HCMC Blogs

Stew, James and I attended a meeting with Colin and Kim to discuss what we can do to ease some of the recovery issues that arose when the disk failed, among other associated things.

We'll be requesting a few new filesystems which will be devoted to the code and data required for a live site as well as one that will be devoted to archival use. Repositories of images, videos etc. that are not called upon in the normal course of events will go in the archive filesystem. This will mean that we can pick and choose which FS gets priority in the event of failure. Obviously, we could have lived without EEBO or the 700GB of TIF files when recovering the data last week. If that stuff had been on a separate FS it would have been faster/easier to restore sites.

The ice has been broken on the new Colonial Correspondence (CC) website!

After ruminating on the design and creating one discarded mockup, I came up with an (I think) attractive search bar that will help define the appearance of the site and future design elements.

I think that overall design of the site will be completed in the next three or so days.

Still to do is the banner, an attractive " historical snippets" column, and an "about this project" box, and an subtle footer.

After that the styling will begin on the text and form elements on the site, which must wait for the core look to be completed before they can be styled.

There were about a thousand instances in the despatch files where underlining code had not been converted appropriately, either because it wasn't in a format my converter was expecting, or because it ought to be a different kind of markup (for instance, where a ship's name was underlined, and we'd actually like it to be marked up as a ship's name so that we can find ships' names as a class of item in the database). I kept coming across these as I continued to work through the footnotes, so I've spent a couple of hours tidying them up. There are still many instances if ships' names to convert, though; those will have to be done in batches as we discover what the names actually are.

Link: http://eigentaste.berkeley.edu/user/index.php

An engaging procrastination tool from Ken Goldberg, an engineering professor at the University of California at Berkeley, and Tavi Nathanson, a Berkeley computer-science student. Jester--based on a computer algorithm that Berkeley patented in 2003--is a website that claims to personalize jokes for your particular sense of humour. You begin by rating several jokes on a scale ranging from "less funny" to "more funny," and the algorithm uses your responses to deliver your kind of humour.

Not quite a procrastination tool, because when I stopped smiling, I started to wonder if we could use a similar tool to find out, for example, our students' favorite medium for learning.

Caveat Lector: if your sense of humour doesn't run to "how many feminists/men does it take to change a lightbulb" (one of the calibrating jokes) or the word sh*t, you'll probably feel better if you skip this one.

Yes, according to a number of experts. Possibly, say pedagogical mavens who worry that PowerPoint undermines active learning. Definitely, yawns anyone who's ever sat through an interminable slideware read-over, disguised as a lecture or presentation.

Read more »

On late duty.

SM has finished capturing 16 videos now, and has burned DVD archives from most of them. While she burns the remaining ones, she's working on transcribing the Shippey video. Greg stripped off the audio track, and I worked on it a bit to bring up the quality. It had a loud continuous hum which had to be removed, and the volume was pitifully low. After bringing it up a bit, I tried time-stretching it so it would run at typing speed, to make the transcription go more easily, but the results were pretty poor; squeaky artifacts are introduced as soon as you get above about a 30% increase in length. This might work better with higher-quality audio, or with a better set of tools (such as Sonar 7, which I've used at home to stretch multi-track audio data with some success), but in this case it didn't help much.

SM has finished about a quarter of the transcription, which is in plain text. When it's done, I'll use Transformer to convert it to a suitable format, probably XHTML based on what we're using for Vic Van Video, leaving gaps for the timings, and she'll go back through and add all the timings. Then we'll see what we can do about a Web version of the video, as a pilot.

Got back into the grind of checking, linking and updating the footnotes for BC 58. There are 506 notes in the file at this point, and I have 324 left to do. It's a long process, but it's going a lot faster now that the search is quicker. Some notes mention CO volume 59, but the documents they point to describe themselves as being from CO 60; wrote to Jim about that. Possibly it's just a typo.

Got back into the grind of checking, linking and updating the footnotes for BC 58. There are 506 notes in the file at this point, and I have 324 left to do. It's a long process, but it's going a lot faster now that the search is quicker. Some notes mention CO volume 59, but the documents they point to describe themselves as being from CO 60; wrote to Jim about that. Possibly it's just a typo.

In the database, deleted all the records in the topics table that had no instances pointing to them (either never had any or the topic had been collapsed into another topic, e.g. ChHx into Chinese History).

Checked all the ids for the about to be deleted topics to ensure no instances pointing to them in the articles table. Found only one instance of a pointer to id 148 (originally coded as "X" by author), which I deleted from the field in the articles record.

Updated topics.txt, and sent JL and email saying if he wanted to collapse any more topics, I'm pretty confident we could do it without problem, though we are no longer able to use the text files to rebuild the db content if necessary, as they contain pointers to non-existant topic id's, though a bunch of global search and replaces would solve that.