Re: Double-underline markup

[Urs Liska]

Am 19. Oktober 2019 18:54:39 MESZ schrieb Thomas Morley <address@hidden>:
>Am Sa., 19. Okt. 2019 um 14:45 Uhr schrieb David Kastrup <address@hidden>:
>>
>> Thomas Morley <address@hidden> writes:
>>
>> > Am Sa., 19. Okt. 2019 um 13:35 Uhr schrieb David Kastrup
><address@hidden>:
>> >>
>> >> Thomas Morley <address@hidden> writes:
>> >
>> >> > Iiuc, you recommend to fix \underline to make it work with most
>simple
>> >> > input like:
>> >> >
>> >> > \markup {
>> >> >   \override #'(offset . 12) \underline
>> >> >   \override #'(offset . 10) \underline
>> >> >   \override #'(offset . 8) \underline
>> >> >   \override #'(offset . 6) \underline
>> >> >   \override #'(offset . 4) \underline
>> >> >   "underlined"
>> >> > }
>> >> >
>> >> > I'll have a look.
>> >>
>> >> No, to have it work with most simple input like
>> >>
>> >> \markup
>> >>   \underline
>> >>   \underline
>> >>   \underline
>> >>   \underline
>> >>   \underline
>> >>   "underlined"
>> >
>> > Ok, understood. I'll give it a try...
>>
>> I mean, I might well be too naive about this.  If the underline is
>> occuring in a fixed position with regard to the baseline, there are
>only
>> a few obvious avenues to have multiple underlines work:
>>
>> a) change the baseline.  That's not really acceptable when mixing
>> underlined and non-underlined text
>> b) change a property (akin to offset but probably unique to
>\underline
>> to avoid unexpected interactions) for the sake of additional
>> underline calls.  That would result in the _innermost_ \underline
>call
>> ending up lowest.
>> c) somehow affect bounding box/outline in a manner that can be
>> interpreted for moving the whole next underline to a different
>position
>> while retaining the baseline.  That's sort of the handwavy "do magic"
>> option that may or may not be workable at all.
>>
>> --
>> David Kastrup
>
>I first tried to tackle the increasing widths of subsequent calls of
>\underline and observed a so far not mentioned problem:
>\underline increases the width of the returned stencil, so that
>word-space adds its value to the _increased_ borders. This is better
>visible with increased thickness:
>
>\markup
>  \column  {
>    \line
>      \override #'(thickness . 10)
>      \override #'(offset . 7)
>        { This is a \underline underlining test }
>    \line { This is a \underline underlining test }
>  }
>
>See attached image (the experimental code used there is not mature yet
>and thus not posted here, p.e. appropriate y-offset is not coded)
>
>The spacing of words looks uneven with default underline.
>With the experimental code word-space is not affected, but the line
>extends into this space.
>
>Admittedly this is all more obvious with (very) thick lines, which is
>likely a rare use-case.
>
>Anyway, should I accept the uneven spacing or the lines sticking into
>the space, provided by word-space?
>
>I'd very much appreciate opinions
>
>

I'd definitely prefer the "ornament" not changing the "substance". I.e. I would 
let the line protrude into the word space.

Urs

>
>Thanks,
>  Harm

-- 
Diese Nachricht wurde von meinem Android-Gerät mit K-9 Mail gesendet.