Ramon
Ramon
For a description of exemplars, see: https://www.youtube.com/watch?v=U72b4Nl0Ftw
Any design for retroactively exporting interesting spans will I think require nodes to retain spans somewhere long enough for the sampling decision to be made. Luckily, it's all best-effort so...
@bogdandrutu agree that the underlying model should be protobuf-based. I would argue against textproto as the format (not widely known) and instead for using the yaml -> json -> proto...
I think this is really critical to making the HTTP and gRPC measures useful. Without this, especially the HTTP measures have too little context to really make them useful for...
We could make the tag map mutable e.g. ```go tag.FromContext(ctx).Mutate(tag.Insert(key, val), tag.Upsert(key2, val2)) ``` I think this is already the case in Java, right? You can replace the Thread-local tags...
I see. Well, in that case I propose that we make the tag map mutable in all languages. The only alternative I can think of would be to store the...
So something like: ```go ochttp.MutateTags(ctx, tag.Upsert(...), tag.Upsert(...)) ``` or ```go ocgrpc.MutateTags(ctx, tag.Upsert(...), tag.Upsert(...)) ```
The concurrency problems with the second approach are no worse than those that come with updating Span attributes (for example).
/cc @dinooliva
Agree that we should drop the idea of mutating tags in general - let's forget about that. @rakyll regarding the second approach (just adding tags to the top-level Record), I...