----- Original Message -----
Sent: Tuesday, January 14, 2003 7:10 PM
Subject: RE: Comments on the recent EDIINT AS2 v12
draft
Rik,
In
the hope that we can reach a reasonable solution to this
matter, I will respond to your points. My comments are bounded
by <db> and </db>.
History of v12 draft:
-
after the as2 v2 draft we attempted to combine the then current as2 draft
with
gisb standard used in the energy industry.
<db> As I
recall, it was both the Energy, represented by NAESB/GISB (http://www.naesb.org) and Automotive
industries, represented by AIAG, http://www.aiag.org that contributed the
sections in AS2 you are referring to.
</db>
- during the recent two
interop rounds with twenty plus products, implementing
the non gisb part, it became apparent that the v11 draft was very confusing
because
of the attempt to combine the two.
<db>The two
interop rounds you refer to were sponsored by UCC and facilitated
by Drummond Group. These tests were conducted in a closed forum and only
those companies that were willing to pay UCC/DGI the
entrance fee (originally $15,000 now I believe it's $35,000) are
allowed to participate. The test plan, which Drummond
Group developed, explicitly excluded certain parts of the AS2 specification (the Energy/AIAG
parts) . However the test is marketed/promoted
as an "AS2 test" even though significant parts of AS2 are not being tested.
To
my knowledge, neither the test plan nor any of the testing experiences or results were
discussed on the EDIINT WG list before, during or after the tests were
conducted. The only public information that I'm aware
of regarding these tests is available at www.drummondgroup.com,
ref:
I
see several issues with the current testing
process:
1.
The entrance fee is a barrier to entry for some
companies
2.
The test plan has not
approved through the IETF process.
3. The test unfairly excluded important parts of
AS2 functionality, which the Energy and Auto industries contributed. Niether
were these industry groups given an opportunity to express
their opinions, because the work took place outside the IETF process in
a private forum.
</db>
- the twenty plus
vendors interop testing help make the v12 draft clearer
<db>The twenty plus vendors, operating in private, changed the
AS2 specification without so much as consulting the co-authors of the spec.
The fact that these vendors aren't implementing certain sections of AS2
does not give them the right to remove them. I know of several AS2
implementations that were negatively affected by this decision. Why weren't
these implementers involved in the discussion to make these
changes? The answer is because the discussions did not follow the IETF
process. None of the proposed changes in V12, including the removal of
an AS2 co-author, were discussed on the EDIINT WG list.
</db>
- Dick
Brooks called very concerned that I dropped the gisb info from the v12
draft
- I asked that he send me the gisb standard and I would append it to
the current v12 draft, call it v13, to
facilitate discussion, while ensuring the non gisb part has clarity,
which it did not have in v11
<db>You state the v11 spec lacked
clarity, I agree it is a bit rough. However this was not a hinderance
for numerous implementers of the V11 spec, as indicated by the number
of interoperable products
listed in your press release, plus the number of production implementations in
the Energy industry (over 50), which I'm aware
of.
Additionally, all of the text you've requested from
me is already available to you in the V11 draft of AS2. Simply cut and
paste the parts of V11 that were
removed in V12 and you
will have what you' ve requested.
</db>
The suggested process at this time:
-
please review the contents of the v12 draft for clarity and correctness for
the non gisb part
- once this is complete we will discuss the gisb
part which Dick Brooks will send me and I
will include in the next release v13
-
Then we will discuss (based on the old v11 draft) if there is a means to
combine them
in the same draft/document
<db>I respectfully disagree. I firmly believe it would be
less work and more efficient to start with V11.
I
propose an alternative approach which begins with the V11
spec. Start by incorporating text from V12 into V11 and the
resulting V11+ spec be issued as draft V13 to the EDIINT WG for
discussion.
I
also propose the following enhancements to improve the
overall process:
-
Conduct all discussion regarding AS2 spec changes on the EDIINT list server
so the entire EDIINT community may participate in the
discussion
-
Seek consensus from the EDIINT community with regard to testing plans.
Testing results
should also
be disclosed/ discussed in the open
on the EDIINT WG list
-AS2 testing should not be cost
prohibitive so that all interested parties may
participate
-
Testing should execrise as much of the AS2 technical specification as
possible, ideally this would be 100% of the defined spec, but that may not
be achievable.
</db>
Non technical issue:
- the
energy industry refers to the gisb spec as as2
- the
retail industry and others using the non gisb part of the spec refers to it
as as2
- both, in some cases mandate the use of as2 specs.. Which
means we have a naming issue
if we decide to split them into two specs
-
fairness is the key to resolving this issue, especially on the use of the
as2 name.
Please review v12 asap for clarity and
correctness…
<db>I've worked in the IETF since 1992 and I
believe the IETF process is fair and open. We simply need to follow the
defined process and use the EDIINT WG list and F2F IETF meetings to hold
open discussions on matters pertaining to AS2.
This will ensure the broadest
exposure to interested parties and provides an opportunity to
participate in a fair and open process.
I
believe there is an opportunity now to take a step in this direction by
issuing a V13 draft using the alternative approach described
above.
</db>
Dick Brooks
Systrends, Inc
7855 South River Parkway,
Suite 111
Tempe, Arizona 85284
Web: www.systrends.com <http://www.systrends.com>
Phone:480.756.6777,Mobile:602-684-1484,eFax:240-352-0714
History of v12 draft:
- after the as2 v2 draft we attempted to combine the then current as2
draft with
gisb standard used in the energy industry.
- during the recent two interop rounds with twenty plus products,
implementing
the non gisb part, it became apparent that the v11 draft was very
confusing because
of the attempt to combine the two.
-
the twenty plus vendors interop testing help make the v12 draft
clearer
-
Dick Brooks called very concerned that I dropped the gisb info from the
v12 draft
- I asked that he send me the gisb standard and I would
append it to the current v12 draft, call it v13,
to facilitate discussion, while ensuring the non gisb part
has clarity,
which it did not have in v11
The suggested process at this time:
-
please review the contents of the v12 draft for clarity and correctness
for the non gisb part
-
once this is complete we will discuss the gisb part which Dick Brooks will
send me and I
will include in the next release v13
-
Then we will discuss (based on the old v11 draft) if there is a means to
combine them
in the same draft/document
Non technical issue:
-
the energy industry refers to the gisb spec as as2
-
the retail industry and others using the non gisb part of the spec refers
to it as as2
- both, in some cases mandate the use of as2 specs..
Which means we have a naming issue
if we decide to split them into two specs
-
fairness is the key to resolving this issue, especially on the use of the
as2 name.
Please review v12 asap for clarity and
correctness…
Best regards,
Rik Drummond
EDIINT Chair
<<...>>
