XML Processing Model WG

06 Sep 2012


See also: IRC log


Norm, Alex, Jim, Mohamed, Cornelia


Accept this agenda?

-> http://www.w3.org/XML/XProc/2012/09/06-agenda


Accept minutes from the previous meeting?

-> http://www.w3.org/XML/XProc/2012/08/30-minutes


Next meeting: telcon, 13, 20, 27 Sep? 4 Oct?

<jfuller> I can be there on all thurs

Norm gives regrets for 13, 20, and 27 September.

Cornelia gives regrets for 13 September.

Next meeting: 13 September, Alex to chair.

Use cases and requirements

-> http://www.w3.org/XML/XProc/docs/langreq-v2.html

Alex: We have a very large document, with lots of stuff we can do. Now that we've been sitting on it for a while, have we really settled on a strategy and what the next version should be?

jfuller, http://www.w3.org/XML/XProc/docs/langreq-v2.html

Alex: There are a lot of things we could do, but what's our strategic goal? How do we make it easier to use?
... It includes simplifying things like parameters but also includes an inventory of steps.
... We could go off the deep end on steps.
... One strategy would be to pick a handful of items.

Jim: I thought we had agreement on a strategy that was more limited.

Alex: But can we be more specific.

Norm: I'll put a stake in the ground, fix parameters and allow non-xml documents to flow through the pipeline.
... I think that's the bare minimum, plausible easier to use story.

Norm: I think the inventory of steps is a red herring, we can do that in notes

Alex: I think it would be nice to look at what we have and see if there are any obvious bits we need to address.

Cornelia: I have a question too, what about a compact syntax

Norm: I have two of those. That was the subject of my Balisage paper.

Alex: I feel that way sometimes too.

Jim: Is that a note or part of a V2?

Alex: I don't know.

Norm: I don't know either.

Jim: Programming in a markup language always requires a certain mindset that leaves some programmers behind.

-> http://www.balisage.net/Proceedings/vol8/html/Walsh01/BalisageVol8-Walsh01.html

Jim: A compact syntax might broaden adoption.

Cornelia: One of the barriers is that uptake hasn't been that big.
... Anything we can do to help would be a good idea.

Jim: I think it's early days for XProc. I think there's more use of it than we think.

Norm: Yes, it turned up a bunch of times at Balisage, very heartwarming :-)

Cornelia: I agree. The non-XML thing is a higher priority than a compact syntax.
... I think compact syntax is still on the short list then.

<jfuller> http://depx.org/

Jim: I did an experiment with dependency management; it's broader than just XProc, it includes XQuery, XSLT, etc.
... It's on github, so you can contribute your own steps.
... We could slice this to be explicitly for XProc.
... What depx is is a client, written in XProc, that provides a library of steps for interacting with the depx repository.
... Do we want to say anything about the automatic loading of steps?
... If you want to use a step, you have to explictly define it and such. Do we want any sort of automatic dependency injection mechanism.

Norm: Do you mean automatically inject the decl or the impl?

Jim: The declaration.

Norm: Uh. I'm not sure that has a whole lot of appeal, but I'll think about it.

Jim: I'm trying to shorten the chain: I see a step on depx I want to use, I push a button, and it works.

Norm: I think that could be confusing. We decided to require the declarations for interoperability.

Cornelia: What would be required of the XProc standard?

Jim: Maybe nothing.

Norm: We'd have to change the standard to say that we could use steps that aren't explicitly declared in scope.
... I could be persuaded to allow "magic" declarations.

Alex: The XQuery folks allow all sorts of implementation-defined stuff.
... And having a declaration doesn't actually require the implementations to do the same thing.

Norm: Let's see if we can take some concrete steps.
...Parameters: I'm on the hook for a concrete proposal.
... Non-XML documents: Cornelia will ask Vojtech to submit a proposal
... Compact syntax: Look at my Balisage paper :-)
... Implicit declarations: Jim, send something to the list.
... New steps: propose them in groups for notes

Norm: If we can drive these to concrete proposals by mid-October, then maybe we can declare victory.

Some discussion of TPAC in Lyon.

Norm: If we can get proposals by mid-October, maybe we can wrestle the details to the ground at the f2f and come out with apublication plan.

Alex: We have a bunch of outstanding action items, what's the strategy for the next few weeks.

Norm: Do what you can. :-)
... I'll be online in two weeks if I can.

Jim: Alex, I might have a first draft of the zip/unzip thing.
... Where is the language v2 requirements and use cases.
... Are we going to publish that?

Alex: We have one, but I don't think it's ready to publish.

Norm: If we've reached a strategic decision, then we can update the document to reflect that and move some of the other material off into separate reference documents.
... But we need Murray, Henry, and Vojtech to consider our planning of today.

Alex: I'd like a more concise statement of our strategy.

Norm: I can do that.

Jim: I'm a little bit concerned about the title of the document.
... In my mind, I'm thinking of this document as a V2 of the original requirements document.
... It's a tool to use to create the new set of requirements.
... One more comment, in the v2 document as it is now, how do people feel about the requirements that aren't satisfied.

Alex: Off the top of my head, we didn't do things like digital signatures. I think on the other hand we did meet a lot of our requirements.
... Some of them require custom steps.

Norm: It would be useful to categorize the ones we didn't satisfy as really unsatisfied (digital signatures) or actually satisfied but requires some custom step or web service that could easily exist but doesn't.

<scribe> ACTION: Alex to categorize the unmet use cases as really unmet or met with external dependencies. [recorded in http://www.w3.org/2012/09/06-xproc-minutes.html#action01]

Any other business?

None heard.


Summary of Action Items

[NEW] ACTION: Alex to categorize the unmet use cases as really unmet or met with external dependencies. [recorded in http://www.w3.org/2012/09/06-xproc-minutes.html#action01]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.136 (CVS log)
$Date: 2012/09/06 14:55:04 $