregg Kellogg wrote:
>>>>>
>>>>>>
>>>>>>
>>>>>> On Tuesday, April 5, 2016, Anders Rundgren <
>>>>>> anders.rundgren.net@gmail.com <mailto:anders.rundgren.net@gmail.com>>
>>>>>> wrote:
>>>>>>
>>>>>>     On 2016-04-05 19:33, Roger Bass wrote:
>>>>>>
>>>>>>         This seems to be a time for big picture reflections.
>>>>>>
>>>>>>
>>>>>>     Indeed.  IMO we need to start with questions rather than with
>>>>>> answers.
>>>>>>
>>>>>>     The IMO #1 question is: Should the imagined effort depend on new
>>>>>> technology in browsers or not?
>>>>>>
>>>>>>     Why is that important?  Because if the answer is "Yes" it
>>>>>> effectively means that the task is (more or less) owned by the browser
>>>>>> vendors [1].
>>>>>>     If OTOH the answer is "No", I don't see that the effort has a
>>>>>> clear binding to W3C except for marketing reasons.
>>>>>>
>>>>>>
>>>>>> This would seem to indicate that the W3C is only bound to efforts
>>>>>> that end up
>>>>>>
>>>>> > being implemented within browsers, which IMHO is a rather narrow
>>>>> reading of W3C's charter.
>>>>>
>>>>> That's correct.  I don't see that W3C is "authoritative" outside of
>>>>> core Web technology
>>>>> including browsers and a limited set of data formats (XML, JSON_LD).
>>>>>
>>>>>
>>>>> To me, the question is more about if the work builds on and extends
>>>>> web standards,
>>>>> which aren't limited to browser implementations. In particular, if the
>>>>> linked data
>>>>> stack is involved, it's clearly a web standard, but of course, that is
>>>>> not yet clear.
>>>>>
>>>>> Since anybody can develop web standards that aren't tied to browsers,
>>>>> it means
>>>>> that the main purpose for using W3C in this case is for marketing.
>>>>> There's nothing
>>>>> wrong with marketing but the W3C membership fees and process
>>>>> requirements certainly
>>>>> exclude a lot of people who may be needed to succeed.
>>>>>
>>>>> BTW, standardizing "Applications" (which include Web Payments) have
>>>>> proved to be way
>>>>> more difficult than standardizing lower layers (core technology).  If
>>>>> you then add
>>>>> "Security" to the puzzle you quickly reach dead-lock which is why
>>>>> community-driven
>>>>> projects like the Linux kernel never succeeded creating a unified
>>>>> cryptographic
>>>>> architecture like featured in Windows and OS/X.
>>>>>
>>>>> Anders
>>>>>
>>>>>
>>>>>
>>>>>> Gregg
>>>>>>
>>>>>>     Cheers,
>>>>>>     Anders
>>>>>>
>>>>>>     1] Experienced standards editor Ian Hickson explains it pretty
>>>>>> well:
>>>>>>
>>>>>> http://manu.sporny.org/2016/browser-api-incubation-antipattern/#comment-29249
>>>>>>
>>>>>>     "Fundamentally, the people who write the code have all the power.
>>>>>> That’s always been the case"
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>
>
Received on Wednesday, 6 April 2016 01:30:40 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:07:46 UTC