W3C home > Mailing lists > Public > www-svg@w3.org > December 2005

SVGT 1.2: Please extend the Last Call period

From: Maciej Stachowiak <mjs@apple.com>
Date: Wed, 28 Dec 2005 19:42:38 -0700
Message-Id: <1F1D37FA-B523-4C46-9909-07355307FBAF@apple.com>
To: www-svg@w3c.org


Now that many have dug deeper into the spec, using the bare minimum  
Last Call period of 3 weeks, in a block that heavily overlaps a major  
holiday season in places where many interested parties live, seems  
inappropriate.

Specifically:

* The specification is extremely large. The PDF version is 352 pages.  
It is aggressive to expect to proofread even a novel of that length  
in 3 weeks, let alone a technical specification.

* Many issues have been identified, some of the quite serious, by  
people who have yet to get all the way through the spec. This  
strongly implies that more issues are lurking in the remainder. I  
have seen serious problems in nearly every section of the spec that I  
have read closely. These include not just "spec lawyer" problems or  
language fine-tuning, but contradictions with other w3c specs,  
contradictions within the spec itself, serious security flaws and  
language that is meaningless as written.

* The spec references many other w3c specs, and in some cases  
textually includes subsets of them. This requires close reading of  
the other specs in addition to the 352 page SVG 1.2 Tiny  
specification itself. This cannot be written off as a minor issue  
because there are in fact many conflicts. In other cases, the spec  
calls for behavior that is randomly different from other XML  
applications such as XHTML, requiring still further specs (and  
implementations, where there is de facto standard behavior) to be  
examined.

* The previous Last Call period generated a flood of comments,  
finding many serious problems. In some cases reviewers gave up part way.

* Many comments from the previous Last Call have not been addressed  
in a satisfactory manner; in some cases the WG did not even make the  
changes they said they were making in response to comments.

For all these reasons, the time alotted is hardly adequate. I don't  
accept the argument that "the industry is pushing to have this spec  
released". No one is well-served by the release of a broken  
specification.

Therefore I request:

* Please extend the last call period to a reasonable amount of time.  
At least three additional weeks seems like the bare minimum, ideally  
considerably more.

* Please have at least on additional Last Call before proceeding to  
CR to ensure that issues raised during this Last Call are addressed  
in a satisfactory manner.

Regards,
Maciej
Received on Thursday, 29 December 2005 03:06:56 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:32 GMT