XML Processing Model WG

Meeting 207, 26 Jan 2012


See also: IRC log


Henry, Norm, Vojtech, Jim, Alex, Cornelia, Carine
Mohamed, Paul


Accept this agenda?

-> http://www.w3.org/XML/XProc/2012/01/26-agenda


Accept minutes from the previous meeting?

-> http://www.w3.org/XML/XProc/2012/01/19-minutes.html


Next meeting: telcon, 2 February 2012

No regrets heard.

Review of open action items

A-206-01: continued

A-206-02: continued

A-206-03: completed

-> http://lists.w3.org/Archives/Public/public-xml-processing-model-wg/2012Jan/0041.html

Processor profiles WD published

Norm: Yay us.

-> http://www.w3.org/TR/2012/WD-xml-proc-profiles-20120124/

<scribe> ACTION: Norm to setup the last call comment list for new LCWD [recorded in http://www.w3.org/2012/01/26-xproc-minutes.html#action01]

Rechartering troubles

Norm: We got pushback on the charter; one of the possible solutions is to put FPWD of 2.0 in the charter.

Carine: We got pushback because the goals weren't defined clearly enough.
... I was also surprised to see that issue of Rec-track documents as pushback. We've done that before for requirements and use cases.

Some discussion of the clarity of our goals.

Norm: So, basically, if we want to do V.next, we'll need to have a workshop or some other event to gauge interest. And if we don't put V.next in the charter, we won't get chartered.

Carine: I think that's basically the case.

Norm: Liam suggests putting FPWD of 2.0 and the possibility of a workshop in the charter. Maybe we should do that.

Cornelia: And why wouldn't we do that, isn't that what we want to do?

Henry: Yes, but earlier conversations suggested that we weren't ready to do 2.0.

Alex: But we have community feedback for 2.0

Henry: Yes, I think Liam just needs help writing that: point to wiki, point to mailing list.

<scribe> ACTION: Norm to work with Liam to get a new charter proposal drafted along those lines. [recorded in http://www.w3.org/2012/01/26-xproc-minutes.html#action02]

XProc V.next discussion

Norm: I sent a list of low-hanging fruit items.

-> http://lists.w3.org/Archives/Public/public-xml-processing-model-wg/2012Jan/0041.html

Norm: And Vojtech observes that it doesn't anything about non-XML.
... And I think maybe we could do something smallish about that.

Vojtech: I was thining especially about small stuff.
... Like if you could save binary data, with p:store.
... To make it more symetrical. We have p:load with p:document and p:data but we have nothing to store binary data.

Norm: Yes, and an option on p:store seems pretty straightfoward.

Vojtech: What I did is just what Norm did, I added an extension to p:store. But mine was a bit more generic in the sense that both and XML and non-XML data can flow through the pipeline. Whatever the p:store gets, it saves it.
... I was thinking about p:document and p:data and their relationship. At the moment I didn't want to change that much. I changed p:data so that it can produce binary data that's not base64 encoded.
... But I wonder if p:document, if you point it to binary data, whether it should do the same thing. Or if p:data should return XML.

Norm: We should consider a proposal to do some work in this area; being able to load XML, HTML, JSON, etc.

Jim: In the past, have we ever talked about p:document*s*?

Henry: Yes, the possibility of having a set of documents flowing through the pipeline was there in the Markup pipeline. We did discuss it briefly, a while ago.
... But it's not low-hanging fruit. You have to talk about how to generate names for these things; it really has to be a map so that steps down the pipeline can extract documents from the set.

Jim: But multiple p:document elements can be used. That might let you implement something like an ant fileset.

Vojtech: So like in ant, you could specify a base URI and some sort of mask, so you get a sequence of files.

Jim: it's a little awkward to work with sets of files and baking in at that level would remove the contortion from some pipelines.

Norm: That seems like it might be low hanging fruit; you could implement it yourself.

Jim: Yes, but it wouldn't bake in at the p:input level.

Norm: Yeah, I can see that.
... I'll add that to the low-hanging fruit.

Alex: With AVTs, I can imagine that we might be able to do the same sort of thing with HTTP URIs.
... So if you had a set of documents, you could iterate with numerical positions, perhaps.

Vojtech: You can also imagine doing this on the p:load step.

Jim: I've got one other thing. I did an experiment with my implementation, I enabled "AVT-everywhere".

Norm: You mean you made "{" and "} expand everywhere all the time.

Vojtech: But what if you want to include an XSLT pipelien?

Jim: You can turn it on and off.

Norm: I don't understand how that works.

Jim: There are lots of details; I just made them up.

Alex: That sounds a little bit like an alternative pipeline syntax.

Norm: I'd like to see some examples.

Norm runs through his list

<jfuller> completely agree on xpath 2.0 going forwards

Vojtech: There's also the question of optional not-specified options.

<jfuller> https://github.com/jpcs/rbtree.xq

Discussion inevitably returns to parameters.

Cornelia: I think another area we should be looking at is mashup-technologies. Those should be using XProc.

<jfuller> be great Cornelia if you have any links to mashups tech

Norm: I'll summarize again the low-hanging fruit and then I'd like everyone to think about whether or not that list is complete.

Any other business

None heard.

Summary of Action Items

[NEW] ACTION: Norm to setup the last call comment list for new LCWD [recorded in http://www.w3.org/2012/01/26-xproc-minutes.html#action01]
[NEW] ACTION: Norm to work with Liam to get a new charter proposal drafted along those lines. [recorded in http://www.w3.org/2012/01/26-xproc-minutes.html#action02]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.136 (CVS log)
$Date: 2012/01/26 21:48:59 $