I am not sure which 'it' you are referring to. If you mean why I think APL was
widely accepted at this company, my best guess is that it started with the
powerful corporate group that had been educated in their field with APL as a
fundamental tool for their work. This group had a good story for upper
management of how APL would help this group provide real time business analysis
and much faster market response in an era when punch card input, green bar
output and multi-month COBOL development cycles were the norm.
Having a powerful sponsor that demanded APL got APL in the door. The initial
marketing of APL as a pay as you go service made the financials attractive for
rapid and relatively low cost startup and allowed almost free incremental usage
by other groups.
I suspect that one of the most important factors helping APL move beyond a
specialized support tool for a single functional area was the fact that this
version of APL came with an international network in an era when national PTT
laws and exclusionary regulations made it exceedingly difficult to set up such a
multinational corporate data network. This APL network supported the
establishment of APL as the international email hub of the company, which got
APL terminals on the desks of almost all corporate office workers and management
(and even on factory floors) world wide.
Finally, the extensive libraries of APL based tools and pre-built applications
that came with this version of APL provided the relatively naive developers and
users with a set of building blocks and examples that allowed them to use the
techniques of incremental real time development to solve real world problems
much more quickly and effectively than competing tool sets could.
As far as I can tell, the beginning of the end of APL at this company came from
much the same genesis as the beginning: how APL was used and portrayed in
education. After the period of APL'S ascendancy, the new crop of MBA power
brokers were taught management theories that had no place for APL (or most other
idiosyncratic computer languages) in the new models of corporate information
management. The original young lions that championed APL were now seen as aging
advocates of a legacy support nightmare.
Some APL developers now started to drop support of APL and retooled to go to
where the corporate sun shone. As the number of developers supporting APL
declined, the self-fulfilling prophecy of "lack of support for a non-standard
language" influenced program managers to chose other options.
And there were now other options available. APL was no longer the only choice.
The software and services market had caught up with APL'S early monopolistic
lead and now supported the new corporate data services model which had shifted
from custom solutions to off the shelf standard deliverables that the new movers
and shakers demanded.
So, as I see it, while the issue of APL character set was used as ammunition in
minor skirmishes, the APL character set was never the overriding issue for the
decisions that led to using or not using APL and APL-developed solutions at this
company.
On 4/12/2013 4:32, Björn Helgason wrote:
what is it then?
On Apr 11, 2013 7:55 PM, "David Mitchell" <[email protected]> wrote:
I'll throw in a comment/story that is mostly true. I was hired at a large
company many years ago. One of my assignments was to eliminate APL usage,
which was seen as overly expensive and not in the preferred company
direction. At the time, APL was probably the most widely used language in
the company.
I succeeded in my assignment to eliminate APL after many decades of work
in 2007. I suspect that there were pockets of continuing APL usage via the
PC versions of APL (or J).
There are many reasons why APL was so difficult to eliminate. In my
opinion, these are mostly the same reasons that led to APL being, from many
points of view, one of the most important computer languages and
implementation environments in this company for many years.
From what I saw of the adoption and later decline of APL usage, the APL
symbols were fairly far down on the list of the reasons for or against the
usage of APL at this company.
In a way, this current discussion of symbology reminds me of the
internecine battles in the past over the theoretical correctness of various
implementations of enclosure.
I will say my reasons for using both APL and J (and the dozen or so other
languages I use regularly) have not much to do with their usage of
symbolics or keywords. After learning the first few dozen languages, I
find that these differences are not very important to me.
On 4/11/2013 14:56, Björn Helgason wrote:
On Apr 11, 2013 5:47 PM, "Joey K Tuttle" <[email protected]> wrote:
I have been tempted, several times, to make one comment and that is a
feeling that the APL character set was perhaps the single most important
reason for lack of widespread acceptance and use of APL...
I believe you are right.
PS: A comment like that to c.l.a would not be popular.
------------------------------**------------------------------**
----------
For information about J forums see
http://www.jsoftware.com/**forums.htm<http://www.jsoftware.com/forums.htm>
------------------------------**------------------------------**
----------
For information about J forums see
http://www.jsoftware.com/**forums.htm<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
