Thanks again for your comments.  I'm oing to go back and scrutinize
those keyboards again.  The result I ave so far is a bit better than
what I was expecting but there is certainly romm for improvement.

 - michael

On Sun, Jan 8, 2012 at 10:26 AM, bob therriault <[email protected]> wrote:
> Thanks for your response Michael.
>
> I agree with Bjorn that this work holds great potential and encourage you to 
> pursue it.
>
> Comments inline
>
> Cheers, bob
>
> On 2012-01-07, at 10:19 AM, Michael Dykman wrote:
>
>> Hi Bob,
>>
>> Comments inline.
>>
>> On Sat, Jan 7, 2012 at 12:29 PM, bob therriault <[email protected]> 
>> wrote:
>>> Hi Michael,
>>>
>>> I think that your qwerty screen is excellent.
>>
>> Thank you.
>>
>>> On the number screen I think that the underscore [_] should be added to 
>>> allow for the entry of infinite and negative infinite into number vectors 
>>> without changing screens, also it would be useful to have y on this screen 
>>> if you are going to include x.
>>
>> The underscore is there, immediately to the left of the '1'.  As for
>> selecting which letters made it to the numpad, I wanted to squeeze in
>> all the characters used for forming literals.  The 'x' is provided for
>> extended integers.  Do you think providing 'y' should take precedence
>> over having all the hex keys there?  Space is pretty tight but I could
>> gain some more by giving up on the oversized digits.  Alternatively, I
>> could just reduce the return key to standard size and put the 'y' in
>> the vacated space.
>>
>> Thoughts?
>>
>
> I see the underscore clearly now that you point it out :) .  I think I would 
> reduce the return key size as this puts the x and y beside each other, but 
> thinking it through I am not sure it is needed at all and if it were there 
> would also be a case for m, n, u, and v. See comments on symbol screen below.
>
>>> I assume that the [...] key takes me to symbols, the [123] key to numbers 
>>> and the [ABC] key to the qwerty.
>>
>> Correct.  The shift on the qwerty keyboard is really a caps-lock at
>> the moment.  I intend to convert it to the classic mobile tri-state
>> shift eventually.
>>
>>>> If this is the case I think that you can change the [...] key that appears 
>>>> on symbols to a shift to take you to your two character primitive layout. 
>>>> This would allow you enter tacit code by just using the shift key (or 
>>>> shift lock if you prefer) and the symbol layout.
>>
>> I thought of that.  In that case the 'shift' would produce identical
>> results to swapping in the alternate keyboard unless I mistake your
>> meaning.
>
> That is what I meant and that keeps all the primitives within one keystroke 
> of each other. It could be nice to have x and y included on the symbol 
> keyboard by making the lowest line smaller as this would allow explicit 
> programming in most cases without having to go back to qwerty. u, v, m, and n 
> would still not be available on the symbol keyboard, but there is only so 
> much space and most programming is done with x an y in verbs.
>
>>
>>> I might also suggest making sure that symbols such as period [.], 
>>> semi-colon[:], and parenthesis [(] and [)] are kept in consistent positions 
>>> across screens to allow for muscle memory. It can be really frustrating to 
>>> have the symbol change location according to the screen.
>>
>> I did keep '.' and ':' consistent between qwerty and sym1, my idea for
>> the numpad layout required a bit of a shift which I may rethink.  The
>> parentheses do deserve more consistent treatment..  thank you for
>> pointing that out.
>
> Actually there is position change of '.' and ':' between number and 
> qwerty/symbol keyboards. I think that this could be pretty important, so you 
> may want to lock down size and position on '.' ':' space '(' and')'
>
>>
>>> I think that is all that I see for now, but I am not a keyboard designer, 
>>> so take this very much as a crowd source approach to design :) .
>>
>> I am no keyboard designer myself.  It's the crowd-sourcing I am after
>> in this forum.  Thank you very much for your comments.
>>
>> Given the wild popularity of mobile devices, I think it is important
>> to get J onto a mobile device to broaden it's popularity.  I used to
>> carry a CE-device with J and it often struck me how the conciseness of
>> J made it an ideal language for mobile use.
>>
>> - michael
>
> Some of these ideas extend further into visual gui and touch interfaces as 
> well. Keep up the good work.
>
> Cheers, bob
>
>>
>> For information about J forums see http://www.jsoftware.com/forums.htm
>
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm



-- 
 - michael dykman
 - [email protected]

 May the Source be with you.
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to