Semconv 1.24.0 #46
Semconv 1.24.0 #46
Conversation
jack-berg
force-pushed
the
semconv-1.24.0
branch
from
683099c to
2755790
Compare
jack-berg
force-pushed
the
semconv-1.24.0
branch
from
2755790 to
9831977
Compare
|
#45 is merged so this is ready for review. |
jack-berg
force-pushed
the
semconv-1.24.0
branch
from
9831977 to
07ceccc
Compare
| * </ul> | ||
| */ | ||
| public static final AttributeKey<String> EVENT_DOMAIN = stringKey("event.domain"); |
There was a problem hiding this comment.
@lmolkova This is an example of an experimental attribute that was deleted rather retained but marked as deprecated.
Based on what you said, in the future this will not be the case. And we would expect to keep this attribute around semi-permanently marked as deprecated. Is that correct?
There was a problem hiding this comment.
I'm not trying to answer for @lmolkova, and I'm not addressing the general question being asked....but in this specific case I do think it makes sense to remove it, because (experimental) Events are starting to gel and we don't want people using this.
There was a problem hiding this comment.
sorry, I missed it.
Yes, starting from 1.24.0 we'll have a check that prevents attributes (and metrics) from being deleted - open-telemetry/semantic-conventions#764
It would apply to experimental and stable semconv.
I think we'll need to do the same for events, updated open-telemetry/build-tools#292 so we don't forget
There was a problem hiding this comment.
Thanks, I guess that's fine, as long as they're marked deprecated and can be removed in the subsequent release. I'd hate for us to have constraints that cause long-term maintenance burden due to changes in experimental semconv (and yes, I realize this isn't the place to drag this out, heh 😅 ).
There was a problem hiding this comment.
Would love to hear your thoughts on removing-experimental-deprecated on open-telemetry/semantic-conventions#740. It'd be great to have a common policy for all language SIGs.
...bating/src/main/java/io/opentelemetry/semconv/incubating/AspnetcoreIncubatingAttributes.java
Outdated
Show resolved
Hide resolved
...ncubating/src/main/java/io/opentelemetry/semconv/incubating/SignalrIncubatingAttributes.java
Outdated
Show resolved
Hide resolved
| * Deprecated, use {@code http.request.header.content-length} instead. | ||
| * | ||
| * @deprecated Deprecated, use `http.request.header.content-length` instead. |
There was a problem hiding this comment.
This isn't a showstopper, but it does really a little redundant.
There was a problem hiding this comment.
Good point, but we don't control that text. It comes directly from the semantic conventions YAML: https://github.com/open-telemetry/semantic-conventions/blob/cafda7127683b7f667e27cdbd3220510b6f998c9/model/registry/deprecated/http.yaml#L34
| @@ -25,10 +25,6 @@ public final class SystemIncubatingAttributes { | |||
| /** The device identifier */ | |||
| public static final AttributeKey<String> SYSTEM_DEVICE = stringKey("system.device"); | |||
|
|
|||
| /** The disk operation direction */ | |||
| public static final AttributeKey<String> SYSTEM_DISK_DIRECTION = | |||
There was a problem hiding this comment.
This might be another set of examples of deleted (not deprecated) semconv?
There was a problem hiding this comment.
Its all generated code. To some extent you can rely on the build passing to help give you confidence in the output, since the build will run code generation and detect any difference. The one thing it doesn't check is if there are are files which are part of the commit, but which are not generated anymore. For example, this commit excludes the |
Depends on #45.