Posts Of The Week 2021-04-15
Lots of small things today.
History
“Those who cannot remember the past are condemned to repeat it”. I love reading about how software came to be the way they are today.
Before we can talk about where generics are going, we first have to talk about where they are, and how they got there.
https://cr.openjdk.java.net/~briangoetz/erasure.html
Go Errors
Error handling is still jacked in Go 1.16. That is, the formatting change is still not present. Why is this a problem? There are two use cases for errors. Error as values, which can be inspected programatically, and error printing, which is not meant for programmatic consumption.
In Go 1.13, an API for error type inspection was introduced. This provides a way to communicate different types of errors. Why would you want to do this? A common example to distinguish an error that can be trivially handled with a retry, such as updating a network resource using a compare-and-set operation, vs a non-retryable error (authentication failed). However, 1.13 API does not help in error printing. I actually spent a bit of time trying to coerce the API into exposing the call stack information, but failed to do so without also exposing unwanted error types. You can see my failed attempt here.
At the present, call stack support is available in golang.org/x/xerrors.
| feature | 1.13 | xerrors |
|---|---|---|
| call stack | no | yes |
| retain the call stack when annotating an error | no | yes |
| retain the error type when annotating an error | yes | yes |
| hide/change the error type when annotating an error | hard to do, so “no” | yes |
The last feature is particularly important. It allows one to expose the call stack to aid in debugging without encouraging programmers to code against implementation details.
Some programmers will code against implementation details no matter what you do.
A new model for content creators?
I followed Bleve for a while. The lead of Bleve, Marty Schoch, had started a new group called Bluge Labs. By the means of this project, I learned that you can create a sponsorship model on Github. It is an interesting model and I’m curious to see how it turns out. Sponsorship often distorts the relationship between two parties. I hope it works out for Mr. Schoch.