See also: IRC log
<dom> [jo lists out the information we may need to record as part of the audit trail]
<trackbot> Tracking ISSUEs and ACTIONs from http://www.w3.org/2005/MWI/BPWG/Group/track/
noted that Dom disagrees that we are a robot so we do not look at robots.txt
<dom> (and we don't do recursive retrieving)
it is not necessary to add a referer[sic] field
<dom> HTTP Request
noted that the extra request(s) probably HEAD will suffice need to be made for caching 'if-none-match' etc.
open questions on Tidy - character encoding, defaulting, normalisation, etc. Need to think abut what we want here
<dom> Sides-based presentation of TagSoup
<dom> CSS Validator source code
Noted that invalid certificates (expired, unrecognised authority) warn
<dom> gtkmozembed.MozEmbed
<dom> SVG of the Mindmap resulting of our morning discussions
<dom> Jo will wrap up the requirements document, including the mind map document
<dom> Sean will send an updated architecture diagram
<srowen> ACTION: Jo to complete a final draft of the requirements statement doc in the next week or two [recorded in http://www.w3.org/2007/04/03-bpwg-minutes.html#action01]
<srowen> ACTION: Jo to complete a strawman of the intermediate doc based on mind map in the next week or two (leading to a schema later) [recorded in http://www.w3.org/2007/04/03-bpwg-minutes.html#action02]
<srowen> ACTION: Jo to complete a strawman draft of the results document in conjunction with Shadi [recorded in http://www.w3.org/2007/04/03-bpwg-minutes.html#action03]
<srowen> ACTION: srowen to circulate a next draft of architecture / design doc in the next week or so [recorded in http://www.w3.org/2007/04/03-bpwg-minutes.html#action04]
<srowen> ACTION: srowen to update / rewrite code after architecture / design doc has been finalized and begin development soon [recorded in http://www.w3.org/2007/04/03-bpwg-minutes.html#action05]
noted that we ought to look at the JHOVE library
Ruadhan observed that we're likely to use them
jo wondered if we would use W3C validator or JHOVE for XHTML
Dom commented that we'd probably just just whatever comes with the XML parser we end up using
<scribe> scribenick: Jo
<scribe> ACTION: Sean to contact JHOVE in pursuit of his actions on code over the next couple of weeks [recorded in http://www.w3.org/2007/04/03-bpwg-minutes.html#action06]
Noted that the decision was made to use local copies of DTDs to avoid counterfeiting
Dom: plus in general we should use cached copies as much as possible to reduce our retrieval footprint
<srowen> http://lists.w3.org/Archives/Public/public-mobileok-checker/2007Mar/0047.html
[sean discusses general framework]
Srowen: Question 1 is to what
extent should output also be available as pure Java
objects?
... it would be nice for any program to use, but would be a lot
of development work
... possibly restrict this to the results and not expose all
the preprocessing as Java objects
[no disagreement]
srowen: Java 5
... default Java XML parser?
jo: what is it?
srowen: not sure may be Xerces
dom: doesn't matter there is a common interface
srowen: Tag soup
jo: noted that we said we needed to study what we mean for tidying more carefully
srown: jakarta commons HTTP client
jo: does that give access to raw HTTP headers?
ruadhan: think so
srowen: using JHOVE
... also planning to use some kind of CSS parser -
implementation of SAC (simple api for CSS)
... basic class diagram
jo: UML?
dom: rough schema of how planning to organise code
sro: key classes, key flow
jo: threading?
sro: yes worth mentioning in this document
sro: valid dom is main goal
jo: also character encoding tidying too?
jo: need to discuss how the testing will be carried out
sro: yes we need to discuss the approach
jo: need to start work on this
dom: yes, we need to start documenting expected outcomes
<scribe> ACTION: Dom to start work on gathering test cases and documenting expected outcomes [recorded in http://www.w3.org/2007/04/03-bpwg-minutes.html#action07]
jo: need to create a gap analysis
dom: will create a list in CVS of test cases against tests as part of above action
sro: architecture document will discuss how tests are to be defined
jo: does shadi have a view from EARL on how the expected outcome should be compared with an actual outcome
dom: in our case we'd check
whether a specific warning is given too
... which means we'll have to have specific error codes for
each one
Jo: we'll need an RDF vocab for this
dom: yes but we don't need to be public about it
[following discussion proposed 12 13 June in London]
<scribe> ACTION: Sean to announce to list proposed meeting date and location [recorded in http://www.w3.org/2007/04/03-bpwg-minutes.html#action08]
<dom> [thanks dotMobi for hosting this first great meeting!]
<scribe> ACTION: Sean to find out about Google hosting meeting [recorded in http://www.w3.org/2007/04/03-bpwg-minutes.html#action09]
<dom> Thanks!
[noted]
Thanks also to Shadi for coming and helping
sro: adjourned