Re: Fixes position of mensural c clef (issue 6503091)

[Phil Holmes]

----- Original Message ----- 
From: <address@hidden>
To: <address@hidden>; <address@hidden>; <address@hidden>; 
<address@hidden>; <address@hidden>
Cc: <address@hidden>; <address@hidden>
Sent: Monday, September 24, 2012 6:34 PM
Subject: Re: Fixes position of mensural c clef (issue 6503091)


> The glyph shape is OK, thanks.  Have you tested the appearance on paper
> of all available sizes, using a 300 or 600dpi printer?

No.  Does this often come out strange?  When you say "paper of all available 
sizes" I don't understand why changing the paper size should have any 
effect.  Could you explain?

> http://codereview.appspot.com/6503091/diff/16001/mf/parmesan-clefs.mf
> File mf/parmesan-clefs.mf (left):
>
> http://codereview.appspot.com/6503091/diff/16001/mf/parmesan-clefs.mf#oldcode902
> mf/parmesan-clefs.mf:902:
> The dimension of the bounding box is not correct.  It's far too large,
> as can be seen in the DVI proof.

I was initially aware of this, but then forgot. Is there guidance on the the 
correct way to calculate the parameters to set_char_box()?

> http://codereview.appspot.com/6503091/diff/16001/mf/parmesan-clefs.mf
> File mf/parmesan-clefs.mf (right):
>
> http://codereview.appspot.com/6503091/diff/16001/mf/parmesan-clefs.mf#newcode791
> mf/parmesan-clefs.mf:791: ..z2+(0, blot_rad)
> This must be `.. {up}z2 + (0, blot_rad)', otherwise the curve doesn't
> start with the right direction.  This wouldn't be that bad, but it were
> different to all other glyph shapes.
>
> Ditto for all other transitions from `--' to `..' (this is, using the
> proper `{up}', `{right}', etc.).

Excellent.  I was aware of the discountinuity, but not the bext way to fix 
it.

> http://codereview.appspot.com/6503091/diff/16001/mf/parmesan-clefs.mf#newcode793
> mf/parmesan-clefs.mf:793: --quartercircle scaled blot_diameter rotated
> 180 shifted (z3r + (blot_rad, blot_rad))
> Using `quartercircle' gives good results in the type1 output.  I only
> ask you to make the line fit into 80 columns :-)

OK.  I really should tell you to get a bigger monitor :-)

> On the other hand, why not simply using something like
>
>   startcurve_point{down} .. {right}endcurve_point
>
> ?  This might be shorter, and easier to read.

Didn't think of this as an option.  The benefit of quartercurve is that it 
takes the place of 2 points, so is actually shorter.  I'd like to stick with 
what I now have, but will remember this for the future.

> http://codereview.appspot.com/6503091/

--
Phil Holmes