http://webstandards.or.kr/2006/11/10/have-your-say-about-the-future-of-html/ The standard evangelism group for universial access and trust of web in Korea Fri, 18 May 2012 04:03:51 +0000 http://wordpress.org/?v=2.0.11 http://webstandards.or.kr/2006/11/10/have-your-say-about-the-future-of-html/#comment-27626 Sun, 11 Mar 2007 11:06:03 +0000 http://webstandards.or.kr/2006/11/10/have-your-say-about-the-future-of-html/#comment-27626 I would like to see this spec come with a well-advertised, set of stress test pages published stressing every conceivable feature of the specification, within practical reason (so that it is as difficult as possible for a user agent to pass all of them that does not implement the specification correctly). The spec would state that a correct implementation of this specification passes each of these stress cases without any errors. Each of these tests would have a well-defined expected result, making it as easy as possible to spot compliance failures (perhaps many of them could have a top half and a bottom half, implemented very differently, but should present themselves identically or clearly indicate what variations are permitted). This would encourage better compliance by helping spot problems with various implementations, and encourage them to be compliant by advertising their compliance failures to their end users. I realize that developing a really good set of stress pages is more challenging than some might realize, but the more we can do the merrier, and would work to address the most common complaints expressed herein. I would like to see this spec come with a well-advertised, set of stress test pages published stressing every conceivable feature of the specification, within practical reason (so that it is as difficult as possible for a user agent to pass all of them that does not implement the specification correctly). The spec would state that a correct implementation of this specification passes each of these stress cases without any errors. Each of these tests would have a well-defined expected result, making it as easy as possible to spot compliance failures (perhaps many of them could have a top half and a bottom half, implemented very differently, but should present themselves identically or clearly indicate what variations are permitted). This would encourage better compliance by helping spot problems with various implementations, and encourage them to be compliant by advertising their compliance failures to their end users.
I realize that developing a really good set of stress pages is more challenging than some might realize, but the more we can do the merrier, and would work to address the most common complaints expressed herein.

]]>