Currently in ActiveJob, `wait` in `retry_on` accepts a proc, but it is only handed the execution count. I think access to the error can also be useful when calculating the next wait time. If you agree, give it a like! github.com/rails/rails/...
Ad for “it’s 10pm do you know where you children are?”
Ad for “it’s 10pm do you know where you children are?”, but with “children” overwritten by “threads”
“It’s 10pm, do you know where your children are?”
I added this to my “when good threads go bad” post
Does anyone else remember these ads?! May have only been a USA thing
Yea I didn’t want to `raise` any suspicions with such a long gap between posts, but sometimes in life you’ve just got to `handle_interrupt`s 😜
Finally got it released, just before the new year 🥲
It’s late, and you get alerts that web requests are failing. Page loads hang endlessly. The server isn’t responding, requests are queueing up. What do you do?!
"When good threads go bad" is a deep dive into stuck threads, how to deal with them, and what to avoid.
jpcamara.com/2025/12/30/w...
🙇♂️ 🙇♂️ 🙇♂️
Awesome! Can you say which one??
My @sfruby.com talk on real-time collaboration with AnyCable, YJS and Rails is up! I felt really happy with how it came out - let me know what you think because I can't watch myself talk 😂
youtu.be/J68QOBLEItY
TIL if you want to return the duration in seconds, returning it as an Integer is a _hard_ requirement
There is a _specific_ check that the class is Integer. If you accidentally return a float, Sidekiq ignores it and reverts to its built-in delay policy. Be careful!
In Sidekiq, you can add a special method to your job class called `sidekiq_retry_in`. This method allows you to dynamically adjust retry behavior. For instance, based on the exception raised you can return a different duration in seconds. /1
It's actually happening! Gonna super busy for the next 139 days. 😄
@sfruby.com Day 3 was the perfect community break!
Started with a Lucasfilm museum walk (thank you, @inazarova.bsky.social), then joined the Golden Gate Bridge bike tour with @kanejamison.com, @jpcamara.com, @jeremysmith.co, @adarsh.ruby.social.ap.brid.gy and so many amazing Rubyists.
Had an incredible time at @sfruby.com
Fantastic talks, amazing people, and great vibes across the board
So much to dig into and unpack around tooling, AI, development patterns and our collective ownership in the future of Rails
Community day was such a great finish
I'm not sure @ankane.bsky.social is human - maybe some kind of open source being from the future 😂
I said to a colleague - "we'll see, but it's ankane, so it's entirely possible it'll get done soon". Less than an hour later, it was updated 🤯
Also very curious! I was big into warp for a while, but a) they literally still don't support full terminal history and cap at 20k lines and b) feels like they're getting further and further from prioritizing the actual terminal part
i'm back on iterm2 again, but i'd love a better tool
If you are anywhere near SF Ruby Conf and haven't bought a ticket yet, the time is now! This conf is gonna be 🔥🔥🔥
ok nice to know you are still human 😂😂😂!
Amazing, thanks for putting a name to it @byroot.bsky.social ! I'm gonna dig into this more
"Today I ran Brussels half marathon and it was 21.1km of pure joy!"
I have never heard a marathon described as "pure joy" 😂 It must have truly been something special!
Thanks for the reference! I'm gonna check this out more. Zeitwerk is something I think would be super interesting to read through - I bet there's some unique stuff you've had to do in there
4) "When good threads go bad" coming soon!
Dan shows how jobs are executed in the `process` method, which has this hefty section utilizing `handle_interrupt`. He does a great job describing how it works, and my next article will dig deeper into if you're interested!
11/11 🧵
If you read my article on ruby thread interrupts and bitmasks, this will be very familiar! jpcamara.com/2025/10/22/b...
I asked koichi if signals were the inspiration, and he confirmed it. It was fun to see the connection and what can be an inspiration for central language apis!
10/11 🧵
3) The design of Ruby thread interrupts seem heavily influenced by linux signals
When referring to the linux signal docs, I noticed a startling resemblance - it included `raise` and `kill` functions, a list of pending signals, and a way to block signals using bitmasks with pthread_sigmask
9/11 🧵
`compare_by_identity` sets the hash to compare keys by their reference instead of their value. Not day-to-day useful but a fascinating option, and utilized nicely in Sidekiq
8/11 🧵
You can change the default return value of a hash by using `default=` on the hash. docs.ruby-lang.org/en/3.4/Hash..... Now instead of nil, you'll return whatever `default` you provided! Very handy.
7/11 🧵
`handle_signal` can do what it wants! No reentrancy requirements! No special requirements at all in fact - we're back in the regular flow of our program! Super clever.
2) Two weird hash methods big Hash doesn't want you to know...
6/11 🧵
Meanwhile, the main loop for sidekiq is always waiting on the io reader, calling a blocking method, `wait_readable` in the while loop. Signal received -> write the signal -> wait_readable returns truthy -> while loop runs and calls `handle_signal` with the signal read
5/11 🧵
The full reasons for that are explained in docs.ruby-lang.org/en/3.4/signa..., and have to do with a concept called "reentrancy".
But mike bypasses all of that by using using IO.pipe. The io writer is safe to write to in the trap, so that's all it does, it writes the signal.
4/11 🧵
