Raul

>Most languages have their deficits. This is one as i spoke up about and tried 
>to encourage good steps at amelioration. Ragged arrays is another, as i have 
>noted earlier.

> Extending the capabilities of extended arrays may well be a good backdoor to 
> this capability. i do hope Dan goes along that line.

>As for myself, i am totally out of commission for the next few years, at 
>least, rebuilding my living situation. i have been absorbed in that for 4 
>years and at least now have a dwelling ... somewhat along the lines 
>~bright_plan_f.

---~
http://i.tgu.ca/bright_plan_f

greg
~krsnadas.org

--

from: Raul Miller <[email protected]>
to: Programming forum <[email protected]>
date: 13 August 2015 at 18:49
subject: Re: [Jprogramming] Bitwise operations utility

>If by "bit type", you mean a type that can take on 1/0, then I disagree.

>If by bit, you mean a representation which takes advantage of all bits in an 
>interval of memory, I also disagree.

>However, if by "bit type" you mean a type which takes on 1/0 values where 
>those 1/0 values are packed to take advantage of all bits in an interval of 
>memory, then I agree.

>But, yes, resistance - some of that is embedded in the hardware design (of 
>modern CPUs), and some of that has to do with the amount of work needed in 
>software to compensate for the hardware issues. This could almost double the 
>size of the interpreter.

>That said, if you feel inclined to develop support for bits as a densely 
>packed type, other people might appreciate your work. (Indexing might be good 
>places to start?)

Thanks,
Raul

--

from: greg heil <[email protected]>
to: Programming forum <[email protected]>
date: 13 August 2015 at 18:06
subject: Re: [Jprogramming] Bitwise operations utility

Raul

>You may have a better word for what i call bit-boolean if so i accede to your 
>greater knowledge. i and my colleagues just referred to it as boolean 40 or 50 
>years ago. Things may have changed in the language. However it is 
>incontrovertible that there is no bit type in J. i also contend there has been 
>a long standing resistance to having such. There may be extenuating reasons, 
>but it is definitely the poorer for those who would use such.

greg
~krsnadas.org

--

from: Raul Miller <[email protected]>
to: Programming forum <[email protected]>
date: 13 August 2015 at 15:35
subject: Re: [Jprogramming] Bitwise operations utility

On Thu, Aug 13, 2015 at 5:15 PM, greg heil <[email protected]> wrote:
>>i wrote my undergraduate thesis (in APL) using the manipulation of Boolean 
>>matrices (categories, and many other algebra objects - with arbitrarily large 
>>sizes). i always disliked J because of its avowed anti-Boolean typology. 
>>Another thing to be worked around.

Boolean means different things to different people.

>There's George Boole's approach, for example, and there's later work which 
>constrains the scope to truth values.

>Which are you talking about, here? And, why do you call J's approach 
>"anti-Boolean"?

>(We can take this to chat, if that helps - if we won't be discussing 
>programming.)

Thanks,

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

Reply via email to