Tired of these yet?
So I ran into a weird problem with my V3 site. I decided I had too many pages with no introductory text and it was proving to be awkward to use. Because of the way I had set up the Hugo site, I also couldn’t use the built-in “_index.md” feature, as it would wipe out the images I was trying to display. So I came up with some bonkers, way over-complicated feature to read in a text file (actually Markdown) so that it would show text, but it didn’t work properly and I reached out to the community for help and one of the very nice and helpful gurus had a “WTF ARE YOU DOING?!” moment (he said it much nicer than that)…
…and then I asked myself the same question. What was I doing?? Why the hell had I gone down that road?
And, also, while I was at it, I was having to sort images into folders, which was still not at my original “I just want a bucket” requirement.
While not back to the drawing board, it was back to rethinking the process:
- Take a bunch of (properly EXIF’d and IPTC’d and geotagged) images.
- Run them through a script to build info files for them, including title, description, EXIF data.
- Build a folder set based on geographic data: country, region, city.
- Build the site.
I’d mostly done these steps, though I hadn’t really paid much attention to the geographic organization – I’d done that with folders to a poor degree. Now it would be automatic, the folders being built as the structure required (it did require some cleanup of geographic data, as even Google can’t seem to provide data consistently), and that allowed me to provide the “_index.md” files – intro text for countries and regions.
But then I realized I needed to tweak nearly 1,000 images (and those are the ones I’ve already processed, there’s another 1,000 or so in the wings). And then I realized that I hadn’t properly tagged them (taxonomy, vs. geo), which meant yet another batch of runs.
However, after all of that, I think – finally – I’m in a place where I can focus on the image, not the app.
Whether the app continues to work properly after I double the number of images remains to be seen, however … it already takes over 10 minutes to build (including the image processing, resizing, cropping, etc.) and I don’t know what the upper stable limit is.