Re: review process not working

[David Kastrup]

Reinhold Kainhofer <address@hidden> writes:

> And with the patch reviews and the patch countdown, at least we get
> some feedback now. A while ago it used to be luck if you posted a
> patch to the devel list and got one response eventually.

Well, that's what I experience with my syntax extensions.  I'm currently
down to eliminating some reduce/reduce conflicts because I am not happy
with the semantics of the last patch version (which received no feedback
like its predecessors), and then am likely to document and commit it.

What worked before will continue to work except for some hypothetical
borderline cases (not to be found in regtests) which you need to rewrite
into something quite more convenient.

The main benefit for the average Lilyponder will likely be that there
will actually be documentation.  The previous semantics were awkward and
undocumented.

So in my case, the review process tends to do nothing.  Could be worse,
but likely nobody except myself would know the difference.  In the case
I brought up here, the review process did a lot, but too superficially.
Of those cases, the second one is much more likely to get an ISO
certification for proper process.

I don't feel good about either.

-- 
David Kastrup