Re: horizontal spacing regression

[Carl Sorensen]

On 1/14/11 4:20 AM, "Joe Neeman" <address@hidden> wrote:

> On Fri, Jan 14, 2011 at 11:57 AM, Carl Sorensen <address@hidden> wrote:
>> On 1/13/11 10:25 PM, "Keith OHara" <address@hidden> wrote:
>>>> Two accidentals remain tucked, but this case does not:
>>>>   ceses!4...  feses! r16
>> 
>> So it seems that making this change affects a desired behavior (as described
>> in a regtest).  If, in your opinion, the output is better overall with the
>> new behavior, we should probably do the following:
>> 
>> 1) Have a discussion on -devel, with images shown before and after
>> 2) Change something about the description of spacing-horizontal-skyline.ly
>> <http://spacing-horizontal-skyline.ly>
>> to indicate under which conditions tucking is *not* allowed.
>> 
>> Otherwise, we'll need to figure out some other method of approaching this.
> 
> I think you'll find that the regression was caused by ee0488, in which case
> the extra-spacing-height override is exactly the thing to do if you want to
> restore the old behaviour. It's really just a question of how much padding to
> add; I don't think it's worth a long discussion.

I agree in general, but the question is "do we *want* to restore the old
behavior?"  We have a regtest that demonstrates the new behavior.  When we
try to restore the old behavior in the case at hand, we also break one
instance in the regtest.

I guess that if we can find a value of extra-spacing-height that will
restore the old behavior in the case at hand, and not break the regtest,
it's the best of both worlds.

Thanks,

Carl