Telcon minutes, Wednesday, Sep. 10

Online at http://www.w3.org/2014/09/10-webfonts-minutes.html
and as plain text below:

                               - DRAFT -

                 WebFonts Working Group Teleconference

10 Sep 2014

   See also: [2]IRC log

      [2] http://www.w3.org/2014/09/10-webfonts-irc

Attendees

  Present
          [Google], Vlad, [Microsoft], +1.250.668.aaaa,
          John_Hudson

   Regrets
   Chair
          SV_MEETING_CHAIR

   Scribe
          kuettel

Contents

     * [3]Topics
     * [4]Summary of Action Items
     __________________________________________________________

   <trackbot> Date: 10 September 2014

   Taking a few note today

   Starting with:
   [5]http://lists.w3.org/Archives/Public/public-webfonts-wg/2014A
   ug/0013.html

      [5] http://lists.w3.org/Archives/Public/public-webfonts-wg/2014Aug/0013.html

   Vlad: should all table directory conformance statements carry
   forward? Would be a good ideal to review, wouldn't want to
   automatically carry everything forward
   ... esp. as there are some that do not apply (e.g. padding
   between tables)

   <scribe> ACTION: vlad review conformance statements for WOFF
   1.0 and transfer the applicable ones over [recorded in
   [6]http://www.w3.org/2014/09/10-webfonts-minutes.html#action01]

   <trackbot> Created ACTION-138 - Review conformance statements
   for woff 1.0 and transfer the applicable ones over [on Vladimir
   Levantovsky - due 2014-09-17].

   Next "5. Compressed data format"

   Make compressing with Brotli a MUST

   John H., yes, sounds good

   Next: ""If the decompression function fails for any table, the
   WOFF file is invalid and MUST NOT be loaded."

   Yes, the file should be discarded if invalid

   Next: "Do we have the same constraints on table data
   immediately following table directory,"

   Vlad is working on updating the spec (addressed last week)

   Vlad: to recap, made sense with 1.0 to reproduce the exact
   binary, in 2.0 will ask decoder to order by OpenType
   recommendations, but order not significant otherwise

   John H: note, that the OpenType specification only captures the
   order for a set of the tables, not any possible one. We would
   just want to convey this in the spec

   John H: the specification has a different list for CFF (vs.
   TrueType). Not a big difference, but good to note

   Vlad: I updated the specification to cover this, along with
   links to the recommendations

   John H: note, that for OpenType this is a recommendation.

   Vlad: our specification will say that the decoder SHOULD follow
   the OpenType spec

   Next: "These differences will invalidate 'DSIG' table"

   Vlad: we covered this in the past, and added the recommendation
   that WOFF 2.0 remove the DSIG table
   ... Sergey had clarified that the DSIG table was used in
   showing the icon type, but that has (likely) since changed

   Jonathan Kew just joined! Welcome

   Vlad: Behdad had elaborated on how an empty DSIG table could
   work
   ... the comment from Chris was that we should be clearer on
   whether the DSIG table should be removed or not
   ... wasn't suggesting any changes that would change the file
   format, rather improving the spec

   Vlad, David: would like to keep the behavior as is, and change
   spec to MUST remove DSIG

   Vlad: accept Chris's suggested wording

   action vlad Update the DISG wording per Chris's recommendations
   (to always remove the table)

   <trackbot> Created ACTION-139 - Update the disg wording per
   chris's recommendations (to always remove the table) [on
   Vladimir Levantovsky - due 2014-09-17].

   Next: "The WOFF 2.0 encoders SHOULD also set bit 11 of the
   'flags' field of

   Vlad: yes, let's require this (MUST)

   action vlad The WOFF 2.0 encoders MUST also set bit 11 of the
   'flags' field

   <trackbot> Created ACTION-140 - The woff 2.0 encoders must also
   set bit 11 of the 'flags' field [on Vladimir Levantovsky - due
   2014-09-17].

   Next: "It is up to the encoder to produce transformed data

   Chris suggested: "The encoder MUST produce transformed data
   that is valid."

   Johnathan: spec may not be giving much value, unless the
   specification clarifies what exactly needs to be done.

   Vlad: let's make a note to revisit this later. something could
   be changed, but not sure what just yet
   ... even if the output is valid OpenType data, how would you
   test that (if a MUST). let's defer to face-to-face

   Next: "Editor's note: Do we need to add the conformance
   requirement for UA, if bounding box is not present?"

   Vlad: yes

   Next: ""a decoder should store for each glyph the corresponding
   offset in the reconstructed glyph table

   Vlad: this could become a normative statement. acept

   <scribe> ACTION: vlad add a normative statement in the loca
   table section (to firm up) [recorded in
   [7]http://www.w3.org/2014/09/10-webfonts-minutes.html#action02]

   <trackbot> Created ACTION-141 - Add a normative statement in
   the loca table section (to firm up) [on Vladimir Levantovsky -
   due 2014-09-17].

   Next: ""Editor's note: Do we need to add the conformance
   requirement for UA, if bounding box is not present?"

   Vlad: yes

   <scribe> ACTION: vlad update spec to require bounding box
   presence [recorded in
   [8]http://www.w3.org/2014/09/10-webfonts-minutes.html#action03]

   <trackbot> Created ACTION-142 - Update spec to require bounding
   box presence [on Vladimir Levantovsky - due 2014-09-17].

   Next: ""The origLength field MUST specify an adequate amount of
   space to represent the reconstructed glyf table"

   Vlad: need to define better.
   ... size of the output data could vary based on the
   optimizations applied. nominal size not there yet...
   ... ultimately it's up to the decoder. possible security
   concern.

   Jonathan: implementor could allocate the wrong amount of memory
   ... note that the size is a "guide", but may not be accurate
   ... other option, specify an exact algorithm
   ... would still be up to the implementor, but at least the
   specification would be clear

   Vlad: exact algorithm would be hard, brute force / largest data
   not ideal, achieving optimal results could be overkill

   Jonathan: could specify the simplest brute force algorithm,
   while allowing the decoder the flexibility to be more efficient

   Sergey: efficiency in space or size?

   Vlad: many unknowns, but the could be a big undertaking to get
   right (to make it very reliable memory allocation size wise)
   ... let's continue at the f2f

   Next: "6. Extended Metadata Block

   Vlad: should be a MUST
   ... in regards to single stream or separate. yes, separate
   compression stream
   ... padding, alignment. had highlighted this as a question

   <scribe> ACTION: rod check padding and alignment in reference
   implementation [recorded in
   [9]http://www.w3.org/2014/09/10-webfonts-minutes.html#action04]

   <trackbot> Error finding 'rod'. You can review and register
   nicknames at <[10]http://www.w3.org/Fonts/WG/track/users>.

     [10] http://www.w3.org/Fonts/WG/track/users%3E.

   Vlad: first should be made explicit, next four byte aligned
   data blocks

   Next: covering the bug that was uncovered, where the spec and
   reference implementation had drifted

   John H: any behavioral impact?

   Vlad: no, data is not changed, only the sequence is not
   modified.
   ... fine updating the spec to reflect this (just a cosmetic
   change)

   <scribe> ACTION: vlad update the spec in regards to the bug
   [recorded in
   [11]http://www.w3.org/2014/09/10-webfonts-minutes.html#action05
   ]

   <trackbot> Created ACTION-143 - Update the spec in regards to
   the bug [on Vladimir Levantovsky - due 2014-09-17].

Summary of Action Items

   [NEW] ACTION: rod check padding and alignment in reference
   implementation [recorded in
   [12]http://www.w3.org/2014/09/10-webfonts-minutes.html#action04
   ]
   [NEW] ACTION: vlad add a normative statement in the loca table
   section (to firm up) [recorded in
   [13]http://www.w3.org/2014/09/10-webfonts-minutes.html#action02
   ]
   [NEW] ACTION: vlad review conformance statements for WOFF 1.0
   and transfer the applicable ones over [recorded in
   [14]http://www.w3.org/2014/09/10-webfonts-minutes.html#action01
   ]
   [NEW] ACTION: vlad update spec to require bounding box presence
   [recorded in
   [15]http://www.w3.org/2014/09/10-webfonts-minutes.html#action03
   ]
   [NEW] ACTION: vlad update the spec in regards to the bug
   [recorded in
   [16]http://www.w3.org/2014/09/10-webfonts-minutes.html#action05
   ]

   [End of minutes]