Re: What's with the spacing code?

[Phil Holmes]

"David Kastrup" <address@hidden> wrote in message 
news:address@hidden
> "Phil Holmes" <address@hidden> writes:
>
> [Please don't write anything important below your signature, as mail
> clients will cut this away on reply].

Apologies.  I have to cut and paste my sig to the bottom and I already had 
something ready to paste, so forgot.

> If you think this is all fine, take out the markup from the example and
> get a really _tight_ fit in contrast.

Wasn't saying it was fine - just that it's not a regression between 13.34 
and 13.35 - it's a change, but compared to 12.3, 13.34 was too tight.  Using 
the test file you provided, 12.3 took 7 pages.  .31 and .34 (and probably 
others - I don't have a full set) took 5.5 pages and leave no room for 
markup.  .35 takes 10.5 pages and leaves too much room for markup.

> The current spacing is not a matter of "too tight" or "too loose".  It
> is a matter of "too unpredictable".

Agreed.  There are currently 2 critical bugs outstanding against spacing 
issues.

> And if you really want to see some hot action, just write
>
> \score { c }
>
> as often as you want.  Regardless of how much of those you put into the
> file, the outcome will be just a single page.

This produces the same output in 2.12.3.   I'll add it as a bug which is not 
a regression.

> All this is rather erratic.

Agreed.  I'll add the initial issue you raised to the tracker as a critical, 
even if it's the same as the other spacing problems, since I guess it'll 
serve as an excellent test pattern.

--
Phil Holmes
Bug Squad