Daniel Chao
Daniel Chao
I'm thinking the schema version should probably just be an extra field on each message.
This actually introduces a new problem--links to modules of a package dependency is now broken. I'm going to hold off on this for now.
I think maybe a better solution is to let users configure what _library_ they want their nullness annotations to come from (this is maybe what @holzensp is talking about too)....
> Pkl’s @NonNull annotation should be deprecated as tools won’t understand it. I don't think this needs to be deprecated; tools like IntelliJ let you configure nullability annotations. The intention...
> It doesn’t need to be, but users who care about nullness annotations should be strongly encouraged to use one of the widely supported libraries instead of configuring Pkl’s proprietary...
> PS: The blue colored line numbers 7 | and 1 | are difficult to read on black background. Hm, it doesn't look too bad for me. But, hard to...
I am not sure we want to do this. This makes it hard to upgrade the collections library, and I'm not sure how much of a performance gain we get...
Sorry for dropping the thread here. Does this PR make any PE optimizations? Or, are there any PE optimizations that we know about that we can follow up with in...
I'd rather not vendor this if it just helps save a wrapper class. Do you mind going into a little more detail? What optimizations do you see that we can...
In most cases, we vendor because we want to eliminate a dependency (e.g. gson, guava, xerxes). In other cases, we vendor because we needed to make modifications to an existing...