LEMDO 2023-06-21 to 2023-06-23
to : Martin Holmes
Minutes: 310
On Wednesday worked on some thorny tickets, and managed to close a couple, one new and one old; also worked on the proliferation of included appendix material problem afflicting edition pages, and reduced the number of bad local links from over 5,000 to under 1,000 in the QME test site, so that’s some definite progress, but there are clearly more red-herring issues disguising the (presumably) relatively few genuine broken links we want to find.
On Thursday did a lot of work on the anthology link checker output, and ended up with a couple of specific things to work on, one being the perennial problem of links within entity records which themselves were imported, so the targets of those links were not processed in the normal way. That will need a policy decision. But another issue, which is being tracked as GitHub #181, is a problem whereby a nasty collision of overlapping ranges which are the subject of both collations and annotations results in the failure to insert the required collation image/link. Figuring out the basics of that just in order to raise the ticket took a little while.
On Friday, made a lot of progress with the remaining linking errors in QME. It turns out that
we have no constraints to check the accuracy of lem/@source and rdg/@wit pointers,
and they’re very easy to mistype, so there were lots of those errors. I wrote some Schematron to
catch these errors in the ODD file, built it, and used it to fix a whole pile of them including
all the current ones in QME, but had to comment it out in the end because there are many to
get through, and they should be showing in the diagnostics anyway.