Here is the first of the four messages. All four are from me. The next
three messages didn't actually get out until two days later (wed October 21,
around 9:00 AM, when the PWG server was rebooted). That may be why you
couldn't find them.
We agree that the June draft are the specification that we are implementing.
However, there was the suspicion that the natural language override
NLO 3 of 4 - for Get-Jobs at the job level
NLO 4 of 4 - at the attribute level for all requests and responses
was not really implemented completely. We agreed at the telecon, that we
really do need the test scripts to test this. Then people could make more
intelligent votes. Unfortunately, the scripts that I wrote last week
uncovered a bug in the test tool. I hope to have some scripts tomorrow that
we can all try. If there are some implementations that implement the June
draft properly, then we will probably agree not to change them. One
compromise suggested today, is to change a MUST to a MAY for the job-level
override in Get-Jobs and to make using textWithLanguage and nameWithLanguage
a SHOULD. That would allow implementations that conform to the June draft
to continue to be conformant.
See the minutes and resulting action items from today's telecon for more
>From: Hastings, Tom N [mailto:hastings at cp10.es.xerox.com]
>Sent: Monday, October 19, 1998 23:16
>Subject: IPP> MOD - NLO 1 of 4: Overview of Natural Language Override
>(NLO) Sim plification
>>>We've had some e-mail and telecon discussions to clarify and
>natural language override (NLO) mechanism, without changing any of the
>functionality. At last week's telecon, we decided to put the
>question to a
>vote on the mailing list for two week period. Bob Herriot and
>have reviewed these four mail messages. The next 3 mail messages will
>> NLO 2 of 4: a clarification to allow the NLO to be used redundantly
> NLO 3 of 4: a vote to eliminate the job level override for Get-Jobs
> NLO 4 of 4: a vote to eliminate the attribute override by
>the data type which specifies the natural language explicitly.
>>A vote to accept will mean a change for those implementations
>to the June 1998 specifications. A vote to reject should mean that an
>implementation either already conforms to the June
>specification or will be
>changed to conform.
>From: Paul Moore [mailto:paulmo at microsoft.com]
>Sent: Tuesday, October 27, 1998 18:17
>To: 'ipp at pwg.org'
>Subject: IPP> NLO votes
>>>Can somebody please state what the proposed changes are. I
>have tried to
>find the orginal proposal somewhere in the mail threads and cannot.
>>I will remind peole that we voted to that ipp1.0 was done
>other then the
>issues we raised at the bake-off. This was not raised then or
>Functioning interoperable implmentaitons can be built using the current
>spec. What suddenly changed?