General punctuation spaces U+2000 to U+200B

From: William Overington (WOverington@ngo.globalnet.co.uk)
Date: Sat Apr 26 2003 - 04:09:11 EDT

  • Next message: Kevin Brown: "Re: Private Use Area"

    I am preparing a font named Quest text which includes various ligatures from
    old printing, using the code points of U+FB00 to U+FB06 and various Private
    Use Area code points in the range U+E700 to U+E7FF. Characters for Old
    English and Esperanto are also included. Some
    other characters will hopefully be included as time goes on.

    I am hoping to implement the spaces in the range U+2000 to U+200B. This is
    so as to allow Quest text to be used as a continuation of the fun of
    handsetting of metal type. The space U+0020 is available for general use
    with Quest text.

    I have been looking at U40-2000.pdf and ch06.pdf and now write to ask for
    advice in the matter please.

    What is the difference between U+2000 EN QUAD and U+2002 EN SPACE please?

    What is the difference between U+2001 EM QUAD and U+2003 EM SPACE please?

    In letterpress printing, with handsetting of type from a metal font which
    had been cast by a typefounder, I always thought of a quad as something like
    a 2 em quad or a 4 em quad, that is, as a multiple em width block of spacing
    used at each end of a line of handset type to give stability, or used within
    a line to space out a fleuron or some such ornament. So an em quad seems
    not to fit with that, so what is the meaning please?

    U+2003 carries the note "nominally, a space equal to the type size in
    points" and the note "may scale by the condensation factor of a font".

    Quest text has vertical values in the range -768 to 2048. In fact, the -768
    is given by U+E700 STAFF so as to force the height, yet no descenders as yet
    go below -512 font units.

    So, how wide should I make U+2003 in font units please?

    If Quest text were a metal type font, then the point size would be the
    equivalent of 2816 font units ( that is 2048 + 768). Yet in electronic
    typography the point size is, as far as I am aware, the equivalent of the
    2048 font units above the base line.

    I realize that ultimately Quest text is just a font produced by one artist
    so if I choose to implement U+2003 as 2816 font units wide and then
    implement U+2002, U+2004, U+2005 and so on in proportion to that width that
    that will probably be acceptable within that font, yet I am hoping that it
    will be a work of art which helps bridge the gap from handsetting of metal
    type in days of old to modern electronic typography so I would like to get
    it right and therefore ask for advice please.

    I had not known of the expression "condensation factor of a font" before
    reading it in the note about U+2003 and wonder if someone could possibly say
    more about this term please.

    Is there a general value which is used to implement U+0020 in electronic
    fonts please? I seem to remember that the space used between words in
    ordinary setting of body text with metal type was to start with a thick
    space, namely one third of one em, when composing the words and then to go
    wider as necessary once the end of the line were reached, so as to achieve
    justification.

    I am hoping that by using SC UniPad (downloadable from the
    http://www.unipad.org webspace) with the chosen option of rendering spaces
    as picture glyphs, that a sequence of spaces from the U+2000 to U+200B range
    can be set in SC UniPad, then copied and pasted into an application such as
    WordPad where the Quest text font is being used, and a simulation of hand
    justification of text can be produced, as a hopefully interesting heritage
    continuation of the typesetting skills of the past. Certainly, this may
    take some investigation due to the pixel nature of the rendering of fonts,
    yet the investigation will hopefully be an interesting educational project
    in itself.

    William Overington

    26 April 2003



    This archive was generated by hypermail 2.1.5 : Sat Apr 26 2003 - 04:45:38 EDT