Introduce a ReadWriteSpan interface and pass it to the onStart of the SpanProcessor

[jkwatson]

Resolves #1569

[codecov]

Codecov Report

Merging #1574 into master will decrease coverage by 0.11%.
The diff coverage is 83.33%.

@@             Coverage Diff              @@
##             master    #1574      +/-   ##
============================================
- Coverage     86.85%   86.73%   -0.12%     
  Complexity     1392     1392              
============================================
  Files           162      162              
  Lines          5369     5369              
  Branches        513      513              
============================================
- Hits           4663     4657       -6     
- Misses          523      527       +4     
- Partials        183      185       +2     

Impacted Files	Coverage Δ	Complexity Δ
...io/opentelemetry/sdk/trace/MultiSpanProcessor.java	100.00% <ø> (ø)	17.00 <0.00> (ø)
...ntelemetry/sdk/trace/RecordEventsReadableSpan.java	81.74% <ø> (ø)	76.00 <0.00> (ø)
...telemetry/sdk/trace/export/BatchSpanProcessor.java	94.63% <0.00%> (-1.35%)	8.00 <0.00> (ø)
...elemetry/sdk/trace/export/SimpleSpanProcessor.java	97.36% <ø> (ø)	10.00 <0.00> (ø)
...ions/trace/export/DisruptorAsyncSpanProcessor.java	86.00% <ø> (ø)	10.00 <0.00> (ø)
...k/extensions/trace/export/DisruptorEventQueue.java	83.51% <90.00%> (ø)	12.00 <0.00> (ø)
.../io/opentelemetry/sdk/trace/NoopSpanProcessor.java	100.00% <100.00%> (ø)	8.00 <1.00> (ø)
...try/sdk/metrics/aggregator/LongMinMaxSumCount.java	96.07% <0.00%> (-3.93%)	6.00% <0.00%> (ø%)
...y/sdk/metrics/aggregator/DoubleMinMaxSumCount.java	96.07% <0.00%> (-3.93%)	6.00% <0.00%> (ø%)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update ca85d8f...f23a997. Read the comment docs.

[Oberon00]

Note that @bogdandrutu was against having such an interface because it might make people cast the ReadableSpan in OnEnd to it. See open-telemetry/opentelemetry-specification#669 (comment)

[jkwatson]

Note that @bogdandrutu was against having such an interface because it might make people cast the ReadableSpan in OnEnd to it. See open-telemetry/opentelemetry-specification#669 (comment)

If they know the SDK codebase, they can cast directly to the RERS implementation as it is today. I don't see that this really changes much, aside from actually matching the current spec. :)

[Oberon00]

Both PRs match the spec, I intentionally wrote it so that two parameters are also OK. That said, personally I also think that a single ReadWriteSpan interface is more clear.

[jkwatson]

Both PRs match the spec, I intentionally wrote it so that two parameters are also OK. That said, personally I also think that a single ReadWriteSpan interface is more clear.

Not sure what other PR you are referencing. Did I miss a PR somewhere?

[Oberon00]

I mean both the solution on this PR and the solution with two parameters proposed in the description of #1569 (that I also plugged in my comment). I assume that's what you referred to when you said "actually matching the current spec".

[jkwatson]

I mean both the solution on this PR and the solution with two parameters proposed in the description of #1569 (that I also plugged in my comment). I assume that's what you referred to when you said "actually matching the current spec".

Oh sorry. I was comparing with the current main branch, not vs another option! Very sorry for the confusion. I do think that either option would meet the spec..this just seemed like the simplest way to go.

[anuraaga]

Nice!