I certainly agree with you about the nature of the problem, when writing J.

It's just that we also need to solve the implementation issues
(hopefully without making a horrible mess of things) if we want to
solve this.

Thanks,

-- 
Raul

On Mon, Jan 21, 2013 at 2:13 PM, Don Guinn <[email protected]> wrote:
> Why do I want to put something like (3 : 0)"0 to start the definition? In
> many cases it makes the definition clearer. And avoids having to repeat the
> "0 in every use to avoid the definition failing if it only works for rank
> 0. Also, sometimes I want to get some information before the enclosing
> local names are hidden.
>
> The problem I ran into when wanting to include rank or some other tacit
> processing on the data before entering the explicit definition is that the
> debugger then treats the entire explicit part as one line. To avoid this I
> have to define the verb without the "0 for testing, then add it afterward.
>
> On Mon, Jan 21, 2013 at 11:21 AM, Raul Miller <[email protected]> wrote:
>
>> Note that this creates a new problem -- what kind of state does the
>> debugger need to track?
>>
>> Without the "0 we only have one instance of the explicit verb state.
>> With the "0 we can have millions of instances of the explicit verb
>> state.  Of course, we only execute one instance at a time, but now we
>> are asking the interpreter to do some kind of extra tracking for verbs
>> derived by (") and not just in verbs derived by ( :) ...
>>
>> Anyways, it's doable, but I'm not convinced that "just addressing this
>> one common case" would yield simpler code than what we'd get from a
>> proper design.
>>
>> But maybe Roger or Henry or others have deeper insights than I (and I
>> will note that there's a related struggle having to do with how we
>> handle the display when intermediate results become both numerous and
>> large).
>>
>> --
>> Raul
>>
>> On Mon, Jan 21, 2013 at 1:04 PM, Don Guinn <[email protected]> wrote:
>> > Being able to step through a tacit definition as Henry's work shows sure
>> > would be great. But as a start it would be nice if the explicit debugger
>> > like the one in J6 did not get confused if a definition included some
>> tacit
>> > expression like "0 in the definition line.
>> >
>> > test=:(3 : 0)"0
>> > (rank 0 processing)
>> > )
>> >
>> > On Mon, Jan 21, 2013 at 9:00 AM, Raul Miller <[email protected]>
>> wrote:
>> >
>> >> That sounds great!
>> >>
>> >> What does it mean?
>> >>
>> >> (Note that Henry's recent efforts might be relevant, here.)
>> >>
>> >> Thanks,
>> >>
>> >> --
>> >> Raul
>> >>
>> >> On Mon, Jan 21, 2013 at 9:54 AM, Don Guinn <[email protected]> wrote:
>> >> > How about the debugger handle tacit definitions as well as explicit
>> ones?
>> >> >
>> >> > On Mon, Jan 21, 2013 at 6:24 AM, Raul Miller <[email protected]>
>> >> wrote:
>> >> >
>> >> >> On Mon, Jan 21, 2013 at 3:58 AM, Linda Alvord <
>> [email protected]>
>> >> >> wrote:
>> >> >> > However, I do not understand how ic and ic2 agree when they don't!
>> >> >>
>> >> >> . . .
>> >> >>
>> >> >> Personally, I sometimes wish that a rank decoration or other modifier
>> >> >> on an explicit definition would not interfere with debugging.  I
>> think
>> >> >> that this issue introduces two different phrasing styles, and there's
>> >> >> a tension between them that I am not totally sure is good for us.
>> >> >> . . .
>> >> > ----------------------------------------------------------------------
>> >> > For information about J forums see
>> http://www.jsoftware.com/forums.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
>> ----------------------------------------------------------------------
>> For information about J forums see http://www.jsoftware.com/forums.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