Open rmarx opened 1 year ago
Personally, I don't think I'd have the ability to pass through enough context to my logging library to capture the trigger. But I think I need a bit more time to work on the implementation.
@LPardue any update on your opinion on this? I still think this would be useful in general, even if quiche can't pass it :) and i still like the values I proposed above. Are you opposed to a PR for this?
I can live with it if you want to add it. But I'd like a better description of how the "trigger" field is a part of event data; see #429
I am wondering if the new H3PriorityUpdated event (see #312) needs some specified
trigger
field values.Conceptually, you can get the necessary context from other events surrounding this event (e.g., if you get a H3FrameParsed with a HEADERS containing a
priority
field preceding this, that's a clear indicator that's the trigger). However, there are cases where this is less clear (e.g., default initialization to a value different from the spec, local override due to configuration or other logic, merging of client and server-sent priorities for the same stream, ...)We don't NEED to define any triggers here (
trigger
is always defined on event data, just not necessarily specced in qlog), but I'm wondering if some default values would be useful here @LPardue?e.g.,