- From: fantasai <fantasai.lists@inkedblade.net>
- Date: Wed, 27 Jun 2007 17:46:10 -0400
- To: Arron Eicholz <Arron.Eicholz@microsoft.com>
- CC: "public-css-testsuite@w3.org" <public-css-testsuite@w3.org>
Arron Eicholz wrote: >> Fantasai wrote: > >> >> Looks good to me. What exactly are the tags you are using? I was >> thinking it would be clearer to use tokens rather than letters, >> e.g. 'ahem' rather than 'a' for the Ahem font requirement. >> > > We are currently using the letters 'a' for ahem and 'i' for interaction, > etc... We are not strongly attached to them though. > > I agree I think tokens would be better since they are more self describing. > If we go with tokens, there is always the possibility of someone misspelling > them (though hopefully the code review would catch that). The scripts can validate them, too. I already need to write one that can parse these, to add in those HTML comments about e.g. Ahem requirements that your team requested. So, here are some possible flag names ahem (a) scroll (c) -- for tests that will only pass on continuous media invalid (f) -- replaces failure test type image (g) interact (i) namespace (n) -- (the harness script shouldn't convert these to HTML) svg (v) paged (p) -- for tests that will only pass on paged media dom (o) Any suggestions for improvement? :) > Also there is the issue of how the tokens should be separated this wasn't an > issue with letters. Semi-colon, comma, other? I personally like semi-colons > in this case. I would just use whitespace, just like class attribute values. ~fantasai
Received on Wednesday, 27 June 2007 21:46:18 UTC