Log in

HCMC Journal

Servitengasse 2025-06-23 to 2025-06-27

to : Martin Holmes
Minutes: 420

Over the weekend started work on the map page and JSON. On Tuesday, did a bit of a dive into the problem of who appears on the Original Residents list versus on the Collected Residents list, and sent a query to HT, because it seems rather inconsistent on the current site. Then carried on working on the GeoJSON, during which I discovered that most of the buildings appear to have coordinate sets which don’t follow the right-hand rule, so I’ve been correcting those, which is quite finicky and tedious. I suppose I’ll have to retro-fit those changes into the SQL db just in case.

On Wednesday, I finished fixing all the geometries to make them follow the right-hand rule, and then when I had valid GeoJSON and was able to look at it, I discovered that two of the buildings had different (less complex) shapes than those that show up on the website. I discovered that the website is using some (presumably directly-edited?) GeoJSON that does not match what is in the database. I was able to retrieve the new, better shapes for the two buildings (8 and 10), and fix them for right-handedness, so our TEI placeography is up to date. However, this means that basically we are done with the original database. We already know it had errors, and I’ve been fixing those as we find them, but now it turns out that other editing was done externally to the database, so from this point, the TEI is the canonical source, and we no longer bother to fix errors in the database. It’s now just a historical relic. This will also free us to fix some oddities and inconsistencies that arise from the actual structure of the database, since the TEI doesn’t have the same constraints.

On Thursday, working from home initially, integrated HTML validation into the build and started fixing all the things it found. Then worked on the map JSON and got the buildings to display correctly, also tweaking them so that they show in the right order. Then I started work on creating entries for people, and got a little way into it. For this, I’ll adopt an interim approach to dividing them into residents and collected people, pending decisions from HT, and then work on the relocation of the site menu into the map menu bar.

On Friday, I got the map working as planned, which means we could publish a new edition any time. However, there are many niggles that need to be addressed, which I’ll outline in another post.