C-LARA

This week, I have been working on four things: “Simple C-LARA”, the ComputEL-7 paper, the second progress report, and an initial version of “reading histories”.

Simple C-LARA

We now have the enhancement to keep track of dependencies between project phases installed on the UniSA server, so that Simple C-LARA will only need to recompile the necessary parts when you tell it to create the multimedia text. In full C-LARA, the tick marks in the project detail view now correctly show which steps are out of date and need to be redone. This functionality seems to be very useful.

I have also added a number of smaller fixes and improvements, most notably a new pane which shows the segmented text and allow the user to edit it. I’m sure we’re not at the end of the improvement process yet. Feedback very welcome! If you haven’t already tried it out, playing with simple C-LARA is a lot of fun, you can create charming little illustrated poems and stories in just ten minutes.

If you haven’t already seen it, check out Claudia’s post describing initial experiences at the Romanian school. Really encouraging.

ComputEL-7 paper, Drehu alphabet book

Pauline has now got a second Drehu resource completed, a version of the classic fable “The Wind and the Sun”. We are in the process of substantially rewriting the paper to respond to the reviewer comments.

Second progress report

The second progress report is coming together:

• Evaluating C-LARA’s performance. Annika, Chadi, Rina and I have agreed to repeat the ALTA experiments with the current model, so we can track progress. We still need to finalise the details though.
• “ChatGPT-4 as a software engineer”. I have been discussing an intriguing idea with the AI: maybe it can help classify the logged interactions, which otherwise will be a rather laborious task? It doesn’t seem impossible.
• Initial user experiences. Claudia and Lucretia will contribute a section on the Romanian school.
• Other. I expanded the introduction considerably.

‘Reading histories’

I have been working on the idea we’ve discussed many times of including functionality to support “reading histories” which combine several texts with the same L2 into a single C-LARA document. Quite a lot of the code is now working on my laptop. I hope to have a first version installed on the server before the end of the week.

Next Zoom call

Thu Feb 1, 2023, 20:00 Adelaide (= 09.30 Iceland = 09.30 Ireland/Faroe Islands = 10.30 Europe = 11.30 Israel = 13.00 Iran = 17.30 China = 20:30 Melbourne/New Caledonia)