RE: Welsh orthography in Unicode

From: Philippe Verdy (verdy_p@wanadoo.fr)
Date: Sat Jun 28 2008 - 04:50:25 CDT

  • Next message: Leo Broukhis: "Re: Welsh orthography in Unicode"

    André Szabolcs Szelp wrote:
    > > BTW, not to forget the elipsis: Never type three dots, use
    > always one
    > > character, U+2026. :-)
    >
    > As I figure it, "Ellipsis" was actually a descriptive name
    > for the character, not a functional one. It made it's way
    > into unicode via legacy encodings where it served as a "three
    > dot leader" (note the grouping with "one dot leader" (2024)
    > and "two dot leader (2025) which were used for setting
    > "leaders" as in TOCs between titles and page numbers[...]
    >
    > Actually you should never use U+2026 for a grammatical
    > ellipsis. Use three periods. And if you are picky, increase
    > tracking. Or commission your local type designer to implement
    > an OpenType feature to replace the three-period sequence by
    > your "ellipsis" glyph.

    I also do agree that the grammatical ellipsis should never look different
    from the succession of three full stops (and so there is no good reason not
    to use regular full stops). Using the encoded "ellipsis" looks ugly,
    especially when using monospaced fonts, because its dots are too small
    compared to other punctuation signs.

    The encoded "ellipsis" is a misnomer, for me it is just akin to joining
    dashes of various length (notably those in the box-drawing block), so it is
    effectively just a three dot leader. But even its use for dot leaders is
    ugly, because you can' really know how many dots you need to create the
    proper column alignment. If you were using HTML for example, you would
    probably better represent the leaders using text decoration (through CSS,
    i.e. style) like a dotted line. Leaders do not represent text by themselves,
    they are just used to decorate tabulating spaces of variable length (the
    actual number of dots does not carry any semantic information in such
    tabular layout, and leaders never play the role of punctuations, they are
    not different from the space or tabulation that they just decorate). Even
    the one-dot, two-dots and three-dots leaders are just there for
    compatibility because the actual number of dots in leaders does not matter
    (so you can group them as if they were a single tabulating space): for me
    they are just like a decorated TAB.

    On the oppposite the full stops are actual punctuation meant by the
    grammatical ellipsis: two dots may be used as an alternative to a
    grammatical en-dash to mean bound (closed) intervals, where it is used to
    mean "from..to". Three dots are typically used for half-open intervals or
    unterminated lists/enumerations. In both case, this is not a decoration.

    Note that there should also be a different presentation between full stops
    (used semantically in ellipsis notations or as decimal separators) and
    leader dots: Imagine a TOC that needs to display a list of exact titles
    including their punctuation. The dots of tabulating leaders should be tiny
    and their displayed number will adapt to the fonts actually used for their
    rendering (and the size and resolution) of the rendering surface, but the
    punctuation full stops (and ellipssis) must NOT varying the numer for dots
    that must still remain identifiable, even if they are immediately followed
    by tabulating leaders. But many proportional fonts do not differentiate the
    dots used in full stop and ellipsis (however the layout of ellipsis is often
    wrong as the left+right bearings are frequently different from the gaps
    between the dots, and frequently the one-, two- and three-dot leaders do not
    align properly as expected)

    My opinion is that two- and three-dot leaders are just meant for round-trip
    compatibility with legacy encodings and legacy monospaced fonts used on
    consoles with ugly text layout grid: as the number of columns was limited,
    it was conceivable to have several leader dots packed into the same cell,
    but using these "ellipsis" characters as punctuation was really even more
    ugly and confusive.

    Anyway this is quite a battle of religions, some other people (and some
    wellknown word processors...) will still insist to correct your own three
    full stops into three-dots leaders (aka "ellipsis"), and don't recognize
    that leaders are in fact ugly and confusive punctuation signs, and that
    there is absolutely no possible confusion when using a *counted* number of
    full stops to write the grammmatical ellipsis.



    This archive was generated by hypermail 2.1.5 : Sat Jun 28 2008 - 10:45:05 CDT