When splitting a j sentence into multiple lines the programmer almost has to use more characters to assign intermediate definitions. These could have appropriately descriptive names. The person reading the code should be told what is the intended input data type, rank, and valence if non-obvious. The intended valence of tacit expressions without : is non-obvious, particularly if definition and use appear in separate sources. Please give the reader a chance to understand the algorithmic approach. Whereas I might easily deduce that a verb produces a binary vector I would like to be told up front, for example, that it produces a binary vector corresponding to numbers matching every third odd terms from Fibonacci sequence.
> Date: Wed, 17 Aug 2011 19:21:46 +0100 > From: mijj <[email protected]> > Message-ID: <[email protected]> > So it's not really a matter of how many characters there are on a > line. > It's a matter of clear grouping. ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
