Pulled in all the food images directly into my repo
Previously, they were served via Google Cloud Storage, but Gridsome has native support for responsive images
Or, it sort of supports responsive images. For it to work in a realistic scenario, you have to do some weird stuff
So now it does a fancy thing where there’s a super low-res base64-inlined image on the page, then it lazy loads the real images as they scroll into view
According to Google Page Insights, performance actually got worse, but I hope that the reality is that browsing the site for real users is now faster
It’s much tidier having all the images in one place and not having to manage my own image resizing
It still involves a bit of babysitting because often my article text chose a different serving size than the USDA’s default serving size, so I do need to do some manual work for each page to make sure all the numbers match up.
Got a lot of the infrastructure in place to start auto-generating pages based on USDA info
My previous policy was to always hand-write all pages, but I’m going to see if auto-generating works.
Learned a little more GraphQL (read: went from 0 to 0.1% competent) to improve my page queries in Gridsome
Submitted a patch to Commento so that all changes to their documentation would go through a dead link checker in CI, but they haven’t responded to the MR
