Re: Feedback on Initial letter for CSS


thank you for sending these details.  I have one or two questions and 
comments.  Sorry it has taken me so long, but I wanted to better 
understand initial letter in general before composing my questions.

[1] on page 2 of the PDF you say "In Indic scripts the top reference 
points is the hanging base line", but not all indic scripts have a 
connecting line at the top (eg. tamil, malayalam, ...). Is this 
information relevant to those scripts?

[2] Suppose we have two articles, both starting with boxed initial 
letters. Suppose that the first article starts with ঞ and the second 
starts with ন্ত্বী (ie. the second is likely to be twice the height of the 
first).  Would it be normal to use the same sized box for both articles? 
  If you did, presumably there would either be a lot of unused space in 
the first case, or a much more shrunken character in the second.

Or would one try to use the same relative size for the characters and 
use different box heights?

[3] I assume that, where there is no box and the top line alignment is 
in effect, the size of the exclusion area would depend on how many 
subjoined glyphs are attached to the main glyph in the syllable.

[4] The approach you outline, where a syllable has 'ink' above the top 
line and the top line is aligned with that of its neighbours, brings in 
the idea of using box size to accomodate the 'ascenders'.  I wonder 
whether this is (a) trying to define the requirements via the technical 
solution, and (b) attempting to mix two different models.  Let me expand 
on (b).

For Latin text, as I understand it, there is a notion of alignment of 
cap height with the neighbours, but that model just lets the accents 
rise above the cap height and deals with them by expanding an exclusion 
area around the main part of the character that avoids overlapping of 
the initial letter ink with the surrounding characters.  This model may 
be sufficient for the Indian case too.  Setting up a box and then trying 
to also align the top lines of teh characters sounds rather complicated 
to me.

But maybe what the document needs to say rather is simply this:

"Where the top line of the initial letter is aligned with the top line 
of the adjacent normal text, vowel signs and other diacritics may rise 
above the top line.  These must not overlap with text before the line on 
which they appear."

[5] As i understand it, your document proposes that the size of the 
initial letter *when not in a box* is determined by the number of lines 
between the top line of the syllable (where there is one) and the after 
edge of the syllable cluster – meaning the lowest bit of ink in the 
syllable where vowel signs, subjoined consonants or other diacritics 
appear.  Is that correct?

Note that, by contrast, in the Latin case the size is determined by the 
distance from cap height to alphabetic baseline – any descenders or 
diacritics below the baseline are not considered for sizing and care 
must be taken to establish an exclusion area, where needed, to ensure 
that they don't overlap with surrounding text.

I just wanted to be sure that there is not some such lower metric in the 
indic case – especially where a script has no top line (eg. Tamil, 

[6] I assume that the alignment of the top line of non-highlighted 
characters is at the top of the thicker top line of the initial letter. 
Is that correct?

[7] there are some instances in the examples where the top lines of the 
initial letter and the following letters don't appear to actually touch. 
What are the factors involved in that?


On 04/02/2015 07:24, Somnath Chandra wrote:
> Hello Richard,
> Please find enclosed the feedback on initial letter requirement for CSS
> In-line. Pl review it and if you find it suitable or necessary to
> include in the Indic Layout requirement , you may pl. do so.

Received on Tuesday, 10 February 2015 17:32:58 UTC