Mariage

Mariage is my pilot project for the notion of static site-building, and I've started on that work today. I've reconfigured part of the repo to provide a location for developing the code, and written rendering for all the AJAX fragments; I'm now starting on the main documents, beginning with the simplest, articles.

My approach is to take advantage of the HTML5 data- attributes to preserve as much of the original TEI info as possible (source element name, attribute values, etc.), so that I'm not trying to predict what bits I will and won't need for the purposes of styling and interactivity. It's quite straightforward and elegant to do this, and rather a relief to get away from the heavily contingent style of rendering that I've ended up getting used to on projects that pre-date HTML5. No major gotchas so far.

We've decided that while CC goes on editing the texts, I will:

• Write some pilot XSLT to generate key fragments (Ajax responses as mini files).
• Write basic XSLT to generate the main documents sans website chrome.
• Write basic CSS for document display.
• Create XSLT to generate website chrome, designed in advance for desktop and mobile.
• Get a Jenkins VM set up for HCMC project static-website builds, with Mariage as the pilot.

Then we'll go on to consider the issues of redesign and the static search engine problem.

Met with CC to discuss recent contact from GS, who's still promising to send us an XML version of his EM French lexicon, enriched from some of our own texts, which we will be able to use as one path for modernization of our texts for search purposes. While we were at it, we fixed a few dozen errors in our encoding revealed by his work on our files.

Per CC, we've changed the default font to Georgia (although on one Mac, in FF, it has trouble with italicized long-s characters), set the default style to white, and removed the background graphic from it. This gives a plainer, more readable site.

This long-outstanding task is still waiting to be done. I worked through all the documents except the Varin, and was able to automate some of it by using regexes. There are some situations in which a <ref> and a correction coincide, where the consequences for rendering would be difficult to figure out, so in those cases I've left a note in place for the moment.

Met with CC, SA and some folks from the library to discuss a possible grant application, details of which will be fleshed out in due course. Did some research first and follow-up afterwards; meeting on Friday by which time I need to have a basic list of bullet-points to discuss.

Meeting and discussion with CC. Fixed a few minor display bugs with Le Blanc, by widening page margins etc and making footnote numbers smaller.

CC reported a bug, where trying to upload le_blanc.xml to eXist failed with a "networking error". It was indeed consistently so; had to delete the file and re-upload it before it would work. No idea why; couldn't find any obvious reason for it. Watch for this happening again.

Reviewed the budget and mobilization plan, and reported to CC.

Met with CC to review the grant application and the budget.