normalize term for instrumentations #539
Conversation
Fixes open-telemetry#476 A single term for instrumentation libraries would be beneficial, as it reduces ambiguation. Removing synonyms "integration" and "adapter". Integration doesn't imply some form of instrumentation, while adapter could easily overlap with middleware or other concepts in existing frameworks that will be instrumented.
toumorokoshi
requested review from
arminru,
bogdandrutu,
c24t,
carlosalberto,
iredelmeier,
jmacd,
reyang,
SergeyKanzhelev,
tedsuo,
tigrannajaryan and
yurishkuro
as code owners
Using the full "instrumentation libraries" form where possible. Fixing capitalization of OpenTelemetry. Adding reference to "Instrumenting Libraries". Adding link to definition in overview
|
@open-telemetry/specs-approvers please review/approve this PR ;) |
|
@toumorokoshi Please resolve the (already resolved) comments so new reviewers can get a more condensed view ;) |
fixing typo Co-authored-by: Giovanni Liva <[email protected]>
|
Mentioned in the commits, but just as an FYI: I couldn't find a clear source to determine how to make "instrumentation" plural, so I renamed all referenced to "instrumentation libraries" instead. |
|
@SergeyKanzhelev I've addressed your concern around the goal of OpenTelemetry. @mtwo would you mind giving the PR another review? I rewrote the first paragraph significantly. After reading your comments I realized that the first paragraph could articulate several details more clearly. |
Co-authored-by: Morgan McLean <[email protected]>
|
this seems to be mostly an alignment. not a new work. And we have enough sign offs. Merging |
* Fix instrumentation library synonym Address issue not resolved in #539: https://github.com/open-telemetry/opentelemetry-specification/pull/539/files#r426546780 * Link to definition in api.md * Link to definition from glossary.md Co-authored-by: Armin Ruech <[email protected]>
|
Going to add some notes as a public system of record is good. As of this post, the following languages have the following conventions: instrumentation
not instrumentation
|
|
JS maintainer here. Should we be renaming existing libraries to use the term "instrumentation", or just using that term in new libraries and keeping existing naming for packages already published? /cc @open-telemetry/javascript-maintainers @open-telemetry/go-maintainers @open-telemetry/ruby-maintainers @open-telemetry/python-maintainers @open-telemetry/technical-committee |
If there is going to be a push to do this, the time is best before the GA release. I'm interested what the @open-telemetry/technical-committee 's guidance here is. |
As @MrAlias mentioned, the best time to do a complete rename is before GA. I'll be filing tickets in all repos to start the discussion, but I think that's a SIG / TC decision. I'll see if I can bring it up in the next spec meeting as well. As a data point, I'll be making a hard push to do a complete rename on the Python side, as well as adhering to the standard from now on. |
* Rename plugin directory to instrumentation As per open-telemetry/opentelemetry-specification#539 * Update package paths * Fix tracer names in othttp instrumentation
* Rename plugins director to instrumentation As per open-telemetry/opentelemetry-specification#539 * Update package names * Plugin -> instrumentation
* Rename plugin directory to instrumentation As per open-telemetry/opentelemetry-specification#539 * Update package paths * Fix tracer names in othttp instrumentation
Fixes #476
A single term for instrumentation libraries would be beneficial, as
it reduces ambiguation. Removing synonyms "integration" and "adapter".
Integration doesn't imply some form of instrumentation, while adapter
could easily overlap with middleware or other concepts in existing
frameworks that will be instrumented.