On Fri, Feb 8, 2013 at 2:45 AM, Keith OHara <k-ohara5...@oco.net> wrote:
> On Thu, 07 Feb 2013 20:56:06 -0800, Trevor Bača <trevorb...@gmail.com>
> wrote:
>
> On Thu, Feb 7, 2013 at 2:02 PM, Keith OHara <k-ohara5...@oco.net> wrote:
>>
>> Trevor Bača <trevorbaca <at> gmail.com> writes:
>>>
>>> > Did VerticalAxisGroup's default-staff-staff-spacing property stop
>>> > respecting the 'padding' attribute between 2.17.9 and 2.17.10?
>>> >
>>> > I ask because I use a custom time signature context in my scores.
>>>
>>> No, padding still works. The latest LilyPond misunderstood what you
>>> want with the zero-dimension TimeSignature.
>>>
>>
>
>> Good point. I set zero-extent time signatures so that the notes in the
>> score will be positioned exactly where they are supposed to begin when
>> proportional notation is turned on. (You can see this in the small PNGs
>> sent along with the original post.) With nonzero-extent time signatures we
>> wind up with notes being pushed slightly to the right of where they should
>> begin in order to accommodate the X-extent of time signatures that come
>> between notes.
>>
>
> Other than that there's no particular allure to the zero-extent time
>> signatures: they're just the mechanism I was best able to use to achieve
>> the precision of horizontal layout. Quite happy to solve the problem a
>> different way if some newer mechanism is recommended.
>>
>>
> Probably you should adjust your solution, because Mike's next patch
> completely removes lines of music that have only zero-size extents.
>
> If I comment out the following four overrides, the output looks correct to
> me.
>
> % \override TimeSignature #'X-extent = #'(0 . 0)
> % \override TimeSignature #'Y-extent = #'(0 . 0)
> % \override TimeSignature #'self-alignment-X = #center
> % \override VerticalAxisGroup #'staff-staff-spacing =
> % #'((basic-distance . 8) (minimum-distance . 8) (padding . 6)
> (stretchability . 0))
>
> Notice that basic-distance, etc., take hyphens where you had underscores
> (and there are no error messages for typos when setting these low-level
> options).
>
> If it is important for the bar line to be close to the following note,
> maybe something like
> \override TimeSignature #'X-extent = #'(0 . 0.1)
>
>
Hi Keith,
These are very helpful, indeed. Thank you so much for taking a look at the
collection of settings. Taking these as a starting point I find that these
adjustments ...
\override TimeSignature #'X-extent = #'(0 . 0.1) % <=== following your
trick of a small dummy distance
% \override TimeSignature #'Y-extent = #'(0 . 0)
% \override TimeSignature #'self-alignment-X = #center
... recoup the original layout.
So this is good.
I also find that ...
\override VerticalAxisGroup #'staff-staff-spacing =
#'((basic-distance . 0) (minimum-distance . 10) (padding . 6)
(stretchability . 0))
... works again with the new settings so that I'm once again able to
vertically position the time signature context exactly where I want it.
Also, when you write that "that basic-distance, etc., take hyphens where
you had underscores (and there are no error messages for typos when setting
these low-level options)" I feel like I may be losing my mind because it
looks on my end like the underscore-formatted versions of the options work
just fine (under both 2.17.9 and 2.17.11). Could this in fact be the case?
But are the hyphen-formatted versions to be preferred? Perhaps as part of
the increasing regularization of the syntax that's taking place globally?
Reflecting a bit, it seems like \override TimeSignature #'X-extent = #'(0 .
0.1) is ever so slightly a bit of a hack because we're using distance as a
type of boolean flag. I'm quite happy with whatever set of solutions emerge
and I don't mean to create extra work for anyone (certainly not Mike as he
continues to improve spacing; I love spacing improvements). Maybe an
explicit flag like \override TimeSignature #'no-X-extent = ##t might be the
clearest way to model what we're actually doing here: just ask the user to
signal her or his intentions directly to the spacing engine.
Thanks again.
Trevor.
--
Trevor Bača
trevorb...@gmail.com
_______________________________________________
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel
