HCMC Blogs

I now have a list of sections along with their titles and page-references from JJ, for the purpose of constructing a modern TOC. It turns out that these sections, since they're going to have to display in multiple pages rather than page-by-page, and since they begin and end in the middles of pages (in other words, they don't correspond with existing <div> elements), can't be handled in the way I had previously anticipated. So I've rethought things, and I think this is what we'll do:

I already have code to display a single page. I need to modify that (wrap it, basically_ so I can retrieve a page-range as well. The TOC will be constructed as a <list type="abstractTOC"> inside the <profileDesc>/<abstract> in the header. I've already added this @type value to the schema. Each item in the list will consist of a <ref> element pointing at a page-range (probably using pbIndex rather than the sometimes-flaky page numbers of the original text, although we'll probably use the original page-numbers for display).

Waiting for approval from the team for this before I can proceed.

Spent all day following these instructions:

http://web.uvic.ca:8080/~ucomcoop/docs/Building_and_Migrating_Sites/Migrating_From_2008_Template

In most cases, the simplest approach is to copy stuff over from an existing site (I used French). Sometimes it helps to screenshot the working site and copy settings from it (because you can't have two simultaneous sessions on Cascade in different locations, even with different browsers). I have virtually everything working except for the News and Events page, which is strangely borked, and the unit logo, which I think needs to be done by an official graphics person.

Leaving early.

I've been struggling mightily with the issue of how to convert BC dates to Gregorian equivalents. I think it turns out to be senselessly complicated. W3C (ISO 8601) dates, which aspire to be computable, include a zero year, which is equivalent to 1BC, but XSLT date-arithmetic stumbles over this with an error; meanwhile, various accounts of the proleptic Gregorian have it including a zero year or not including one. It's difficult to know what a conversion to proleptic Gregorian could mean in such a context. Nothing at that period seems to be accurate enough for the Julian/Gregorian difference (which is only a couple of days at that point) to be worth worrying about.

So what I've decided is this: for any negative Julian dates, I just add the little mouseover saying "This is a Julian date." I don't provide any suggestion of what its Gregorian equivalent might be.

This (and some appropriate error-trapping) means that TRIU2 is working again.

Figured this out after seeing a sample XEP config for Oxygen on the Oxygen forum (thanks Adrian).

In the process of updating Ubuntu, a bunch of apps got removed because of repo problems. I reinstalled all the ones I noticed, but it looks like the "xep" script file that runs the app was deleted. That would have presented Oxygen with a problem, presumably, and it seems to have responded by setting the XEP config to this:

/media/Data/mholmes/apps/RenderX/XEP/lib/xep.jar -fo ${fo} -${method} ${out}

In other words, it was pointing directly at the jar file, not at the now-missing script. I've restored the xep startup file from a backup, and pointed Oxygen at it, and now everything works.

The XEP error message didn't help figure this out.

Met with SS, TC and DC about the Versioning Machine version 5. Created ten tickets on SourceForge for the tasks that need to be completed before we can do a beta release. Also planned a poster proposal for TEI Conference.

One in PDF rendering, where an extra period was appearing when imprint contained only date and nothing else; and one in XHTML rendering, where all monograph titles were being double-tagged in bibliography output. The latter was only recently apparent because I'd added handling for genuinely-nested titles, setting italics back to normal for the inner one.

Encoded the French intro, and built a volume for JT and HT to proof.

I'm logging this here because I haven't found a solution to it yet; I'll add one if/when I do.

My installation of Oxygen which has been working fine for a couple of years, currently at 15.2. I use XEP a lot. Recently I updated the OS (from Ubuntu 12.04 to 14.04, both 64-bit), which may or may not be the cause of the problem, which is this:

When I run a transformation scenario which does XML-XSLT-XSLFO-PDF using XEP, I get this error:

invalid file (bad magic number): Exec format error

However, when I run the same process on the command line, starting with the FO file, it works fine. The XEP license file and xep.xml file are in place, and haven't been changed. Even if I run the process pointing explicitly at the /bin/java in the Oxygen JRE folder, it still works fine, so it's not a problem with Oxygen's Java being different from my system default (which is OpenJDK-7, as it has been for a long time).

Spent the morning learning the best ways to create nice app-like interfaces for shell scripts on the three platforms.

Windows:

Used "Bat To Exe Converter" to create an exe: http://www.f2ko.de/programs.php?lang=en&pid=b2e

Mac:

Used Platypus. Found a number of things I had to work around to get it going properly. Tested on a couple of other Macs.

Linux:

Settled on a solution where the first time the script is run (or whenever it's run without a folder parameter), a .desktop file is generated, pointing to the path of the script (which can't be relative in a .desktop file). This seems to work OK. Parameters are passed to the script automatically.

This is now ready for testing on any platform.