JT has decided that the regular volume is going to be ready before the film volume, so we're switching the volume numbers. Changed the ids, filenames and header tags appropriately, then deleted and re-uploaded all the files. This also served as a check on the stability of eXist, after its flakiness yesterday: no problems at all this morning. Also did an rsync --delete to Rutabaga; over the years a lot of alternative filenames and versions have accumulated, so this is a good time to clear out the cruft.
Did a few pages; the next section will require some XSLT work to handle the drama quotations, which I don't think we've encountered before.
I was updating the ScanCan collection this morning, and ran into some serious errors, which eventually resulted in all the documents being unavailable, and any attempt to upload or reindex giving an error. The site is still running, but no documents show up. A bit of research reveals a message from WM on the gmane eXist list to the effect that a restart usually fixes it, so I've asked for Tomcat to be restarted. A subsequent message from Wolfgang says this:
I finally managed to find out what caused the infamous "document id and proxy id differ" exceptions, which have been reported by various users during the past months. The error was the result of a conflicting access to the same document by two threads, one of them updating nodes, the other just reading them. The reading thread modified properties of the document object without having acquired a lock. So far I thought the "document id and proxy id differ" exception should not be fatal, but I have to correct myself now: the error may easily lead to corruptions in all indexes, which makes a reindex necessary.
Following the restart, I'll start reindexing all the collections. Fingers crossed on this one. It's the first indication of a potentially serious issue with the rather old eXist version that's running all the older projects. With luck, it'll happen rarely and we won't be forced to upgrade eXist and port all the projects to a new one...
The Durbach article (surely the last for vol 19) is now in, so I've set it up, done the metadata and the first paragraph, and also the blessedly short bibliography.
Another review for Vol 19 has come in, and is now marked up and on the proofing site.
The Blackwell article is now complete, and posted for proofing.
Finished the body of Blackwell; only the biblio left to do. In the process, added a couple of fixes and a new feature to Transformer.
Moving a bit more quickly through this now, with the help of Transformer, but I need to finish the next version of Transformer to get a bit more flexibility. I'm going to take a look at that today.
This one is slow, partly because of the huge number of listed movie titles with glosses (I created some Transformer sequences to automate handling those) and also because the end notes are part of the document body, so you can't click to jump to them in OOo; I have to track down to find them, and then back to find my place. I've also experimented with auto-markup of (for instance) the film title which is the topic of the document.
Did the first few paras, including lengthy notes.