Explain the logical reasoning behind local Abseil snapshot + namespace renaming

[maxgolov]

No code changes.

This PR is a supplementary doc that explains the logical reasoning / merits of the process used to create / refresh the Abseil snapshot in PR #572

Since some changes were not trivial, a separate section is added to elaborate on these changes.

Hopefully we should not employ this (semi-manual) process too often.

[codecov]

Codecov Report

Merging #797 (327c006) into main (703a2a5) will decrease coverage by 1.85%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##             main     #797      +/-   ##
==========================================
- Coverage   95.48%   93.64%   -1.84%     
==========================================
  Files         156      190      +34     
  Lines        6614     8953    +2339     
==========================================
+ Hits         6315     8383    +2068     
- Misses        299      570     +271     

Impacted Files	Coverage Δ
...pentelemetry/context/propagation/noop_propagator.h	60.00% <0.00%> (-40.00%)	⬇️
sdk/src/trace/span.cc	84.50% <0.00%> (-6.96%)	⬇️
api/test/trace/provider_test.cc	73.34% <0.00%> (-6.66%)	⬇️
...lemetry/exporters/memory/in_memory_span_exporter.h	94.29% <0.00%> (-5.71%)	⬇️
...de/opentelemetry/trace/propagation/b3_propagator.h	94.03% <0.00%> (-2.46%)	⬇️
...ude/opentelemetry/common/key_value_iterable_view.h	86.67% <0.00%> (-2.22%)	⬇️
...lude/opentelemetry/sdk/trace/samplers/always_off.h	77.78% <0.00%> (-2.22%)	⬇️
sdk/test/trace/always_off_sampler_test.cc	84.62% <0.00%> (-2.05%)	⬇️
sdk/test/trace/always_on_sampler_test.cc	89.75% <0.00%> (-1.92%)	⬇️
...i/include/opentelemetry/trace/propagation/jaeger.h	96.50% <0.00%> (-1.06%)	⬇️
... and 85 more

[maxgolov]

We seriously have to do something about codecov/project. It is:

• rarely adds much (or any) value
• noisy
• creates an impression of CI failure when, in fact, CI is completely clean.

We may need to edit the associated GitHub action to avoid triggering red marks, and only treat it as optional, always pass showing the code coverage rate changes. Maybe there's an option to configure a min threshold that trips the failure, excluding anything that is under 1%.

Also, perhaps, if the list of changed files covers only Markdown or in /docs folder, we should avoid triggering any builds at all. That'd save environment and electrons that are unnecessarily disturbed.

[lalitb]

incorrectly closed by bot.