Here's something about memory in Rails background jobs that I sort of encountered over the years but only recently fully figured out. I'll jump straight to the point.

The rails query cache causes ruby to be less effective at garbage collection. This is an acceptable tradeoff in web processes which do not run for a long time . . .

Read More

In a discussion recently someone was asking for feedback on how to go about expanding a system to cover more use cases without clumsily breaking the abstraction. I had these thoughts:

Maybe questions to ask are:

• How easy is this to change later?
• How easy is this for others to use?
• Are there any fundamental . . .

Read More

Just a quick post here about something that doesn't seem to be covered much elsewhere.

If you have a query like this (this one generated by pg_search):

SELECT COUNT(*)
FROM "things"
INNER JOIN
 (SELECT "things"."id" AS pg_search_id,
 (ts_rank((
 to_tsvector('simple',  . . .

Read More

I recently did a big cleanup of how fonts are used on a project. Here's a quick rundown of some things I learned along the way. Each building block is pretty straightforward and easy to find documentation for, but overall best practices, and explanations for how it all comes together, are lacking, so hopefully this guide will fix that . . .

Read More

If you want to run rake tasks from within an already-running rails instance (such as when using ruby-clock) without forking and making a new rails process (system 'rake foo:bar'), it's very possible, but there are a few things to keep in mind.

First, load the jobs once when the app is being initialized (I think this is . . .

Read More

The best approach for creating a new column that has a default has changed since Postgres 11. In the case where we want the existing rows to have a different value from future created rows (e.g. a boolean which signifies if a user should see the onboarding tour), the best approach has also improved significantly in a way that perhaps . . .

Read More