LEMDO work 2022-10-24 to 2022-10-28
to : Martin Holmes
Minutes: 410
Did a little more work on the overlap-detection stuff, but it’s not going as smoothly as I’d hoped, and I might switch it from XQuery to XSLT. I also closed one ticket by adding a processing-instruction, and fixed another display bug which had not made it as far as a ticket.
Project meeting on Wednesday resulted in one new ticket and some ad-hoc schema changes; I’ve also been investigating the possibility of a sparse anthology build, which would build lemdo-dev first but only with the subset of pages required for the anthology, making it a fast and targetted build that would not break due to irrelevant bugs in pages not included. That seems tricky but doable.
On Thursday we finally figured out why lab machines could not build the lemdo.odd file into schemas: the bleeding-edge plugin needs to be installed. This suggests that the Oxygen framework has changed its layout or folder structure and so the same paths will not work. We must connect with Oxygen about this.
Friday was a concentrated LEMDO day, addressing the very thorny issue of the broken site interface when the same text links to multiple annotations. It turned out to be the result of changes to the HTML structure which had been made due to requests for CSS changes, which had put the JS out of sync with the HTML structure, combined with an overly-complicated JS function which attempted to hide annotations when the same text was clicked on a second time, something that is unintuitive anyway. I simplified the HTML structure of the popup, rewrote the JS, and made a lot of CSS changes to deal with fallout, and to fix a couple of other bugs including an unnecessary scrollbar. Finally I added CSS for double-underlining text which is associated with multiple annotations, although I’m not sure whether that’s a keeper or not because it’s a bit ugly. All around, a win, though, and a nasty annoyance removed.