W3C

- DRAFT -

XProc telcon

12 Jun 2008

Agenda

See also: IRC log

Attendees

Present
Paul, Mohamed, Alex, Henry, Vojtech
Regrets
Richard, Norm, Rui, Andrew
Chair
Henry S. Thompson (pro tem)
Scribe
Henry S. Thompson

Contents


Last week's minutes: http://www.w3.org/XML/XProc/2008/06/05-minutes.html

Accepted.

Next meeting: 19 June 2008, regrets from Andrew only ones known as yet

split-sequence change

http://lists.w3.org/Archives/Public/public-xml-processing-model-wg/2008May/0040.html

MZ: Features suggested by Alessandro's suggestions based on Haskell:
... p:is-empty is covered by limit attr on p:count
... p:pack has been added
... split a sequence up until the first failure to match some pattern still needs a way to happen

http://www.w3.org/TR/xproc/#c.split-sequence

HT: Sounds straightforward
... anyone see any difficulties?

MZ: This gives us functionality for dealing with sequences which it's difficult, if not impossible, to get any other way

HT: Better name than 'stop-test-after-first-false' ?
... We get the initial subsequence which matches

AM: I get it

HT: How about 'initial-only'

AM: It's going to be opaque, people will have to look it up to understand it

RESOLUTION: Add an 'initial-only' attribute to p:split-sequence, a boolean

<scribe> ACTION: Alex Milowski to add an 'initial-only' attribute to p:split-sequence, a boolean [recorded in http://www.w3.org/XML/XProc/2008/06/12-minutes.html#action01]

Serialization question

http://lists.w3.org/Archives/Public/public-xml-processing-model-wg/2008Jun/0008.html

AM: The origin of the impl-defined for the defaults for unspec'd serialization options comes from the QT Serialization spec.

MZ: That's not what I was concerned about, rather that I read the spec. in 5.6 as allowing other attributes which are not specified in the spec.

HT: Oops, this section needs to be cross-referenced from 7.3

AM, MZ: Discuss what 'unspecified' means here

HT: I think that the intent was that 'unspecified' refers to _options_ which are missing for a particular step

AM: Right

MZ: Yes

HT: But, alas, the Serialization spec. does not say that defaults are impl-defined
... Propose to delete the "default value" sentence from 5.6

RESOLUTION: delete the "default value" sentence from 5.6

HT: Propose to amend the reference to 7.3 in 5.6 to read as follows:

The semantics and defaulting behaviour of the attributes on a p:serialization are as described for the corresponding options Section 7.3, “Serialization Options”.

RESOLUTION: Amend the reference to 7.3 in 5.6 to read as follows: "The semantics and defaulting behaviour of the attributes on a p:serialization are as described for the corresponding options in Section 7.3, “Serialization Options”."

<scribe> ACTION: Alex Milowski to draft a Note to add to 7.3 explaining that we don't give simple defaults, behaviour wrt missing options is complex and you have to read [Serialization] to find out. [recorded in http://www.w3.org/XML/XProc/2008/06/12-minutes.html#action02]

<scribe> ACTION: Alex Milowski to add an error to 7.3 to cover all other parameter-related Serialization errors [recorded in http://www.w3.org/XML/XProc/2008/06/12-minutes.html#action03]

C14N support

MZ: My email also proposes adding support for c14n to serialisation

HT: I am opposed, it's a new feature, and it can be easily fitted into the existing spec. as an impl-defined serialization method

AM: I'm opposed also, we should wait for Serialization spec. to provide for this, so we don't find ourselves isolated when they do

MZ: So, do you mean there would have to be two new methods, x:c14n-with-comments and x:c14n-without-comments

HT: No, I think you would have new options to go with impl-defined x:c14n method

AM: What happens when Serialization does add support for c14n -- do those attributes/options move from prefixed to unprefixed?

HT: MSM would say "we should say 'Serialization or its successors'"

AM: The problem is that it's not hard to allow for new serialization options, when there's a new Serialization spec., but adding attributes in no namespace to p:serialization is not allowed
... Maybe we should provide for this ahead of time, with a specific namespace.

HT: We're out of time, AM please start an email thread on your idea, we'll pick it up next week.

HT: Adjourned

Summary of Action Items

[NEW] ACTION: Alex Milowski to add an 'initial-only' attribute to p:split-sequence, a boolean [recorded in http://www.w3.org/XML/XProc/2008/06/12-minutes.html#action01]
[NEW] ACTION: Alex Milowski to add an error to 7.3 to cover all other parameter-related Serialization errors [recorded in http://www.w3.org/XML/XProc/2008/06/12-minutes.html#action03]
[NEW] ACTION: Alex Milowski to draft a Note to add to 7.3 explaining that we don't give simple defaults, behaviour wrt missing options is complex and you have to read [Serialization] to find out. [recorded in http://www.w3.org/XML/XProc/2008/06/12-minutes.html#action02]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.133 (CVS log)
$Date: 2008/06/12 16:16:04 $