On Mon, 8 Apr 2024 14:46:32 -0000 (UTC), Ben Collver wrote:

> On 2024-04-08, Lawrence D'Oliveiro <ldo@nz.invalid> wrote:
>
>> You’ve got FFmpeg in there, the universal multimedia “sonic
>> screwdriver”, that should already be able to do nearly all the
>> necessary conversions/encoding.
>
> I wanted to produce a video showing the tracker interface as it scrolled
> through the score while playing the song, using the same tracker that
> was used to compose the song.

I suspect it would have been less effort to just fake it.

Lawrence D'Oliveiro <ldo@nz.invalid> wrote:
> On Mon, 8 Apr 2024 13:31:33 -0000 (UTC), Rich wrote:
>
>> Lawrence D'Oliveiro <ldo@nz.invalid> wrote:
>>>
>>> That is inherent in the nature of the problem it is trying to solve:
>>> automate an interface--the GUI--which was never designed for automatic
>>> operation. No matter how you slice and dice it, that kind of thing
>>> inherently turns out to be fiddly and fragile.
>>
>> But that part wasn't the problem. AutoHotKey included function calls to
>> do that, and once one found the function call to perform "step x" that
>> one wanted, it worked.
>
> In my experience, that is very much part of the problem. I have done this
> sort of thing in the past--thankfully, not lately.

Yes, that can be problematic.

> E.g. bring up a dialog box to enter a number. I have a template for the
> event sequence, to do the initial menu selection, then have my code insert
> the (variable) sequence of keystrokes for the number, then end with the
> sequence for clicking the OK button.
>
> Only if you do it a little too fast, the text field or the OK button might
> not actually be enabled yet, and so some part of the sequence gets ignored
> and the whole thing goes badly wrong. So you stick in a tenth-of-a-second
> delay somewhere, entirely through trial-and-error, until it seems to work.

None of which is AutoHotKey's fault. All of which is directly caused
by the GUI not ever having been intended to ever be automated in any
way.

> Then later on somebody reports that it sometimes fails in a different
> place. But of course it works for you. So you then have to try the
> sequence multiple times, until you see it fail that one time, and then you
> have to figure out why and come up with another tweak to the sequence to
> fix it.
>
> All this take programming skills--which ordinary users don’t have.

Yup. My complaint, however, was with the syntax irregularities of
the AutoHotKey script language. The actual language syntax looks like
it was designed by a group of drunken monkey's after a night of bar
hopping, with an LSD trip added at the end for good measure. No two
functions used similar calling conventions, and for those that required
some of the same parameters it looked like the language designer
intentionally randomized the order of the parameters for each function
call.

> So who are these macro-recorder thingies aimed at?

I suspect their aim might be twofold:

1) folks who want to use it to change keystrokes (i.e., make their
keyboard be Dovrak but their OS does not have a Dovrak layout); or

2) folks who literally just want to "record" a macro sequence, without
any conditionals and no error checking of any form.

The actual rest that provides a seemingly Turing complete language
looks to have been randomly bolted on using parts from plural
different bodies. It felt like a language that took the worst parts of
Perl, Javascript, Java, Python, C, COBOL, Ada, etc. and combined it all
into a new ugly monster.

> They seem designed to pretend that there is no programming involved
> in using them, but on the other hand it can take a lot of work to use
> them reliably. And so you end up with the worst of both worlds.

That indeed. Maybe their 'recording' part inserts delays equal to what
time it took the user to move/click through things, which would, for
the most part, create an appearance of working most of the time.

On Tue, 9 Apr 2024 00:59:10 -0000 (UTC), Rich wrote:

> None of which is AutoHotKey's fault.

My point exactly.

> All of which is directly caused by
> the GUI not ever having been intended to ever be automated in any way.

My point exactly.

> My complaint, however, was with the syntax irregularities of the
> AutoHotKey script language.

OK, so that looks weird as well. My double condolences. ;)

> Maybe their 'recording' part inserts delays equal to what
> time it took the user to move/click through things, which would, for the
> most part, create an appearance of working most of the time.

That may very well work, if you are repeating a set of GUI actions
exactly. But how often would you, for example, perform the same set of
edits on the same file? You will more likely want to repeat the actions on
a set of files, one after the other. And like it or not, that gets into
actual programming, with variables and loops and all the rest of it.

And this programming is trying to automate something (a GUI) which was
never designed to be automated.

Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 00:30 this Tuesday (GMT):
> On Mon, 8 Apr 2024 14:30:06 -0000 (UTC), candycanearter07 wrote:
>
>> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 06:49 this Monday (GMT):
>>
>>> On Mon, 8 Apr 2024 03:20:02 -0000 (UTC), candycanearter07 wrote:
>>>
>>>> The problem happens when there's no way to access the underlying
>>>> functions, or it requires bad workarounds like the GUI scripting.
>>>
>>> That kind of thing seems less common in the Open Source world. Perhaps
>>> because it is more likely to be seen as a bug and fixed.
>>
>> Of course, the opposite is usually more true. A lot of stuff is
>> terminal-only.
>
> Feel free to offer examples.

Installing things?
--
user <candycane> is generated from /dev/urandom

On Tue, 9 Apr 2024 15:10:10 -0000 (UTC), candycanearter07 wrote:

> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 00:30 this Tuesday (GMT):
>> On Mon, 8 Apr 2024 14:30:06 -0000 (UTC), candycanearter07 wrote:
>>
>>> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 06:49 this Monday (GMT):
>>>
>>>> On Mon, 8 Apr 2024 03:20:02 -0000 (UTC), candycanearter07 wrote:
>>>>
>>>>> The problem happens when there's no way to access the underlying
>>>>> functions, or it requires bad workarounds like the GUI scripting.
>>>>
>>>> That kind of thing seems less common in the Open Source world.
>>>> Perhaps because it is more likely to be seen as a bug and fixed.
>>>
>>> Of course, the opposite is usually more true. A lot of stuff is
>>> terminal-only.
>>
>> Feel free to offer examples.
>
> Installing things?

Scriptable.

Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 05:39 this Wednesday (GMT):
> On Tue, 9 Apr 2024 15:10:10 -0000 (UTC), candycanearter07 wrote:
>
>> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 00:30 this Tuesday (GMT):
>>> On Mon, 8 Apr 2024 14:30:06 -0000 (UTC), candycanearter07 wrote:
>>>
>>>> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 06:49 this Monday (GMT):
>>>>
>>>>> On Mon, 8 Apr 2024 03:20:02 -0000 (UTC), candycanearter07 wrote:
>>>>>
>>>>>> The problem happens when there's no way to access the underlying
>>>>>> functions, or it requires bad workarounds like the GUI scripting.
>>>>>
>>>>> That kind of thing seems less common in the Open Source world.
>>>>> Perhaps because it is more likely to be seen as a bug and fixed.
>>>>
>>>> Of course, the opposite is usually more true. A lot of stuff is
>>>> terminal-only.
>>>
>>> Feel free to offer examples.
>>
>> Installing things?
>
> Scriptable.

Yes? That was my point, that more options are available in the
terminal.
--
user <candycane> is generated from /dev/urandom

On Wed, 10 Apr 2024 14:40:05 -0000 (UTC), candycanearter07 wrote:

> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 05:39 this Wednesday
> (GMT):
>> On Tue, 9 Apr 2024 15:10:10 -0000 (UTC), candycanearter07 wrote:
>>
>>> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 00:30 this Tuesday
>>> (GMT):
>>>> On Mon, 8 Apr 2024 14:30:06 -0000 (UTC), candycanearter07 wrote:
>>>>
>>>>> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 06:49 this Monday
>>>>> (GMT):
>>>>>
>>>>>> On Mon, 8 Apr 2024 03:20:02 -0000 (UTC), candycanearter07 wrote:
>>>>>>
>>>>>>> The problem happens when there's no way to access the underlying
>>>>>>> functions, or it requires bad workarounds like the GUI scripting.
>>>>>>
>>>>>> That kind of thing seems less common in the Open Source world.
>>>>>> Perhaps because it is more likely to be seen as a bug and fixed.
>>>>>
>>>>> Of course, the opposite is usually more true. A lot of stuff is
>>>>> terminal-only.
>>>>
>>>> Feel free to offer examples.
>>>
>>> Installing things?
>>
>> Scriptable.
>
> Yes? That was my point, that more options are available in the terminal.

No, you said “the opposite is usually more true”, in reply to my saying
that lack of scriptability is “more likely to be seen as a bug and fixed”.

Lawrence D'Oliveiro <ldo@nz.invalid> wrote:
>On Sun, 7 Apr 2024 16:51:22 -0000 (UTC), Rich wrote:
>
>> AutoHotKey is the most absymal "programming" languge ...
>
>That is inherent in the nature of the problem it is trying to solve:
>automate an interface--the GUI--which was never designed for automatic
>operation. No matter how you slice and dice it, that kind of thing
>inherently turns out to be fiddly and fragile.

Apple did a pretty good job of this with AppleScript by building the GUI
with automation in mind. That fell by the wayside long ago, however.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 03:07 this Thursday (GMT):
> On Wed, 10 Apr 2024 14:40:05 -0000 (UTC), candycanearter07 wrote:
>
>> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 05:39 this Wednesday
>> (GMT):
>>> On Tue, 9 Apr 2024 15:10:10 -0000 (UTC), candycanearter07 wrote:
>>>
>>>> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 00:30 this Tuesday
>>>> (GMT):
>>>>> On Mon, 8 Apr 2024 14:30:06 -0000 (UTC), candycanearter07 wrote:
>>>>>
>>>>>> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 06:49 this Monday
>>>>>> (GMT):
>>>>>>
>>>>>>> On Mon, 8 Apr 2024 03:20:02 -0000 (UTC), candycanearter07 wrote:
>>>>>>>
>>>>>>>> The problem happens when there's no way to access the underlying
>>>>>>>> functions, or it requires bad workarounds like the GUI scripting.
>>>>>>>
>>>>>>> That kind of thing seems less common in the Open Source world.
>>>>>>> Perhaps because it is more likely to be seen as a bug and fixed.
>>>>>>
>>>>>> Of course, the opposite is usually more true. A lot of stuff is
>>>>>> terminal-only.
>>>>>
>>>>> Feel free to offer examples.
>>>>
>>>> Installing things?
>>>
>>> Scriptable.
>>
>> Yes? That was my point, that more options are available in the terminal.
>
> No, you said “the opposite is usually more true”, in reply to my saying
> that lack of scriptability is “more likely to be seen as a bug and fixed”.

Oh, sorry.
--
user <candycane> is generated from /dev/urandom

On 11 Apr 2024 14:30:07 -0000, Scott Dorsey wrote:

> Lawrence D'Oliveiro <ldo@nz.invalid> wrote:
>
>>On Sun, 7 Apr 2024 16:51:22 -0000 (UTC), Rich wrote:
>>
>>> AutoHotKey is the most absymal "programming" languge ...
>>
>>That is inherent in the nature of the problem it is trying to solve:
>>automate an interface--the GUI--which was never designed for automatic
>>operation. No matter how you slice and dice it, that kind of thing
>>inherently turns out to be fiddly and fragile.
>
> Apple did a pretty good job of this with AppleScript by building the GUI
> with automation in mind. That fell by the wayside long ago, however.

As someone who was a champion of AppleScript back in the day, made a point
of supporting it in my own apps, and even made a living creating scripts
for others, let me just say that, in retrospect, it was the solution to
the wrong problem.

In article <slrnv15e4i.rk6.bencollver@svadhyaya.localdomain>,
Ben Collver <bencollver@tilde.pink> wrote:
>Is Programming Obsolete?
>========================
>Brian Harvey
>University of California, Berkeley

If you don't teach basic programming, you can't teach students what a
computer actually is.

If you don't teach assembler, you can't teach students how a computer
actually works.

You don't need to teach a fancy programming language or a giant assembler.
You don't even need to teach ones that are used in the real world because
the point isn't to give the student a marketable skill but to give him the
background where he can understand what a computer is and what it does.

You will find people in the IT community don't think teaching any of this
stuff is important, in part because they don't know what a computer is or
what it does.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

kludge@panix.com (Scott Dorsey) wrote in news:v1qem3$g2t$1@panix2.panix.com:

> In article <slrnv15e4i.rk6.bencollver@svadhyaya.localdomain>,
> Ben Collver <bencollver@tilde.pink> wrote:
>>Is Programming Obsolete?
>>========================
>>Brian Harvey
>>University of California, Berkeley
>
> If you don't teach basic programming, you can't teach students what a
> computer actually is.
>
> If you don't teach assembler, you can't teach students how a computer
> actually works.
>
> You don't need to teach a fancy programming language or a giant assembler.
> You don't even need to teach ones that are used in the real world because
> the point isn't to give the student a marketable skill but to give him the
> background where he can understand what a computer is and what it does.
>
> You will find people in the IT community don't think teaching any of this
> stuff is important, in part because they don't know what a computer is or
> what it does.
> --scott

Sad, but true.

AI is no substitute for being able to reason and express your own solutions.

Worse, recent 'college' grads don't want to do work and will take shortcuts
that create possibly hidden problems that an employer was willing to pay them
to use their knowlege to create a correct solution.

On Sun, 12 May 2024, Scott Dorsey wrote:

> In article <slrnv15e4i.rk6.bencollver@svadhyaya.localdomain>,
> Ben Collver <bencollver@tilde.pink> wrote:
>> Is Programming Obsolete?
>> ========================
>> Brian Harvey
>> University of California, Berkeley
>
> If you don't teach basic programming, you can't teach students what a
> computer actually is.
>
> If you don't teach assembler, you can't teach students how a computer
> actually works.
>
> You don't need to teach a fancy programming language or a giant assembler.
> You don't even need to teach ones that are used in the real world because
> the point isn't to give the student a marketable skill but to give him the
> background where he can understand what a computer is and what it does.
>
> You will find people in the IT community don't think teaching any of this
> stuff is important, in part because they don't know what a computer is or
> what it does.
> --scott
>

Let me add another perspective. Companies, influecing course programs,
don't want to waste time on fundamentals. They want programmers who can do
one nieche thing fast, as fast as possible, so no time to focus on the
basics.

D <nospam@example.net> wrote:
>
>Let me add another perspective. Companies, influecing course programs,
>don't want to waste time on fundamentals. They want programmers who can do
>one nieche thing fast, as fast as possible, so no time to focus on the
>basics.

And this makes sense if you intend to only hire them for a couple years
and then dump them. If you're doing that, then hiring people who can
learn new systems quickly doesn't actually help you. Unfortunately this
policy has other problems but is very popular.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

On Sun, 12 May 2024 13:02:59 +0000, Scott Dorsey wrote:

> In article <slrnv15e4i.rk6.bencollver@svadhyaya.localdomain>,
> Ben Collver <bencollver@tilde.pink> wrote:
>>Is Programming Obsolete? ========================
>>Brian Harvey University of California, Berkeley
>
> If you don't teach basic programming, you can't teach students what a
> computer actually is.
>
> If you don't teach assembler, you can't teach students how a computer
> actually works.
>
> You don't need to teach a fancy programming language or a giant
> assembler.

Indeed. When I was doing my Master's, I saw undergraduates being taught
using a 'paper computer'. 16 words of RAM, 12 bit word. It was a set of
boxes with index cards with contents written on them. Plus a couple of
boxes for accumulator and flags. It was ctually a PDP-8 clone, but only
for simplicity.

--
Using UNIX since v6 (1975)...

Use the BIG mirror service in the UK:
http://www.mirrorservice.org

On Sun, 12 May 2024, Scott Dorsey wrote:

> D <nospam@example.net> wrote:
>>
>> Let me add another perspective. Companies, influecing course programs,
>> don't want to waste time on fundamentals. They want programmers who can do
>> one nieche thing fast, as fast as possible, so no time to focus on the
>> basics.
>
> And this makes sense if you intend to only hire them for a couple years
> and then dump them. If you're doing that, then hiring people who can
> learn new systems quickly doesn't actually help you. Unfortunately this
> policy has other problems but is very popular.
> --scott
>

Spot on scott. That is the game plan as far as I can tell, sadly.

kludge@panix.com (Scott Dorsey) writes:

> In article <slrnv15e4i.rk6.bencollver@svadhyaya.localdomain>,
> Ben Collver <bencollver@tilde.pink> wrote:
>>Is Programming Obsolete?
>>========================
>>Brian Harvey
>>University of California, Berkeley
>
> If you don't teach basic programming, you can't teach students what a
> computer actually is.
>
> If you don't teach assembler, you can't teach students how a computer
> actually works.
>
> You don't need to teach a fancy programming language or a giant assembler.
> You don't even need to teach ones that are used in the real world because
> the point isn't to give the student a marketable skill but to give him the
> background where he can understand what a computer is and what it does.
>
> You will find people in the IT community don't think teaching any of this
> stuff is important, in part because they don't know what a computer is or
> what it does.

Precisely.

D <nospam@example.net> writes:

> On Sun, 12 May 2024, Scott Dorsey wrote:
>
>> In article <slrnv15e4i.rk6.bencollver@svadhyaya.localdomain>,
>> Ben Collver <bencollver@tilde.pink> wrote:
>>> Is Programming Obsolete?
>>> ========================
>>> Brian Harvey
>>> University of California, Berkeley
>>
>> If you don't teach basic programming, you can't teach students what a
>> computer actually is.
>>
>> If you don't teach assembler, you can't teach students how a computer
>> actually works.
>>
>> You don't need to teach a fancy programming language or a giant assembler.
>> You don't even need to teach ones that are used in the real world because
>> the point isn't to give the student a marketable skill but to give him the
>> background where he can understand what a computer is and what it does.
>>
>> You will find people in the IT community don't think teaching any of this
>> stuff is important, in part because they don't know what a computer is or
>> what it does.
>> --scott
>
> Let me add another perspective. Companies, influecing course programs,
> don't want to waste time on fundamentals. They want programmers who
> can do one nieche thing fast, as fast as possible, so no time to focus
> on the basics.

In other words, bring me profits, forget yourself and the the world.

Bob Eager <news0009@eager.cx> wrote:
>
>Indeed. When I was doing my Master's, I saw undergraduates being taught
>using a 'paper computer'. 16 words of RAM, 12 bit word. It was a set of
>boxes with index cards with contents written on them. Plus a couple of
>boxes for accumulator and flags. It was ctually a PDP-8 clone, but only
>for simplicity.

Bell Labs used to sell a paper card thing called the CARDIAC which was
vaguely like this, and the US Army training folks had a thing called TAGSAC
which was a paper computer for teaching programming in the days when
computers were too expensive to waste on students.

But these days you can buy an ST evaluation board that plugs into your PC
and has a marvelous 8-bit instruction set. It even has a divide! Take that,
8051.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

On Mon, 13 May 2024 13:10:35 +0000, Scott Dorsey wrote:

> Bob Eager <news0009@eager.cx> wrote:
>>
>>Indeed. When I was doing my Master's, I saw undergraduates being taught
>>using a 'paper computer'. 16 words of RAM, 12 bit word. It was a set of
>>boxes with index cards with contents written on them. Plus a couple of
>>boxes for accumulator and flags. It was ctually a PDP-8 clone, but only
>>for simplicity.
>
> Bell Labs used to sell a paper card thing called the CARDIAC which was
> vaguely like this, and the US Army training folks had a thing called
> TAGSAC which was a paper computer for teaching programming in the days
> when computers were too expensive to waste on students.
>
> But these days you can buy an ST evaluation board that plugs into your
> PC and has a marvelous 8-bit instruction set. It even has a divide!
> Take that,
> 8051.

The great thing about the one with index cards is that it came out of the
stationery budget!

--
Using UNIX since v6 (1975)...

Use the BIG mirror service in the UK:
http://www.mirrorservice.org

Bob Eager:

> Indeed. When I was doing my Master's, I saw undergraduates
> being taught using a 'paper computer'. 16 words of RAM, 12
> bit word. It was a set of boxes with index cards with
> contents written on them. Plus a couple of boxes for
> accumulator and flags. It was ctually a PDP-8 clone, but
> only for simplicity.

This reminds of a great site absolutely on-topic here:

<https://wiki.xxiivv.com/site/research.html>
and especially:
<https://wiki.xxiivv.com/site/paper_computing.html>

--
() ascii ribbon campaign -- against html e-mail
/\ www.asciiribbon.org -- against proprietary attachments

Anton Shepelev <anton.txt@g{oogle}mail.com> wrote at 17:31 this Monday (GMT):
> Bob Eager:
>
>> Indeed. When I was doing my Master's, I saw undergraduates
>> being taught using a 'paper computer'. 16 words of RAM, 12
>> bit word. It was a set of boxes with index cards with
>> contents written on them. Plus a couple of boxes for
>> accumulator and flags. It was ctually a PDP-8 clone, but
>> only for simplicity.
>
> This reminds of a great site absolutely on-topic here:
>
> <https://wiki.xxiivv.com/site/research.html>
> and especially:
> <https://wiki.xxiivv.com/site/paper_computing.html>

Interesting, I'd never heard of that.
--
user <candycane> is generated from /dev/urandom