Efficient XML Interchange Working Group Teleconference

30 Aug 2016

See also: IRC log




<scribe> scribe: TK

<scribe> scribeNick: taki


TK: It was published.
... DP sent an announcement on public mailing list.
... We need to think about what is the next step.

DP: I integrated DB's comment.
... We allow dedicated schema.
... We might want to take a look at that.
... I think it is the next logical step.

<dape> http://w3c.github.io/wot/current-practices/wot-practices.html#mapping-to-xml-schema

DP: WoT current practice document already describes how JSON schema is mapped to XML schema. We can do the similar.

<dape> http://w3c.github.io/wot/current-practices/wot-practices.html#mapping-to-xml-schema

<brutzman_> The link didn't work on Internet Explorer.

<brutzman_> TK: w3c.githu.io works on Firefox (Chrome works OK too).

<brutzman_> seems like a styling issue somehow, thanks for the advice

DP: If we have this mapping, we will have real schema-based solution.

<dape> DP: I think we should take a look at object, and array as in the WoT Practices document but also looking at other JSON types...

<dape> ... this would allows us to have a complete mapping to XML Schema and allows us to use schema-informed EXI4JSON representation

<brutzman_> Technical question about example 24: "EDITOR'S NOTE - JSON schema object does NOT define any order. Therefore, in order to capture the constraints of JSON schema object, we need to use xsd:all constructs instead of xsd:sequence."

<brutzman_> Isn't order preserved for children in a JSON object?

DP: In JSON objects, properties order might change.

<brutzman_> www.json.org notes "An object is an unordered set of name/value pairs."

DB: This is an important work that we can capture.

<dape> TK: This new work on supporting JSON Schema and its mapping to XML schema might get a new document or might be integrated in the existing EXI4JSON document. Any Comments?

DP: I think we can first describe it in the appendix section before it becomes stable.
... The document is still very short.

<brutzman_> DP: we can put this information in an appendix in EXI4JSON but would need to keep in synch with WoT Current Practices document

<brutzman_> In some sense we are growing a bit beyond strict EXI4JSON to describe adaptability of XML Schema to JSON.

<brutzman_> Perhaps this is something like "suggested practices for defining data representations in XML Schema that map well to JSON and also EXI4JSON"

<brutzman_> That seems worthy in its own right, we're not trying to correlate "everything XML" with "everything JSON"

<brutzman_> ... it also seems like putting it in EXI4JSON document will let evolve and improve it, our understanding might broaden/deepen with time

<brutzman_> Editorial alternative: we could simply point to WoT Current Practices from EXI4JSON and keep editing with that group.

<brutzman_> extra points award for WoT draft: thanks Daniel for updating the EXI4JSON reference there. 8)

<brutzman_> a true magician is not allowed to reveal how the trick really works..

<brutzman> anyway, with respect to WoT Current Practices, first it is good to keep that section there - a lot of eyes are likely on it, and perhaps even comments will emerge

<brutzman> we should at least point to it from EXI4JSON and think about whether it needs to be adapted or expanded

<brutzman> DP: more belongs to EXI4JSON, we should continue to work on it here.

<brutzman> possibly related issue: defining how to consistently express information in an XML Schema that provides (a) EXI options, (b) EXI4JSON options, and possibly (c) XML Schema to EXI mapping?

<brutzman> having EXI and EXI4JSON options embedded as well-defined hints or annotations in an XML schema would let tools apply EXI consistently with designer's intent

<brutzman> ... and having Current Practices for XML Schema that map well to EXI means that a data-representation designer can achieve best of all worlds simultaneously: XML, EXI, JSON and EXI4JSON

Canonical EXI

TK: I understand all comments were addressed.

DP: We need to upload the schema.

<caribou> http://www.w3.org/2009/exi/options

<brutzman> ... sounds big perhaps, but it appears to simply require following good practices and capturing corresponding options in the XML Schema for anyone to use later.

DP: CB uploaded it. I need to integrate the link.
... We need to convert it into HTML5.
... I will try to fix the content. After that I need CB's help to convert into HTML5.

Ben: I am from Facebook web speed team.
... Desktop, mobile site, etc.
... Infrastructure. We are also working with Browser vendors.
... Better formats down to the client is an option.
... Javascript and CSS, with new compression, we had some improments.
... Parsing text files, css/javascript/html parsing. Javascript is the biggest target. Currently, it is text, and we need to parse the whole document.
... Javascript comes at the top. Binary format for Javascript.

DP: When we started, many device were not ready for general compression scheme.
... You cannot stream, this was another issue.
... EXI reduces the size, and it is typed.
... The goal of JSON and CSS are the same.

<brutzman> (switched browsers to Chrome - IE keeps kicking me off of IRC)

<brutzman> (wide-ranging and interesting discussion with Ben and Nate of Fast Facebook team)

<brutzman> besides near-term opportunities with EXI for JSON and CSS development, we are also looking ahead to HTML5, SVG, perhaps JavaScript itself - essentially Open Web Platform strategy.

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.144 (CVS log)
$Date: 2016/08/30 15:46:35 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.144  of Date: 2015/11/17 08:39:34  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/wtih/with/
Succeeded: s/maps well/map well/
Found Scribe: TK
Found ScribeNick: taki
Present: CB TK DP

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 30 Aug 2016
Guessing minutes URL: http://www.w3.org/2016/08/30-exi-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]