- From: olivier Thereaux <ot@w3.org>
- Date: Wed, 14 Apr 2004 07:35:57 -0400 (EDT)
- To: www-style@w3.org
- Message-Id: <DA951767-8E07-11D8-BE1E-000393A63FC8@w3.org>
Dear CSS Working Group. Below is a short review of the first public working draft for css3-hyperlinks (2004-02-24). The document is globally very pleasant to read, very well presented, especially for a first WD. I will try to give this review a "QA" approach, even though most of the comments were made by simply thinking as an implementor[*] rather than just having strict QA in mind. [* for the context: QA Tools => CSS validator ] I am purposefully including comments which will be naturally addressed as the document evolves and matures, as I believe it would be an important benefit for the spec and implementors if they were addressed at an early stage of the spec development (hence the first WD review). **************** [1] The most notable aspect of this WD is its dependency on a specification that does not exist yet. The abstract mentions rapidly that this is a CSS3 module, without explaining how this module would interact with others. Nor is there a definition of what a CSS3 module is would make early implementation (attempts) possible, but that does not seem possible at that stage, and will not be until there is a master document clearly defining conformance for CSS3 in general and integration of modules in this big picture. [2] The Dependencies section (1.) states that this module depends on CSS2.1 but does not explain in which way. [3] The Dependencies section (1.) does not make any mention of dependencies on CSS3-syntax. Is that on purpose? [4] The Dependencies section (1.) mentions references, without explaining how these differ from the list of references given in "References" at the end of the document. This is not necessarily a bad idea, but an explanation would make it more useful and less confusing [5][edit] Requirements and Principles (2.) would probably benefit from more context and a smoother beginning. Add context. [6] Requirements and Principles (2.) is actually, mostly, about giving the scope of the module. Marking this up more clearly (in scope - NOT in scope) would be beneficial [7] Requirements and Principles (2.) bullet 1,3 and 8 are very similar in their idea. Regrouping them in some way could make the text clearer, especially if, as the text states, it is supposed to be "summarizing" the principles. [8] Requirements and Principles (2.) has a lot of principles that seem very generic. Is there a plan to coordinate all CSS3 documents to achieve consistency on how high level principles are addressed? [9][edit] Missing a transition between Requirements and Principles (2.) and Target Properties (3.) [10][edit] in (3.3#example), "she" is a very contemporary form of Politically correct talk that may well be obsolete before CSS3 is. Consider neutral form for such sentences maybe? [11] Conformance (4.) is obviously still at an early stage, but it's great to already have a mention of a future test suite. Would be good, for next draft maybe(?), to have a planned timeline and modalities for the test deliverables. [12] Conformance (4.) has no information on conformance per se. What class of products does CSS3-hyperlink apply to? How does a product or document conform to CSS3-hyperlink? The spec so far is very declarative "this property is suchandsuch", which is hardly implementable if there is no explanation on what course to follow to implement this specification. See related comment [1]. [13][edit] Default style sheet, Index and Property index give a good example and outline, but may benefit from being clearly marked as appendices, and given an introduction. **************** Thanks for your good and very promising work on CSS3, Regards, -- olivier Thereaux
Received on Thursday, 15 April 2004 04:20:43 UTC