Let's pass the doctype01.dat tests again, please
Created attachment 62244 [details] Patch
Comment on attachment 62244 [details] Patch Isn't our dom2string.js directly from upstream? Does the upstream version have this bug?
Yeah, I've been discussing that with gsnedders in #whatwg. There isn't a more awesome version anywhere. I think he might grab this diff and apply it upstream. He also mentioned wanting to rev the whole runner.html.
Eric: which upstream version? As far as I know the upstream versions of it are just what myself and Simon have up on html5.org, which are in varying states of out-of-dateness. Both have this bug. I was intending on updating the copy I have to fix a number of issues in the runner soon, I'll try and include the fix for this in it. However, the patch is wrong; it should be: if (current.publicId !== null || current.systemId !== null)
Will fix before landing.
Hum... That doesn't actually work. I think that's because of we're bad at handling NULL public and system identifiers. I'll change it in a future patch.
Comment on attachment 62244 [details] Patch Clearing flags on attachment: 62244 Committed r63871: <http://trac.webkit.org/changeset/63871>
All reviewed patches have been landed. Closing bug.