Just noticed this today :-(
https://www-01.ibm.com/common/ssi/ShowDoc.wss?docURL=/common/ssi/rep_ca/4/899/ENUSLP21-0094/index.html&request_locale=en
It's the end of an era.
On 3/24/2021 12:49 PM, mkr wrote:
Just noticed this today :-(
https://www-01.ibm.com/common/ssi/ShowDoc.wss?docURL=/common/ssi/rep_ca/4/899/ENUSLP21-0094/index.html&request_locale=en
It's the end of an era.
I had heard about that a little while back (IIRC it was on a British APL
zoom webinar) ...
Worth noting - there is a suggestion that https://log-on.com is
stepping up to provide a 'solution' for current APL2 customers...
No idea how that will go.
On 3/24/2021 2:17 PM, Kerry Liles wrote:
On 3/24/2021 12:49 PM, mkr wrote:
Just noticed this today :-(
https://www-01.ibm.com/common/ssi/ShowDoc.wss?docURL=/common/ssi/rep_ca/4/899/ENUSLP21-0094/index.html&request_locale=en
It's the end of an era.
I had heard about that a little while back (IIRC it was on a British APL
zoom webinar) ...
Worth noting - there is a suggestion that https://log-on.com is
stepping up to provide a 'solution' for current APL2 customers...
No idea how that will go.
funny thing... just received the following email invitation related to >Log-On:
Invitation to a BAA Webinar
25 Mar '21 16:00 GMT
Log-On Software APL an update:
A brief explanation of what else Log-On does
and their immediate plans for Log-On APL2
Conor Hoekstra:
Conor looking for feedback ahead of his presentation
"Algorithms as a Tool of Thought" for APL Seeds on 31 March '21.
Floor will then be open for any other business.
Join the Webinar https://zoom.us/j/858532665
Passcode: 391680
See the Webinar Schedule 2021 for more information.
Join early to meet with whoever else is around. NB this meeting is recorded.
To present at future BAA Webinars please contact >events@britishaplassociation.org .
Kind regards,
The British APL Association
FWIW, it looks like Log-On is now the official provider. I hope this
works out better than eComStation.
Just noticed this today :-(
https://www-01.ibm.com/common/ssi/ShowDoc.wss?docURL=/common/ssi/rep_ca/4/899/ENUSLP21-0094/index.html&request_locale=en
and for another, there's always J.
I remember that there was a discussion elsewhere where the
question "What killed APL?" came up.
The unusual character set is an obvious suspect. If a computer
doesn't have support for the APL character set, that limits the >attractiveness of APL software for that computer.
But if there were strong demand for APL, it would have been easy
enough to provide support for the APL character set, for one thing,
and for another, there's always J.
So I felt that while the unusual character set certainly *helped*,
it wasn't quite enough to do the job by itself. There would also have
to be an alternative around to lessen the demand for APL from those
who might have benefited from it.
BASIC, of course, already reduced the potential demand for APL by
a big amount, as it provided a way to quickly and easily program
computers for answers to at least some problems.
But BASIC was sort of a given - so I felt that something else in addition
to the character set _meant_ something else in addition to the character
set, given BASIC was already around, so something else besides the
two of them was needed to apply the _coup de grace_ to APL.
And I think there were two suspects.
One is the _spreadsheet_, which provides an easy way to handle
several numbers at once without even doing much programming.
Another is _Mathematica_. Here's an interactive programming
environment that does even *more* of what the science and
technology crowd is likely to want than APL could.
There was the IBM 5100. There was the MCM/70. There was the VideoBrain
Family Computer with APL/S, based on the Fairchild F8 microprocessor.
Uh, John, what leads you to believe that APL is dead?
There are three open source versions and three commercial vendors of
whom I am aware and it is still widely used in finance.
Right now the major threat to APL is R and Python which can do
everything APL can do nearly as concisely as APL and can do a lot more besides.
On Wednesday, April 14, 2021 at 11:45:07 PM UTC-6, J. Clarke wrote:
Uh, John, what leads you to believe that APL is dead?
It does depend on what you mean by "dead". It's just under the radar...
There are three open source versions and three commercial vendors of
whom I am aware and it is still widely used in finance.
Right now the major threat to APL is R and Python which can do
everything APL can do nearly as concisely as APL and can do a lot more
besides.
You have a point... but at the moment, APL is sufficiently far under the radar >that while Python is certainly a popular and useful language, I wouldn't think >of it... _in connection_ with APL enough to think of it as being a threat to >APL. After all, there will always be alternative programming languages.
And the _reason_ why languages like R and Python are being invented to >provide some of the same facilities as APL is now no longer because
people are being inspired by APL to do what APL did and more besides...
but because, not having _heard_ of APL, they're independently seeing a need >to do some of the same kinds of things.
I may be exaggerating; it's certainly possible that Guido van Rossum is >knowledgeable enough about the history of computing that he not only
knows what APL is, but he could even talk about MATH-MATIC and
FLOW-MATIC and the Klerer-May system.
Mathematica and the spreadsheet helped to define the climate in the >relatively early days of the personal computer - VisiCalc dating from
the Apple II, and Mathematica from the Macintosh.
IBM included BASIC with the IBM PC; APL was also provided later,
as software that cost extra and which required the 8087 co-processor.
If APL had been "a thing" to the extent of thriving as it did in the 1970s, >it would have had to have been to be expected that keyboards would have
the APL character set on them, and text displays would include the APL >character set as an alternate mode of operation.
There was the IBM 5100. There was the MCM/70. There was the VideoBrain
Family Computer with APL/S, based on the Fairchild F8 microprocessor.
Of course, though, pointing to the spreadsheet, or even APL's character set, >is really looking at secondary causes.
The _primary_ cause is clear; when personal computers *first* came out,
BASIC was the programming language that could have sufficiently small >implementations to be included with them.
Later on, in the age of the Macintosh and Windows 3.1, computers didn't
come with programming languages period, so by the time they were ready
to have Fortran compilers or APL interpreters, personal computers, as
a consumer product, didn't include programming languages.
You could, of course, _get_ programming languages for them, and C was
what was typically used to develop for microcomputers.
You go to the store and buy a computer... it will have a graphical user >interface, which a mini or mainframe from the 1970s would not have
had, but it won't have a Fortran compiler, a COBOL compiler, an assembler,
an APL interpreter... or even a BASIC interpreter.
With Linux, one can indeed turn a personal computer into an approximation
of a mainframe computer of yore (and then there's the mighty Hercules)...
but that's just not what they're _for_.
J Clarke: "Uh, John, what leads you to believe that APL is dead?"
Me: What leads you to believe that APL is NOT dead?
If I needed to use a language to get through a phase, say Actuarial examinations, I would happily use APL.
If I wanted a career in Software Development, I would NOT start by learning APL (no up-to-date references, dearth of worked examples and what exists is hard to find,).
I love APL as the tool that gets the job done WITHOUT the intrusion of usual programming considerations. However, I find MicroAPL & IBM leaving the APL market rather disturbing; it does not bode well for the future of APL.
On Wed, 14 Apr 2021 23:34:22 -0700 (PDT), Quadibloc
<jsa...@ecn.ab.ca> wrote:
With Linux, one can indeed turn a personal computer into an approximation >of a mainframe computer of yore (and then there's the mighty Hercules)... >but that's just not what they're _for_.
What's your point? I have C, C++, BASIC in two flavors, APL, Python,
R, and Powershell on my work machine, plus C, FORTRAN, COBOL and REXX
on the mainframe. The days where a programmer works in one language
for his entire career or even his entire work day are pretty much
over. You use the right tool for the job. Most of the time that's
APL, BASIC, or Python, and to perform a task I may have pieces in all
three of them.
I really wish that people would get over their fixation with
keyboards. NOBODY I know uses a special keyboard. Mine right now is
a Logitech gamer board. The one on my laptop works fine with it. So
does the "official" Raspberry Pi keyboard (and note that APL is free
for personal use on the Pi--so is Mathematica).
And it's 2021--the only text display that is relevant to APL at this
point is the 3270, and it has the APL character set.
Well, for one thing where I work there are more APL developers than I
can find, and management has been trying to kill it for years without success. I understand that this is not atypical in finance.
The hard part is getting people to try it IMO. Once they've used it
for a bit they get hooked.
On Thursday, April 15, 2021 at 10:56:17 AM UTC-6, J. Clarke wrote:
Well, for one thing where I work there are more APL developers than I
can find, and management has been trying to kill it for years without
success. I understand that this is not atypical in finance.
Why would management be trying to kill it, if it's working well?
I'm not saying it isn't working well. Instead, we all know what the
reason is. They're worried that, because there aren't enough shops
using APL - in other sectors besides finance, preferably - one of
these days, the bottom is going to fall out from under vendor
support.
The hard part is getting people to try it IMO. Once they've used it
for a bit they get hooked.
Hey, we're all agreed that APL is a wonderful language.
Now that APL2 is no longer an IBM product, though, APL has lost
its last remaining shred of visibility. Of course it's harder to get
people to try something they've never heard of!
And, strange to relate, there are actually people old enough to
shave these days who don't remember when you usually used
a computer by means of punched cards, but had to go to a 2741
time-sharing terminal for APL.
In fact, there are even adults these days who never lived through the
days when 8-bit microcomputers were a thing. Someone born in
1981, the year IBM announced the IBM Personal Computer, which
transitioned the industry to 16 bits, would be around 40 years old now.
I know it's hard to even imagine the perspective of a 30-year-old,
for whom computers for direct use (as opposed to embedded
processors) always had 32-bit processors and graphical user
interfaces...
There is Tcl/Tk. There was Clarion. While a revival of APL would be
nice, in my opinion, the most single glaring deficiency of today's world
of computers is this:
Back in the days of command-line interfaces, with BASIC,
nearly anyone could learn to write a computer program.
It might take some time before one could write one with a professional >polish, true.
Today, though, the learning curve for writing a Windows application
is a very steep one. You can't really just write a program at
a keyboard - you will also need graphical screen builders.
And the part you do write at the keyboard will involve
insanely complex things like the Microsoft Foundation
Classes.
So no wonder writing programs is left to the large companies that
can hire large teams of programmers to develop them.
On Thursday, April 15, 2021 at 5:17:27 AM UTC-6, J. Clarke wrote:
I really wish that people would get over their fixation with
keyboards. NOBODY I know uses a special keyboard. Mine right now is
a Logitech gamer board. The one on my laptop works fine with it. So
does the "official" Raspberry Pi keyboard (and note that APL is free
for personal use on the Pi--so is Mathematica).
And it's 2021--the only text display that is relevant to APL at this
point is the 3270, and it has the APL character set.
Oh, you mean like this...
http://www.quadibloc.com/comp/images/3270apl2.gif
(from the page
http://www.quadibloc.com/comp/kyb01.htm
)
And in part because they want to be
"cool kids" and "the cool kids" don't use "klunky old languages".
What leads you to believe that every program needs a GUI? Geez, talk
about making mountains out of molehills.
Again with the keyboard. Grok the concept: NO APL USER CARES ABOUT
THE KEYBOARD.
It is the display character set that matters, not the pictures on the keyboard that nobody ever looks at except hunt-and-peck typists.
On Thu, 15 Apr 2021 10:31:58 -0700 (PDT), Quadibloc
<jsa...@ecn.ab.ca> wrote:
Now that APL2 is no longer an IBM product, though, APL has lost
its last remaining shred of visibility. Of course it's harder to get
people to try something they've never heard of!
Hardly. APL2 hasn't been particularly visible for a long time. Dyalog
is the big player.
On Thursday, April 15, 2021 at 1:32:22 PM UTC-6, J. Clarke wrote:
Again with the keyboard. Grok the concept: NO APL USER CARES ABOUT
THE KEYBOARD.
It is the display character set that matters, not the pictures on the
keyboard that nobody ever looks at except hunt-and-peck typists.
1) To be able to touch-type in _APL_ would require that one has had quite
a bit of experience in it, and uses it a lot.
2) For characters to reach the display, the keyboard layout must
generate them, whatever may be printed on the keys.
3) But the keyboard is more a consequence than a cause. Back when
APL was actually popular, there was a version of the DECwriter,
there was a version of the Tektronix 4010, and so on and so forth,
with an APL keyboard. If you don't see APL characters on the keyboards
of computers today, that's a sign it isn't being used as much.
On Thursday, April 15, 2021 at 1:50:08 PM UTC-6, J. Clarke wrote:
On Thu, 15 Apr 2021 10:31:58 -0700 (PDT), Quadibloc
<jsa...@ecn.ab.ca> wrote:
Now that APL2 is no longer an IBM product, though, APL has lost
its last remaining shred of visibility. Of course it's harder to get
people to try something they've never heard of!
Hardly. APL2 hasn't been particularly visible for a long time. Dyalog
is the big player.
That may be.
But surely you can see the problem.
Dyalog may be the foremost company in the APL field. But it isn't Microsoft, >it isn't Apple, and it isn't IBM. Nor is it Google or even Samsung., Intel, or Nvidia.
Or AMD.
Still, you may be quite right that I am... exaggerating. There's a big difference
between "dead" and "not totally in your face". Of course APL can survive for >decades more, languishing in relative obscurity.
On Thursday, April 15, 2021 at 1:50:08 PM UTC-6, J. Clarke wrote:
And in part because they want to be
"cool kids" and "the cool kids" don't use "klunky old languages".
To phrase this in a way which makes visible the _rational_ objection: it
will be easier to recruit staff if the position provides experience in >languages that are currently in demand.
What leads you to believe that every program needs a GUI? Geez, talk
about making mountains out of molehills.
It's surprising to me that the point I was making is not clear and
obvious.
Back in the command-line era...
Command-line programs could be written in a simple and natural
manner by anyone who was trained in a compiled language.
Thus, if you were a programmer, you could write applications
programs.
Today, though, with the prevalence of the GUI - *and* the way
operating systems are designed to support the GUI - there is no
longer a simple path from "learn how to program" and "write an
application that at least approaches those which are commercially
sold and distributed".
Now the tools used to build applications are much more complex
than a compiler for a programming language.
Of course, though, what with APL not normally producing
compiled executables, this is not so much of an issue for APL
specifically.
I have two APL keyboards. I don't use either because the glyphs
marked on the keys do not match the mappings for the version of APL
that I use, and I find them more confusing than helpful.
On Fri, 16 Apr 2021 07:25:52 -0700 (PDT), Quadibloc
<jsa...@ecn.ab.ca> wrote:
On Thursday, April 15, 2021 at 1:50:08 PM UTC-6, J. Clarke wrote:
Dyalog may be the foremost company in the APL field. But it isn't Microsoft, >it isn't Apple, and it isn't IBM. Nor is it Google or even Samsung., Intel, or Nvidia.
Or AMD.
Geez, Ford is no Tesla. So what?
On Thu, 15 Apr 2021 19:47:39 -0700 (PDT), Quadibloc
<jsa...@ecn.ab.ca> wrote:
Of course, though, what with APL not normally producing
compiled executables, this is not so much of an issue for APL
specifically.
Once again you seem to be falling into the IT ghetto. I need numbers crunched. I don't need a pretty GUI.
That's like saying "if you don't see Chinese characters on the
keyboards of computers today, that's a sign it isn't being used as
much".
On Friday, April 16, 2021 at 9:16:50 AM UTC-6, J. Clarke wrote:
On Thu, 15 Apr 2021 19:47:39 -0700 (PDT), Quadibloc
<jsa...@ecn.ab.ca> wrote:
Of course, though, what with APL not normally producing
compiled executables, this is not so much of an issue for APL
specifically.
Once again you seem to be falling into the IT ghetto. I need numbers
crunched. I don't need a pretty GUI.
And people who need numbers crunched have plenty of tools... spreadsheets, >Mathematica, and, yes, even Python. APL now has more competition than it
used to have in this area, which has forced it into a narrower niche.
On Friday, April 16, 2021 at 9:11:41 AM UTC-6, J. Clarke wrote:
I have two APL keyboards. I don't use either because the glyphs
marked on the keys do not match the mappings for the version of APL
that I use, and I find them more confusing than helpful.
And you don't think that's a sign that APL is declining in popularity?
On Friday, April 16, 2021 at 9:19:31 AM UTC-6, J. Clarke wrote:
On Fri, 16 Apr 2021 07:25:52 -0700 (PDT), Quadibloc
<jsa...@ecn.ab.ca> wrote:
On Thursday, April 15, 2021 at 1:50:08 PM UTC-6, J. Clarke wrote:
Dyalog may be the foremost company in the APL field. But it isn't Microsoft,
it isn't Apple, and it isn't IBM. Nor is it Google or even Samsung., Intel, or Nvidia.
Or AMD.
Geez, Ford is no Tesla. So what?
I would have put it that Tesla is no Ford (or GM or AMC... unless AMC isn't still
around). So perhaps part of my problem is that I'm a dinosaur...
Sysop: | Keyop |
---|---|
Location: | Huddersfield, West Yorkshire, UK |
Users: | 296 |
Nodes: | 16 (2 / 14) |
Uptime: | 60:32:34 |
Calls: | 6,654 |
Calls today: | 6 |
Files: | 12,200 |
Messages: | 5,331,394 |