APL - the APL keyboard is a non-problem

Author

the APL keyboard is a non-problem

Ibeam2000

2006-06-17, 8:01 am

Although I have no particular problem with keyboard support as provided
by any of the vendors (except that the APL2 demo keyboard loses the "Q"
key on my fancy M$ non-carpal-causing keyboard for some reason), many
of my clients felt that the reliance on yet another piece of
specialised hardware was very detrimental to APL's future in their
organisation. Also, the presence of strange characters in a computer
language and the need to type them in also helped any decision to stay
with APL more questionable.

I am the kind of person who could be easily (and correctly)
characterised as sort of a looney, perhaps the kind of person that APL
was made for. I started using APL for just about everything when I saw
just how much (or little) code I needed to deal with in order to solve
my particular problem that day in the autumn of 1975. My filter
modelling problem consumed some pages of Fortran code, but was replaced
by lines of APL - plus, I could plot, albeit crudely, my results. Life
was sweet. Also, I thought that the special characters, not to
mention the exquisite way the 2741 dealt with overstrikes, was kind of
special. For me, mainstream programming took on a new form of
mediocrity.

31 years later, my clients, being normal and sensible people, feel no
such attachement or attraction whatsoever to APL. Well, one liked it,
but he owned the company and was somewhat of a looney too.
Regrettably, he's no longer in business. But for most existing APL
deployments, APL is used because it was there before, because it seems
suitable to the problem at hand, and because the replacement of the
application may incur extra-orbital astronomical costs and have a
probability of success which is considerably less than 1.000. They
have no attachment to the technology, other than the adverse benefits
of neglecting it.

Generally, the "management" feeling is that something must be terribly
wrong with anything, be it a computer language or tool, which imposes
such conditions upon its users, both that they have to type in these
funny characters and that they may need a special device to use it
efficiently. Management feels that APL licences are not cheap, and now
to have to buy additional keyboards for its staff is over the top.

The management mentality may be completely alien to us freethinking
anti-corporate types. These people look for problems, or what they
perceive to be problems, and base their decisions on what they consider
to be more problems lurking underneath. We all know that the keyboard
problem isn't, but management (with the power of no) may disagree. And
therein lies the death of APL.

APL is being rejected for very stupid reasons.

I showed someone APL for the first time yesterday, and the first thing
the guy noticed is the presence of strange characters not ordinarily
found in programs.

One of my clients mitigates the problem by simply not requiring the use
of APL characters, or "APL signs", as they call them in end-user
programs. In this particular case, the client built a domain-specific
language, and APL is simply the scripting and execution environment.
Anyone remember 39 MAGIC from Sharp? Same sort of idea. MAGIC was
built such that it could be used from non-APL hardware. Interestingly,
the addition of control structures to the language makes it that much
easier to write scripts without the use of those dreaded APL signs.
That in and of itself eliminates the need to use the branch arrow, and
depending on how you write branches, rho, take, rotate, times, iota.
The quality of a program is judged by the lack of APL symbols.

'CAT' is 1 plus getdata 'FOOBAR' No APL signs.

To be continued........

AA2e72E

2006-06-17, 8:01 am

Ibeam2000 wrote:
> The quality of a program is judged by the lack of APL symbols.

And, there was I thinking that the quality of a programme/application
is judged by its reliability, its usability, and whether it did what it
was supposed to do and, above all, it empowered users to achieve what
they set out to do.

I have never known anypne judging an application by scrutinising it
code; however, I have seen users extract every ounce of functionality
from an application that they like. Users could not care less about the
semantics of the language or the actual construction of an application.

Perhaps if APL engraved keyboards were more common, fewer people would
be diamayed at the sight of the APL symbols. An APL engraved keyboard
might be specialist hardware but requireing one is no more of a
'hardware dependence' that requiring a UK keyboard with a Euro symbol.
Multinational organisations routinely acquire specialist keyboards that
support, say, the Chinese or Japanese characters.

Special symbols on a keyboard do not get in the way either. If they
did, there would not be such a plethora of keyboards that have
engravings for launching dedicated functionality for such things as
multimedia, internet browsing, Microsoft Office etc.

There is no product on any market that jeopardises its usability in the
same way as APL.

Nancy Wheeler

2006-06-17, 8:01 am

The problem with the Q key is fixed and the fix was shipped in Service
Level 8 (April 2006).

Nancy Wheeler
APL Products and Services

Ibeam2000 wrote:
> Although I have no particular problem with keyboard support as provided
> by any of the vendors (except that the APL2 demo keyboard loses the "Q"
> key on my fancy M$ non-carpal-causing keyboard for some reason), ...

Ibeam2000

2006-06-18, 7:56 am

To set the record straight, I have just downloaded the latest APL2 demo
version (my previous version was Service Level 7, this one is Service
Level 8) and the built-in keyboard program does indeed correctly
display the "Q" key.

I wrote:
> Although I have no particular problem with keyboard support as provided
> by any of the vendors (except that the APL2 demo keyboard loses the "Q"
> key on my fancy M$ non-carpal-causing keyboard for some reason), many
> of my clients felt that the reliance on yet another piece of
> specialised hardware was very detrimental to APL's future in their
> organisation. ...

I apologise for this untimely (parenthetical) remark, as this problem
was corrected long before the post. I should stay more current.

Sponsored Links