Disagree with recent DOC change

[David Kastrup]

I just see in the most recent commit:

 
+Even if you are not familiar with git or are not able to compile
+LilyPond you can still help to narrow down the cause of a regression
+simply by downloading the binaries of different LilyPond versions
+and testing them for the regression.  Knowing which version of
+LilyPond first exhibited the regression is helpful to a developer
+as it shortens the @code{git bisect} procedure described above.

I don't think that this is useful advice since binaries often are
incompatible with Scheme/Lilypond trees of other versions, and since
many regressions are not introduced in the binaries.

The effort to do this outside of a working git controlled tree is not in
any sane relation to the result which will not be reliable most of the
time.

I think that the time of non-developer/non-gituser contributors can be
spent better.

-- 
David Kastrup