Friday 13 November 2026
I started Friday, as I seem to start most days now, editing the Workbench note from the day before and preparing it for publication. It’s becoming quite an established routine, though the process is still fairly iterative and invariably ends with me making plenty of manual changes to the text. I’m gradually getting a better sense of how I want these pieces to read, though, and the exercise of reconstructing the previous day’s work is proving useful in its own right
After that I returned to Richard Craven’s publishing work. I’d arranged to see him at three, so I wanted to remind myself where I’d left things and get some outstanding tasks moving before he arrived and that led back to the Extraction Workbench, particularly the reconciliation of the editorial records behind Richard’s novels. There are a considerable number of proposed changes in the system, and some of the earlier material needs checking against records that have already been incorporated.
The latest exercise covered 1,914 source records, identifying 639 candidates for administrative review, along with 362 relationships requiring further attention and 229 possible gaps in the links between entities and books. The subsequent database checks and backend tests passed, leaving me with a much clearer picture of the remaining editorial work.
An unexpected launch
By the time he arrived things were mostly reconciled and we spent a couple of hours going through the new website which I’d given him for testing a few weeks ago. I hadn’t expected us to launch it that afternoon, but he was sufficiently happy with what we’d built that he was keen to get it live.
We were able to make some edits to the live site, and discussed the editorial and marketing operation we have coming up now the site it live.
One minor problem emerged with the administrator-only editorial interface I’d developed to make the more complex page content easier to edit. The intention was that Richard could make changes on the website, see those changes reflected in the pages, so that I could then send the final edits back to the Extraction Workbench. The interface was recording proposed changes correctly and producing candidates for export, but wasn’t updating the displayed content as I’d intended. That needs fixing, although it also prompted me to reconsider the export process itself.
Manually transferring changes between the website and the Workbench is rather cumbersome, particularly when both systems already handle structured information. I’m now considering building an API that would allow the Workbench to retrieve the editorial changes directly from the live site, while retaining the review and approval process.
We finished the meeting with a pint at my local. I try to see Richard every week or so and these meetings are useful markers in a project where much of the development happens independently. They give us an opportunity to review progress, agree on priorities and make sure we’re both happy with the general direction. Plus, I get to talk about literature and philosophy which is always interesting.
An inconvenient interruption
After the meeting and after dinner, I returned to an application I’d been preparing for an Executive Advisor role in AI and Digital Transformation. I’d been working on the CV and supporting material and noticed nicheclever.co.uk went down, which wasn’t especially convenient. The outage lasted about an hour, during which I spent time investigating the IONOS hosting environment and working out how to access the server logs through SSH and SFTP. I also built a reusable skill for retrieving those logs, so that if something similar happens again I should be able to investigate it rather more quickly.
I didn’t establish a definitive cause from the information available at the time, but the website came back and I was able to continue. The application materials were prepared, although I hadn’t submitted them.
For reasons I won’t go into right now, I also made a small adjustment to the custom instructions used by the Interactive Professional Portfolio.
A few thoughts at the end of the day
I think the most satisfying part of Friday was that the website launch happened rather naturally. I’d gone into the meeting expecting to review the work, make some decisions and probably come away with another list of things to do. Instead, we decided it was ready enough to put online.
I’m also finding that this daily Workbench writing is becoming more interesting as an exercise in its own right. The entries aren’t simply records of what I’ve done. Editing them makes me reconsider the work, sometimes correct my own account of it, and occasionally notice connections I hadn’t quite appreciated at the time.
The process of developing the articles is itself becoming more systematic. I’m gradually defining the style through successive drafts and manual revisions, and I’ve now updated the automated instructions to take those published edits into account. I’m still finding the right voice for these pieces, but I think they’re becoming more recognisably mine.
Export and share