Dan,

You just wrote a part of my "Read This First" page that I am proposing 
for the NuVoc site. Thanks.

Skip Cave

Dan Bron wrote:
>>  You could show the J code using variables such as x and y, and display the
>>  values of x and y graphically without showing how they were created.
>>     
>
> In the front matter of the dictionary, we could declare some notational 
> conventions, such as
>
>       N  is a noun and always refers to a scalar nonnegative integer
>       (e.g. the output of monad  #).
>       Np is a noun and always refers to a scalar positive integer.  
>       Nn is a noun and always refers to a scalar negative integer.  
>       Nn0 is a noun and always refers to a scalar non-positive integer.
>       P is a noun and always refers to a scalar prime integer.
>
>       T is a noun and always refers to a table of numbers,
>       unless otherwise specified, T is equal to the rank 2 numeric
>       array 
>                  i. 4 4 
>                0  1  2  3
>                4  5  6  7
>                8  9 10 11
>               12 13 14 15
>
>       that is,  T=:i. 4 4  .
>               
>       B is a noun and always refers to an (arbitrary rank) array of
>       boxes.
>
>       G is a gerund, normally a vector of boxes (each box the atomic
>       representation of some entity, but normally the entity is a verb).  
>
>       f is a verb, and if another verb is needed, it is named g.  
>       Further verbs are named v0, v1, v2, etc
>       ....
>
>       u is the left argument to a conjunction or adverb; it may refer to a 
>       noun or verb. Unqualified, it often its often a proverb; context will 
>       resolve any ambiguities.
>
>       v is the right argument to a conjunction; it may refer to a pronoun 
>       or proverb. Unqualified, it often its often a proverb; context will 
>       resolve any ambiguities.
>
>       m is the left argument to a conjunction or adverb when it 
>       unambiguously refers to a noun.
>
>       n is the right argument to a conjunction when it unambiguously
>       refers a noun.
>
>       When defining adverbs or conjunctions which consume verbal
>       arguments, we may refer to  the ranks of these verbal arguments as 
>       follows:
>               
>               adverb or left argument to a conjunction:
>               lr     the entire (3 atom) rank of the verbal argument
>               lmr the monadic rank of the monadic argument
>               llr    the left rank of the dyadic argument
>               lrr    the right rank of the dyadic argument
>
>       with intial "l"s replaced with "r"s for verbal right arguments to 
>       conjunctions.
>
>       Mnemonic: left arguments always have single-letter names which,
>       in the alphabet, appears one letter to the left of the letter selected
>       for the corresponding right argument; x vs y, u vs v, m vs n, f vs g.
>       Note that this may run counter to the mathematical convention 
>       which reverses these roles (e.g. a math function with a single 
>       argument will name it x, and if a second argument is added it is 
>       named y; but in J the initial argument is always named y and a 
>       second argument  is always named x)
>
>       ....
>
>       Unless otherwise specified, "vector" and "list" mean a rank 1 noun.
>       For example,  0 1 2 3  and  'abcd'  are both (unqualified) vectors or 
>       lists.  We may also refer to vectors or lists "of"  items, such as 
>       vectors of tables or vectors of vectors.  For example  i. 4 4   can be 
>       seen as a table, or as a vector of vectors, and i. 4 4 4 can be seen 
>       as a cube or a vector of squares.  
>       
>       Unless otherwise specified, "table" means a rank 2 noun.  For 
>       example, i. 4 4 is an (unqualified) table of  16 integers.  We 
>       may also refer to tables OF items, such as tables of vectors. 
>       For example
>       ....
>
> This is in the tradition of the current DoJ [1]  (and other reference texts), 
> while being more verbose and "tutorial", which is how
> the new dictionary differs from the current one.
>
> -Dan
>
> [1]  http://www.jsoftware.com/help/dictionary/dict3.htm
>
>
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
>
>
>   
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to