Re: Strange beaming error

[David Kastrup]

Urs Liska <address@hidden> writes:

> Before testing further I assume the current issue is due to *not*
> using the latest state (which isn't in a release yet).

Why are there different states?  Did the original plan fail on paper or
did it never reach paper?  Or was it not correctly implemented?

> But of course I'll test with the material and with all iterations of
> the code once I have it.

"Iterations of the code" makes me queasy.  It suggests that there is no
plan but instead one pokes the code until it happens to work most of the
time.  Then it cannot be documented, maintained, or changed because
nobody really knows the reason that it works.

I am aware that large amounts of code come into being in that manner,
partly because of resource limitations.

And there is some justification for it in the "reaping" phase of a
product, where the main issue is selling it while investing the least
amount of work (Sibelius anyone?).

LilyPond is not due for the reaper yet.

-- 
David Kastrup