Map Of London

Worked with Tye on making breadcrumb links more sensible (always link to a listType=subcategory view if there are any subcategories).

I've been working on the blog posts. Done so far:

• Three complete blog posts, validating against our schemas and rendering appropriately.
• A new doc type, mdtBlogPost.
• More sophisticated rendering for the mdtList of blog post types.
• Detailed instructions for creating a blog post.

Still to do: an aggregation page which shows them concatenated, or extracts from them concatenated.

Changed "right" to "margin-right". Don't know why it was originally "right".

Added a basic "Send feedback" feature to the site.

Using the consistency checker, I tracked down and fixed about 40 bad internal links, and sent questions to team members on a few more.

Fixed a JS bug reported by PS, made a change to the home page, and looked at another issue that's been raised: it's been suggested that we hide the list of resources at the bottom of a document, but I think that's a bad idea because it means anyone printing the document, or accessing it without JS, will not be able to see the sources cited. In some cases, such as CAMP3, the single source is cited only in a footnote, so looks odd in a print view, but that's because the footnote is not printed at the bottom of the text. I think that's what we should be printing.

We've also identified an Oxygen bug working with PERS1 and BIBL1, where using Find can crash the app. GN is working on a bug report and testing for workarounds. It doesn't affect my machine, but all other Linux machines seem to be affected.

The "this [thing] is mentioned in the following documents" component of the popup reference box was listing all documents, irrespective of publication status. It's now restricted to TEI[descendant::revisionDesc[@status=('final', 'published', 'stub', 'generated')]].

Names heading popup references consisted only of forenames and surnames; this resulted in oddities for monarchs, whose names were too short without roleName and genName components. That's now fixed.

My hacked-together XSLT for @rendition/<rendition> from a few weeks ago was only handling labels and notes, but we obviously need a better approach, so I've done the initial implementation. As it stands:

• Any <rendition> elements in the header are turned into classes in a local stylesheet.
• Any elements which are already being produced with classes have rendition-based classes added to the class attribute.
• Any elements which are output without classes are processing @* (hopefully), and this triggers a template matching @rendition which creates a class attribute.

I've purposely allowed for the use of the rendition system in born-digital as well as primary source documents, mainly because there are overlapping templates so it's hard to distinguish which context you're in half the time, but I don't want or expect rendition to be used on born-digitals; if it is, then something is probably not right with the encoding.

This gives you a copy-pastable <bibl> tag with all content appropriately marked up.