EMLS

I'm presuming I've completed this, because the code is working on all the current browsers...

James C. on the TEI list suggested that clicking again on a note anchor should hide the note that was shown; look at implementing that. Also, the code uses offsetTop to position the note; that's probably a bit too loose, so we need to implement the solution from the Devonshire project, using the findPos() function at the end of this file.

I've gradually evolved a policy for handling XHTML and its mime type. We've determined that there's no reliable way to get good results if you serve XHTML as application/xml+xhtml, so unfortunately the policy for the moment is to serve them as text/html using the xhtml11_compat serializer. See the following posts for more background:

• This post
• This post
• This post

I need to check that EMLS is doing that for all pages, and check that the search works OK on all supported browsers.

The link to the blog was pointing at the old location instead of the current blog URL.

The search page was failing to show results in Safari. Running a search would result in "DOM exception 7", which is "NO_MODIFICATION_ALLOWED_ERR", whenever I the JavaScript tried to set the innerHTML of an element.

This flummoxed me for ages, until I finally figured out that it's caused by the response header of the page. If the page is server as application/xhtml+xml then Safari presumably considers innerHTML inappropriate, since the page is not HTML. If the page is served as text/html, then it works. This is logical, if rather infuriating. For now, I've set the search page to serve as text/html using the xhtml11_compat serializer. I'll now go and investigate ways to use proper DOM methods rather than innerHTML, which is a non-standard hack in any case (although it's supported by all browsers).

The EMLS search capability doesn't work in Safari -- there's no response at all. I also noticed that the logo is overlapping the text in IE7, but I'll try to fix that. Logo position is now fixed.

Just posting minutes spent looking at what is completed, and discussing what should be done.

Ray requested that a link to grab the entire xml corpus of the archive be added.

Ray has requested PDF output of each article and issue. Page numbering should be disregarded, as they are not present in the original.

Karin's old stylesheet is still on the site, and is now available as an option in the browser View menu.

Implemented these changes requested by Ray:

1. Remove the View column, in favour of...
2. The title of the article should click through to the XHTML version.
3. Give the XML and Text links a class, so they can be styled small and icon-like.
4. The Vol/Issue column can simply read Issue.