On 5/16/08, Patrick van Beek <[EMAIL PROTECTED]> wrote:
> I was wondering how members of the forum go about construction of Tacit
> statements?

I approach tacit programming rather like algebraic manipulations,
in that I first focus on getting something that works properly and
then I try and find various small changes which are equivalent for
the case I am interested in, with an eye towards eventual simplicity.

Note also that I use J's command line rather extensively here.
I am most productive when I have some representative and
manageable sample data for my tacit expression(s) to work
on.  Then I can just recall a previous line, change something
and try and see if I like the result.

However, for complicated expressions I may break out
sub-expressions and focus on them with suitable arguments.
(Mostly I do this when I am trying to understand other people's
code, but sometimes I find this useful with my own code --
especially when some experiment did not turn out like
I expected.)

-- 
Raul
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to