comments on CVS from 2002-07-14 15:58 (long)

[Han-Wen Nienhuys]

address@hidden writes:
> 
> I'm baack!

coooool!

Some quick replies. 

> 
>   . Section 3.3, fourth example
> 
>     The key change from two sharps to three flats doesn't look OK
>     (I check this with the dvi output at 600dpi): The two flats are
>     too close together.

yeah, something should be done about the padding in general, right
now, each acc is represented by a single box, flats & naturals should
have two and three respectively. Added some more padding in between.

>   . Section 3.8, first example
> 
>     The arpeggio line doesn't look good.  Its wiggles are too big.

Too wide or too tall?

>     I don't know whether the appearance of the beam knees is a bug
>     (since at one place the stem sits on the wrong side of the note)
>     or whether some fine tuning is missing; I'm just reporting here
>     that they look wierd.

Opps. They sure look strange, yes, but I added some fixes to CVS for
this yesterday. I'll have a look.

>     The slur between the grace note and the main note is definitely
>     buggy; close inspection shows that one end isn't rounded but ends
>     needle-like.

yep. Noted.

>     The last three slurs in the upper stave are not correct.  They are
>     notated with `( ... )' but they look like being written with `~'.
>     I think such slurs should start and end exactly above the stems.

Slurs suck, post 1.6.  

>     It is possibly better to center the `3' (denoting a triplet)
>     exactly over the middle stem.

I am not sure.  I tried looked at some examples, and often the 3 is
also above the exact center or above the note head. I agree that this
looks goofy though.

>     The trill sign is too big in comparison to the notes.

Still?! (It's already a lot smaller than in 1.4)

>     The rests in the fourth bar should be moved up automatically -- we
>     are still using \VoiceOne, aren't we?  The same should be done
>     for the half rest in the left hand.

Ours refs aren't clear. We talked with a guy from Universal Edition
who maintained that rests should not be moved at all except to
prevent a collision.  OTOH, he might have been talking about single
voice music (like in beams, as you note below).

>     To improve readability I suggest that you use @smallexample for
>     the code fragment.

good idea. The examples take too much place, I think. I added the
option smallverbatim to lilypond-book.

>   . Section 3.9, first example
> 
>     According to the code, I expect the treble clef in the left hand
>     to be inserted in front of the lower staff's first note, but it
>     appears at the beginning immediately after the 3/8 sign,
>     overlapping with the beams.  Looks like a bug...

well, it's sort of experimental. It took a lot of effort to move the
clef under the notes in the upper staff. Noted. 

>     The default vertical extension of the small slurs is too big.  I
>     don't know (yet) how this is handled by LilyPond, but I assume
>     that the height of a slur is dependent on the length, right?
>     Which formula is used?

Slurs suck (see above).  The the control points  of the slur are
calculated as

           (0,0),(0,h),(w-h,h),(w,0)

where h = h_infinity * F (x * r_0 / h_infinity) and F (x) = 2/pi *
atan (pi x/2). See bezier-bow.cc

>   . Section 3.10.2
> 
>     The staff lines extend stick out slightly at the right of the
>     final bar.  Either the metrics of this bar line type is not
>     correct, or there is a positioning error in LilyPond.

That is curious. Are you sure it is not a rounding error of the
viewing software?

>   . Section 4.5.1, second example
> 
>     This must be a bug.  If I write [f'8 r16 f' g' a'], I should get
>     this:
> 
>        -----------
>        |     -----
>        |  7  | | |
>        x  7  x x x
> 
>      not this:
> 
>        -----------
>        |  --------
>        |  7  | | |
>        x  7  x x x
> 
>     The latter form may be used in modern music (?), but it shouldn't
>     be the default.

Can you formalize that for me? Do you want the beams to be connected
as if there were no rest?

>   . Section 4.8.5, first example
> 
>     Is a quick notation available to give a note more horizontal
>     space?  Something like `c\wwide\ppp c\wide\pp' comes to my mind...

Nope. 

>   . Section 4.9.4
> 
>     The code
> 
>         \repeat tremolo 4 { c'16 d'16 }
> 
>       gives the following warning:
> 
>         Interpreting music...
>         programming error: Skipped something ?! (Continuing; cross thumbs)
> 
>       (the last line is repeated 3 times).
> 
>       Besides that, the result is ugly (beam too short).  I think this
>       is another place where the width of a graphical element (the
>       tremolo beam) isn't taken into account for computation of the
>       spacing.  BTW, selecting `8' for the repeat counter two other
>       bugs show up: An incorrect (or rather inappropriate) warning
>       appears
> 
>         warning: beam has less than two visible stems:
>         \repeat "tremolo" 8 { c'16 d'16 }
> 
>       and the tremolo beam's vertical position is far too high.

Tremolo beams are low on my priority list. I know that the spacing is
weird (it's even documented :^). If anyone desperately needs them to
be correct, I am sure I will hear about it.

>   . \clef "bass_8" is ugly.  The `8' glyph is too big and should be
>     below the lowest staff line.

I made it smaller, and added the staff to the support of the 8. 

> Keep going!

yeah, you too! :)

-- 

Han-Wen Nienhuys   |   address@hidden    | http://www.cs.uu.nl/~hanwen/