Comments on Addendum to Mobile Web Best Practices

I had actually included these comments in the "rationale" field of the questionnaire, but François pointed out to me that they should preferably be sent to the mailing list directly. So here they are.

E.Casais
-----------

Rationale: 
I have one general remark, and a few observations of detail.

1. GENERAL REMARK

The W3C has now at least the MWBP, its addendum, mobileOK tests, an emerging MWABP, accessibility guidelines, possibly internationalization rules. All these 
documents embody best practices, but adopt different formats to present them; they do not actually target the same classes of terminals, but without specifying clearly what they are -- with the exception of two documents 
explicitly based on the DDC. 

Section 1.2 of the Addendum constitutes a representative example: the document purports to "fill gaps left by limitations of automated tests and complete the 
set of Best Practices" -- but does not defines precisely these gaps. Furthermore, one should be vary of "completing best practices", i.e. extending or refining them, in an addendum rather than in the main document itself. There is finally the issue of redundancy and consistency between the various best practices, which may become increasingly tedious to manage as the set of guidelines grows.

As a developer, I would appreciate to have a brief, overall roadmap document that would guide the developer among the set of BP and that should include the following elements:

a) A list of the best practices with
a.1) a definition of the delivery context or terminal environment each one targets;
a.2) the relationship of each BP to the other BP (subset-superset / complementary / specialization).

b) A suggested roadmap to apply the best practices:
b.1) identification of best practices from the target devices;
b.2) ordering / prioritization of the application of BP according to the elements in a.2).

c) A complete list of references to the relevant BP and guideline documents..


2. DETAILED OBSERVATIONS

General
It is unclear whether the order of bullet points in the "evaluation procedures" is prescriptive or not. 
The Addendum provides for "evaluation procedures". However, several sections indicate "ensure that..." (a prescription) rather than "check if..." or  "determine whether...": 3.1, 3.5, 3.28, 3.37.

Section 2.1
URIs instead of URIS. 
If POWDER is not finalized, why recommend its utilization?

Section 3.1  
"Make sure to indicate which access are being used..." -- meaning "which access keys are being used..."? 
"found on all pages within a Web space", "across all pages" -- technically not tractable for dynamically generated pages. Needs some rephrasing.

Section 3.9
The second bullet has an unfortunate formulation. What is probably meant is "In the English language, the Fogg test can give an indication of complexity, a level of roughly 7 or 8 being ideal." The sentence on other languages makes one doubt whether Fogg applies or not to them, whether Fogg applies but not the levels indicated, or whether any such kind of test applies or not.

Section 3.14
Replace "funcionality" with "functionality".
Replace "Disable cookies supported" with "Disable cookies". 
What is the difference between point 3 and point 4?

Section 4.22
Replace 4.22 with 3.22.
Eliminate the redundant "bad example:" in bullets.

Section 3.30
"Check if all style properties are used in the page" -- to be nuanced. There are general site style sheets defining properties that are not necessarily all used in every page. Perhaps "check the percentage of style properties used in the page" instead.

Section 3.33
Replace "Sumbit" with "Submit".

Section 3.36
The evaluation procedure is a typical example for the issue raised above under "General".

Received on Thursday, 5 March 2009 15:40:33 UTC