[wg11] Re: Part 28 in Seattle?
Martin Hardwick
hardwick at steptools.com
Mon Sep 13 13:26:55 EDT 2004
David,
Gosh you got me. The community wants a sensible STEP XML.
This means:
a. OSEB is disqualified because it puts data into XML attributes.
These attributes are supposed to be used for control flags
not information.
b. ETEB is disqualified because it puts typing data into the XML
(specifically the inheritance information is included).
We have been selling an OSEB tool (under duress) for about three years.
No one asks for it twice.
I guess I should have extended my first list of requirements to include:
3. The STEP/XML binding should be at least as easy to understand as P21
and at the very least it should not be harder to understand.
Most people are expecting more than this because nesting can make data
easier to understand.
Specifically, XML gives you the opportunity to aggregate values into
larger structures that make the meaning of the data easier to understand.
ETEB sort of understood this but chose the wrong data to aggregate.
OSEB was just P21 with labels.
If we drop the requirement to translate EXPRESS into another schema
language then it becomes very easy to allow aggregations (in the full
sense of the word not just the EXPRESS sense) in the exchange file.
Martin
At 06:05 PM 9/13/2004 +0100, you wrote:
>Hello Martin,
>
>>From all this one could surmise that you're now a big fan of Part 28 Edition
>1 since it seems to satisfy all the requirements you mention without any of
>the drawbacks. Now, we just need to pick one of the bindings. Which is it,
>OSEB or ETEB?
>
>Cheers,
>David
>
>> -----Original Message-----
>> From: wg11-bounces at steptools.com
>> [mailto:wg11-bounces at steptools.com] On Behalf Of Martin Hardwick
>> Sent: 13 September 2004 17:38
>> To: edbark at nist.gov; Dave Loffredo
>> Cc: SC4 WG11
>> Subject: Re: [wg11] Re: Part 28 in Seattle?
>>
>>
>> All,
>>
>> We (STEP Tools) will not invest in P28 until we see
>> a way for it to be used for conformance tested data
>> exchange.
>>
>> As soon as we believe the project is going to produce
>> something that can be used for conformance tested data
>> exchange then we will make every resource available
>> because the user community really, really wants to use
>> XML to exchange STEP data.
>>
>> In our opinion there are two obstacles that prevent the
>> current P28 being used for conformance tested data
>> exchange:
>>
>> 1. P28 is used to generate an XML Schema that is then used
>> to defined the exchange.
>>
>> This leads to a situation where the blind are leading the
>> blind. The STEP community does not fully understand XML
>> Schema and will never be sure what is allowed by a 1,000 plus
>> lines of XML Schema definition. The XML Schema community does
>> not understand the STEP models and has no need or motivation to do so.
>>
>> This needs to be fixed by using EXPRESS to define XML data.
>> Projects will then be able to define alternative schemas for
>> that XML data using DTD's, XMLSchema, Relax or ancient Etruscan.
>>
>>
>> 2. P28 has far too many configuration options.
>>
>> The original intent of these options was to handle legacy
>> data without forcing the mainstream users to adopt the stupid
>> choices made for that legacy data.
>>
>> However, the options have morph'd into a language to allow any
>> XML Schema to be adopted as a STEP Schema because a
>> configuration exists to map the EXPRESS model into that XML
>> Schema. Recently this has become worse because it is no
>> longer necessary to prove the mapping using an execution engine.
>>
>> The configurations need to be removed. If someone wants to
>> morph XML data defined by EXPRESS into another model they
>> should use XSLT like everyone else.
>>
>> This project has been an example of everything that can go
>> wrong with the ISO standard process. It is too easy to be an
>> XML expert, too tempting to develop a business plan that
>> requires capture of the STEP market, and too difficult for
>> the ISO process to stop these hijackings.
>>
>> Martin
>>
>> At 01:12 PM 9/9/2004 -0400, Ed Barkmeyer wrote:
>> >Dave Loffredo wrote:
>> >
>> >>I have not yet heard from anyone regarding this ...
>> >> - Dave
>> >>------------------------------
>> >> From loffredo Fri Sep 03 13:33:57 -0400 2004
>> >>From: Dave Loffredo <loffredo at steptools.com>
>> >>To: edbark at nist.gov
>> >>CC: hpreston at ebmail.gdeb.com, staub at pdtec.de,
>> >> david.price at eurostep.com, Jochen.Haenisch at epmtech.jotne.com,
>> >> Masaru.Suzuki at unisys.co.jp, abf at cme.nist.gov,
>> >> hardwick at steptools.com, bgischne at ebmail.gdeb.com
>> >>...
>> >>Ed> I expect that Heidi et al. will have posted a "proposed DIS" of
>> >>Ed> Part 28 to the server by 15 September, after which
>> Heidi will be
>> >>Ed> taking a rewarding leave of absence. I would suggest we (i.e.
>> >>Ed> you) schedule some official "WG11 review" of that draft in
>> >>Ed> Seattle, prior to releasing that document, or its immediate
>> >>Ed> successor, to DIS-FDIS ballot.
>> >>Hi Ed, I am certainly happy to schedule any meeting or
>> review session
>> >>that the Part 28 team would like to hold. What day/time is
>> best and
>> >>who will be leading the review?
>> >
>> >1) Because of funding limitations and other commitments, NIST can
>> >provide no further resources for the development of Part 28 after 15
>> >October 2004.
>> >
>> >2) After discussion with Heidi, I now believe it is unlikely that a
>> >complete DIS draft, incorporating text for all agreed-on
>> ballot comment
>> >resolutions, will be available before she goes on maternity leave.
>> >
>> >3) It is desirable that an interim editor be appointed. If
>> no interim
>> >editor is appointed, it is likely that Part 28 v2 will go
>> into limbo,
>> >awaiting commitment of resources to complete the DIS.
>> >
>> >4) If a number of the NB representatives will be present in
>> Seattle, it
>> >would be useful to schedule a session to ensure that the proposed
>> >resolutions are satisfactory, and possibly to develop/repair
>> explicit
>> >text for them. Heidi referred to this as a "continuation of
>> the Groton
>> >Editing meeting".
>> >
>> >5) If a number of the NB representatives will be present in Seattle,
>> >and believe such a meeting worthwhile, I would be willing to
>> chair that
>> >meeting. If, however, an interim editor is appointed, s/he
>> may better
>> >chair such a meeting.
>> >
>> >6) If, on the other hand, Japan, Norway, Sweden, and the
>> United Kingdom
>> >(the NBs who commented on the CD) would not be represented at such a
>> >meeting, it would be reduced to a working session of such technical
>> >experts as may be present, and it is not clear that that has much
>> >value.
>> >
>> >4 hours on any of the days is acceptable to NIST. (If there is need
>> >for additional technical discussion, we will find a time and
>> place to
>> >continue.) I prefer that the meeting not conflict with the
>> Part 25 v2
>> >session.
>> >
>> >Related topic, in expectaion of imminent change of the guard:
>> >
>> >For the plenary, I recommend that SC4 WG11 acknowledge
>> Electric Boat's
>> >unflagging commitment of resources, both technical and editorial, to
>> >the development of an XML representation standard for the standard
>> >exchange schemas developed in SC4, and specifically acknowledge the
>> >personal contribution of Heidi Preston to this effort, as Editor of
>> >Part 28 v2, whose only reason for abandoning us now is that she has
>> >truly more important commitments!
>> >
>> >If I am in Seattle at the appropriate time, I would be happy to make
>> >this motion in person. But if not, I would be grateful to
>> the Convenor
>> >if he could find a willing participant to do so in my stead.
>> >
>> >-Ed
>> >
>> >--
>> >Edward J. Barkmeyer Email: edbark at nist.gov
>> >National Institute of Standards & Technology
>> >Manufacturing Systems Integration Division
>> >100 Bureau Drive, Stop 8264 Tel: +1 301-975-3528
>> >Gaithersburg, MD 20899-8264 FAX: +1 301-975-4694
>> >_______________________________________________
>> >wg11 mailing list
>> >wg11 at steptools.com http://lists.steptools.com/mailman/listinfo/wg11
>> >
>> >
>> >
>>
>> _______________________________________________
>> wg11 mailing list
>> wg11 at steptools.com http://lists.steptools.com/mailman/listinfo/wg11
>>
>
>
>_______________________________________________
>wg11 mailing list
>wg11 at steptools.com
>http://lists.steptools.com/mailman/listinfo/wg11
>
>
>!DSPAM:4145d38b97161075691429!
More information about the wg11
mailing list