-
Notifications
You must be signed in to change notification settings - Fork 244
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Experiment to switch to nanoserde
instead of serde
#860
base: main
Are you sure you want to change the base?
Conversation
SingleModule(PathBuf), | ||
MultiModule(BTreeMap<String, PathBuf>), | ||
SingleModule(String), | ||
MultiModule(HashMap<String, String>), |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Check if using a HashMap here doesn't impact buildtime determinism.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
That's a good point. I think it's likely best to just stick to BTreeMap.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nanoserde
doesn't support BTreeMap
unfortunately, that's why I changed it.
but yes logically it should be a BTreeMap
here. though the builds are rather non-deterministic anyway so probably quite fine. but should comment it
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks like it would be pretty easy to add BTreeMap to nanoserde if we wanted to, just saying. But yeah we can stick to hashmap while we evaluate.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nanoserde
has a closed issue on this here: not-fl3/nanoserde#7
Apparently the owner of the library prefers to use a proxy in this case
[...] instead of adding all possible containers into the library. "
Might be worthwhile to just open a PR over there if the additional buildtime determinism is worth it.
I think I just need to clean up some clean up some code here wrt to the file IO handling and remove some unwraps and we can get this in! |
Part of #858