Mailing List Archive


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [tlug] Re: Piping stderr?



>>>>> "Jiro" == Jiro SEKIBA <jir@example.com> writes:

    Jiro> Stephen J. Turnbull <stephen@example.com> wrote:
 
    >>> uxterm totally depends on UTF-8.

    >> What do you mean by "depends"?  That is, what restrictions does
    >> this impose on me?

    Jiro> You have to use UTF-8 codeset or filter to use uxterm, isn't
    Jiro> it restriction?

Not as far as I'm concerned.  The filters are freely available.  I
only have to actually type the filter pipeline once, after that I make
a shell function.

    Jiro> And uxterm's text processing depends on UTF-8 code set, I
    Jiro> can say.

Sure, but all that matters is (1) chunking characters in the buffer
and (2) converting to integers for font indexing, both operations are
trivial and efficient.  This is very different if you want to handle
all of EUC and Shift JIS and ISO-2022-JP.  In Shift JIS you can't even
handle all Japanese characters; EUC is the best of the three and
reasonably consistent but what if you want to write Hangul or Hrvtska
(Croatian); and ISO-2022-JP means you don't even know where characters
begin and end without going all the way the back to the beginning.

Why burden a stupid little *term with that complexity?

    >> Software should be written to be general, with as little
    >> dependency on external text representation as possible

I mean it should be easy to put filters on it so as to handle
arbitrary external representations.  These filters can be internal as
with Mule coding systems or Python codecs, or external such as piping
through iconv.  All knowledge of non-Unicode coding should be severely
limited to a few modules.

It should be like a Japanese house.  You may wear sandals, shoes,
sneakers, or boots outside, but when you enter the house you take off
your outdoor shoes and put on slippers.  In the same way, at the time
text enters a program, it should be converted into a universal
representation, ie, Unicode.  It doesn't matter whether the getabako
is inside the program (as the genkan in most homes) or outside (as at
my daughter's school).


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
 My nostalgia for Icon makes me forget about any of the bad things.  I don't
have much nostalgia for Perl, so its faults I remember.  Scott Gilbert c.l.py


Home | Main Index | Thread Index

Home Page Mailing List Linux and Japan TLUG Members Links