Brian Dunn comments

Results 24 comments of


                                            Brian Dunn

Address already in use - connect(2) for flatware-sink (Errno::EADDRINUSE)

I'd love to see a PR that does any / all of the following. 1. deletes the socket file as a part of `flatware clear` 1. adds a `force` flag...

Address already in use - connect(2) for flatware-sink (Errno::EADDRINUSE)

Also, an `at_exit { DRb.stop_service }` in `Sink` could help prevent.

Worker always dies after change to catch CLD signal

🤔 great point. thanks for the issue. I'll look into adding a check that the worker exited early, as apposed to naturally.

Worker always dies after change to catch CLD signal

@AesSedai I've tried to address this issue in #89. Could you test that branch and see if it's working for you?

Worker always dies after change to catch CLD signal

@jdelStrother thanks for trying it out! Merged #89 and released a patch. Optimistically closing.

Introduce the duration provider pattern

@yuki24 Thanks for your work on this! Curious, have you measured timing for that giant suite with and without time balanced jobs. I added the feature because other parallel runners...

Introduce the duration provider pattern

I did not know time balancing offered such gains that's awesome. The[ original design](https://edotor.net/?engine=dot#digraph%20%7B%0A%20%20node%20%5Bcolor%3Dcornflowerblue%5D%3B%0A%20%20subgraph%20%7B%0A%20%20%20%20node%20%5Bshape%3Drecord%2C%20fontsize%3D10.0%5D%3B%0A%20%20%20%20dispatcher%20%5Blabel%3D%22dispatcher%20%7C%20%3Csub%3E%20SUB%20%7C%20%3Crep%3E%20%20REP%22%5D%3B%0A%20%20%20%20worker%20%20%20%20%20%5Blabel%3D%22worker%20%20%20%20%20%7C%20%3Csub%3E%20SUB%20%7C%20%3Creq%3E%20%20REQ%20%7C%20%3Cpush%3E%20PUSH%22%5D%3B%0A%20%20%20%20sink%20%20%20%20%20%20%20%5Blabel%3D%22sink%20%20%20%20%20%20%20%7C%20%3Cpub%3E%20PUB%20%7C%20%3Cpull%3E%20PULL%22%5D%3B%0A%20%20%7D%0A%20%20subgraph%20%7B%0A%20%20%20%20sink%3Apub%20%20%20%20%20%20%20-%3E%20dispatcher%3Asub%20%5Blabel%3Dquit%5D%3B%0A%20%20%20%20sink%3Apub%20%20%20%20%20%20%20-%3E%20worker%3Asub%20%20%20%20%20%5Blabel%3Dquit%5D%3B%0A%20%20%20%20worker%3Apush%20%20%20%20-%3E%20sink%3Apull%20%20%20%20%20%20%5Blabel%3Dresults%5D%3B%0A%20%20%20%20worker%3Areq%20%20%20%20%20-%3E%20dispatcher%3Arep%20%5Blabel%3Dready%5D%3B%0A%20%20%20%20dispatcher%3Arep%20-%3E%20worker%3Areq%20%20%20%20%20%5Blabel%3Djob%5D%3B%0A%20%20%7D%0A%7D) for flatware held a queue in a ["Ventilator"](https://zguide.zeromq.org/images/fig19.png). When a worker finished one `Job` it would...

accepts_nested_attributes_for allow partial updates

Rails achieves this by 1) presence or absence of an `id` , and 2) an `update_only` flag. https://api.rubyonrails.org/v7.1.3.4/classes/ActiveRecord/NestedAttributes/ClassMethods.html An implementation like that would keep with active record api expectations.

Issue when tests finish

Interesting. I wonder if this is a bug with printing pending results or with the way SimpleCov determines when to calculate coverage. Does your suite have 3 or 6 pending...

How to extends Link component from router file

@oedotme I wonder if you can provide an example of using the `Link` from the generated `router` file and extending that? I like the way that component takes a `Path`...