In article <1156917038.977615.324430@[EMAIL PROTECTED]
>,
Romanise <jo****dm@[EMAIL PROTECTED]
> wrote:
>> Romanise <jo****dm@[EMAIL PROTECTED]
> wrote:
>> >If programmers come to recognise that a given language can effectively
>> >communicate with certain least number of letters and diacritic marks
>> >then they would be halfway to devising the optimal tool for given
>> >language.
>> > Herman Rubin wrote:
>> The least number is 1, or 2 if one considers spaces to be
>> characters.
>Please elaborate your point.
>Space would be one charater, besides letters and diacritic marks Indian
>scripts require.
This is well known in the area of Turing machines, and
the proof is trivial.
Suppose we have a finite, or even countable, "alphabet",
including spaces and diacritical marks; I assume that
the marks can be made linearly, or we use a method of
indicating where they go which is linear. It is not
necessary that the marks occupy all the integers, just
that no two marks correspond to the same integer. Then
the mark labeled "k" would be represented by k 1's followed
by a space.
>> Of course, what do you mean by "effectively"?
>Languages often carry a lot of historical baggage with them
>particularly when those with writing skills want to keep their club
>exclusive. Writers even create unnecessary letters and symbols in name
>of "accomodating" more than one language.
This can be handled by setting up a system to add new
characters.
Note: I do not recommend the above. But all of the
communications in this newsgroup use 0's and 1's, with
spaces having their code. So 2-character representations
are not that unusual.
--
This address is for information only. I do not claim that these views
are those of the Statistics Department or of Purdue University.
Herman Rubin, Department of Statistics, Purdue University
hrubin@[EMAIL PROTECTED]
Phone: (765)494-6054 FAX: (765)494-0558
|
