- From: Henri Sivonen <hsivonen@iki.fi>
- Date: Wed, 13 Jun 2007 08:21:27 +0300
- To: Ben 'Cerbera' Millard <cerbera@projectcerbera.com>
- Cc: "HTMLWG" <public-html@w3.org>
On Jun 13, 2007, at 01:18, Ben 'Cerbera' Millard wrote: > On Thursday, May 31, 2007 6:42 AM, Karl Dubost wrote: >> for your tests, give them (using parts of [test metadata]) > > My understanding of cases files is that they should be minimal, > containing as little code and content as possible. Has either > approach been proven useful in practise? Small parsing test cases are good if one tries to step through them in a debugger. > On Thursday, May 31, 2007 9:53 AM, Henri Sivonen wrote: >> [...] file named conforming.txt that lists on LF-terminated lines >> the names (one name per line) of conforming test cases in that >> directory. > > I like the idea of doing this in as simple a format as possible. > Would CRLF-terminated lines be acceptable? Newline normalisation is > easy when processing files; there's even a spec for it [2]. But > authoring LF-only lines is troublesome in Windows (which I use). Yeah, LF, CRLF or CR would work. > A test might conform to HTML4; or HTML5; or both. Should each of > these be represented? I'm pretty sure a document cannot conform to both HTML4 and HTML5 at the same time due to different doctype requirements. -- Henri Sivonen hsivonen@iki.fi http://hsivonen.iki.fi/
Received on Wednesday, 13 June 2007 05:20:49 UTC