issue_comments: 719809780

This data as json

html_url issue_url id node_id user created_at updated_at author_association body reactions issue performed_via_github_app
https://github.com/simonw/datasette/issues/1072#issuecomment-719809780 https://api.github.com/repos/simonw/datasette/issues/1072 719809780 MDEyOklzc3VlQ29tbWVudDcxOTgwOTc4MA== 9599 2020-10-30T21:32:28Z 2020-10-30T21:32:28Z OWNER

Here's an alternative that would definitely work and would be a lot simpler, at the cost of a fair amount of RAM:

  1. Before rendering the template, load ALL of the most-recent-versions of the templates that are stored in the DB. Use those to populate a DictLoader.
  2. Render the template.

This does mean loading template bodies that we won't use. Provided an instance has less than 100 templates I imagine this will work just fine.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
 733499930