Re: EXI Working Group restart/recharter goals

Still looking for feedback.

Here is a major one:  an ongoing common challenge is picking the "right" compression scheme for JSON.

Wondering if we should tackle the larger problems as part of a restart/recharter.

a. EXI for JSON Recommendation is now available, which applies EXI algorithms to JSON compatibly, so that might provide interesting future options if a growing mix of XML and JSON data is expected.

      EXI for JSON (EXI4JSON)
      W3C Working Draft 23 August 2016
      http://www.w3.org/TR/exi-for-json

b. OpenEXI Nagasena is also an Apache incubator, so pursuable work is to add native support for EXI to Apache http/https server implementations everywhere.  This would likely optimize performance even further, and make EXI compliance an ongoing/maintained part of web servers everywhere.  This is a worthy goal for simple long-term Navy-wide adoption of EXI capabilities.

c. Reference performing extensive quantitative analysis of EXI versus JSON available at

      Hill, Bruce W., Evaluation of Efficient XML Interchange (EXI) for Large Datasets and as an Alternative To Binary Json Encodings,
      Masters Thesis, Naval Postgraduate School, Monterey California, March 2015. Co–advisor Don McGregor.
      https://wiki.nps.edu/display/NOW/2015/06/12/EVALUATION+OF+EFFICIENT+XML+INTERCHANGE+%28EXI%29+FOR+LARGE+DATASETS+AND+AS+AN+ALTERNATIVE+TO+BINARY+JSON+ENCODINGS

d. Must note that people often get different results... in our analysis of EXI versus JSON, with EXI consistently dominating.  It would be good to get clear on differences.  I typically caution against any rush to judgement.  Often there are more options to optimize than might currently get applied, so often the primary conclusion might be that "significant compression and performance efficiencies can be achieved."  Establishing best practices would be good.

e.  The coverage of EXI/XML Post-Schema Validation Infoset (PSVI) provides a high degree of overlap with JSON infoset.  Might we even go all the way to providing common mappings between XML/JSON schemas so that interoperability can become straightforward?

Thanks for considering the possibilities.
  

On 2/12/2018 9:33 AM, Don Brutzman wrote:
> Looking ahead for EXI: there is still a lot of opportunity, and a lot to do!
> 
> Here are potential goals for an EXI Working Group restart/recharter.  All feedback is welcome.
> 
> Background.
> - The EXI working group charter timed out in 2017.  Restart or recharter is necessary for formal W3C approval of new work.
> - Efficient XML Interchange Working Group Charter
>     https://www.w3.org/XML/2015/05/exi-charter.html
> - W3C Process Document, 5 Working Groups and Interest Groups
>     http://www.w3.org/2018/Process-20180201/#GAGeneral
> 
> Motivations.
> - A large amount of important work has been accomplished by EXI working group, with a few tasks pending completion.
> - Further adapting EXI offers significant potential value to a wide range of W3C activities and capabilities.
> - W3C Design Principles: Web on Everything, Web of Data and Services, Web of Trust
>     http://www.w3.org/Consortium/mission
> 
> Restart or Recharter.
> - Restart is possible based on success to date and need for continuing work.
> - Recharter might better engage members in broader goals.
> - In either case, a workshop might help to fully establish wide range of goals supportable by current and prospective W3C members.
> 
> Name. Possible rename to better communicate value proposition of efficient (small+fast+correct) data handling.
> - EXI: classic, known (sorta)
> - noparse: fast, high performance, low memory
> - ED Efficient Data: indicates broad scope
> - others?
> 
> Functional goals.
> - EXI 2 goals have been collected as we have progressed: https://www.w3.org/XML/EXI/wiki/EXI2
> - Aggregate document collections (similar to zip or tar)
> - Recommended (canonical?) EXI encodings for HTML SVG CSS JavaScript JSON MathML
> - Referring to canonical EXI compression of an XML Schema to allow valid out-of-band transmission.
> - Integration into Web infrastructure (http servers, native-EXI XSLT parsers, etc.)
> 
> Application areas.
> - Digital publishing
> - Small devices
> - Big data
> - XML security
> 
> Related W3C Working Groups.  Each group presents needs of related interest, with several being quite major.
> - Automotive WG
> - CSS WG
> - Dataset Exchange WG
> - HTML WG
> - Publishing WG
> - Semantic Web IG (e.g. special handling for triples needed?)
> - SVG WG
> - Technical Architecture Group (TAG) guidance
> - Timed Text WG
> - Web Application Security WG
> - Web Authentication WG
> - Web Assembly WG
> - Web of Things WG/IG
> - Web Performance WG
> - XSLT WG
> 
> Appreciation and thanks for feedback from multiple EXI participants to build this initial list  of issues.
> 
> Looking forward to further dialog on email, also online at the next EXI teleconference 26 FEB 2018.
> 
> all the best, Don


all the best, Don
-- 
Don Brutzman  Naval Postgraduate School, Code USW/Br       brutzman@nps.edu
Watkins 270,  MOVES Institute, Monterey CA 93943-5000 USA   +1.831.656.2149
X3D graphics, virtual worlds, navy robotics http://faculty.nps.edu/brutzman

Received on Monday, 12 March 2018 15:01:37 UTC