Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.javascript > #37752 > unrolled thread

[OT] Object orientation: is this still a thing?

Back to article view | Back to comp.lang.javascript

Contents

  [OT] Object orientation: is this still a thing? Julio Di Egidio <julio@diegidio.name> - 2019-11-11 08:51 -0800
    Re: [OT] Object orientation: is this still a thing? "Michael Haufe (TNO)" <tno@thenewobjective.com> - 2019-11-11 10:54 -0800
      Re: [OT] Object orientation: is this still a thing? Julio Di Egidio <julio@diegidio.name> - 2019-11-12 07:09 -0800
        Re: [OT] Object orientation: is this still a thing? "Michael Haufe (TNO)" <tno@thenewobjective.com> - 2019-11-12 12:30 -0800
          Re: [OT] Object orientation: is this still a thing? Thomas 'PointedEars' Lahn <PointedEars@web.de> - 2019-11-14 19:29 +0100
            Re: [OT] Object orientation: is this still a thing? Julio Di Egidio <julio@diegidio.name> - 2019-11-14 10:40 -0800
            Re: [OT] Object orientation: is this still a thing? "Michael Haufe (TNO)" <tno@thenewobjective.com> - 2019-11-15 11:09 -0800
              Re: [OT] Object orientation: is this still a thing? Thomas 'PointedEars' Lahn <PointedEars@web.de> - 2019-11-16 04:05 +0100
          Re: [OT] Object orientation: is this still a thing? Julio Di Egidio <julio@diegidio.name> - 2019-11-13 05:45 -0800
            Re: [OT] Object orientation: is this still a thing? Julio Di Egidio <julio@diegidio.name> - 2019-11-13 05:53 -0800
              Re: [OT] Object orientation: is this still a thing? "Michael Haufe (TNO)" <tno@thenewobjective.com> - 2019-11-13 10:44 -0800
            Re: [OT] Object orientation: is this still a thing? "Michael Haufe (TNO)" <tno@thenewobjective.com> - 2019-11-13 10:42 -0800
      Re: [OT] Object orientation: is this still a thing? "Michael Haufe (TNO)" <tno@thenewobjective.com> - 2019-11-14 08:18 -0800
      Re: [OT] Object orientation: is this still a thing? Julio Di Egidio <julio@diegidio.name> - 2019-11-14 09:44 -0800
        Re: [OT] Object orientation: is this still a thing? "Michael Haufe (TNO)" <tno@thenewobjective.com> - 2019-11-15 11:16 -0800
          Re: [OT] Object orientation: is this still a thing? John G Harris <niam@jghnorth.org.uk.invalid> - 2019-11-17 15:13 +0000
      Re: [OT] Object orientation: is this still a thing? John G Harris <niam@jghnorth.org.uk.invalid> - 2019-11-13 18:46 +0000

#37752 — [OT] Object orientation: is this still a thing?

On Saturday, 9 November 2019 14:39:13 UTC+1, Julio Di Egidio  wrote:
[<https://groups.google.com/d/msg/comp.lang.javascript/mJ2OT-alxrg/ponJu3CvAwAJ>]
> On Saturday, 9 November 2019 12:55:22 UTC+1, John G Harris  wrote:
> > On Sat, 9 Nov 2019 02:02:47 -0800 (PST), Julio Di Egidio
> > <julio.diegidio.name> wrote:
> >   <snip>
> > >For the chronicle, OOP uber alles is a plain fraud, 
> >   <snip>
> > 
> > Is it OO *Programming* or OO *Design* that you criticise? Certainly
> > OOD suffers from a lot of hype and extremism. People who don't really
> > know what they're doing are encouraged to go mad and describe
> > everything as an object.
> 
> Sorry, just a slip, I meant the infamous OO uber alles, which of course is
> about the whole paradigm, not about this or that particular construct (and
> you forget analysis, OOA: and who could blame you, since analysis has been
> literally and purposely obliterated from the panorama top begin with, in
> order to open the doors to that ungodly nonsense called vendor-driven
> development? Of software, courses certificates, books, conferences, and
> what not, including *necessarily* broken development paradigms...).

"Political" issues aside, from a strictly technical point of view I would
maintain that object oriented programming (meaning the whole idea) in its
essence is *misplaced meta-programming*!

Thoughts?

Julio

[toc] | [next] | [standalone]

#37753

On Monday, November 11, 2019 at 10:51:41 AM UTC-6, Julio Di Egidio wrote:
> 
> "Political" issues aside, from a strictly technical point of view I would
> maintain that object oriented programming (meaning the whole idea) in its
> essence is *misplaced meta-programming*!
> 
> Thoughts?

I would say that BOTH the Functional and Object-Oriented language paradigms are obsolete. In regards to OOAD as a whole, I wouldn't want to make the same claims and I think it should be discussed separately. 

Due to the Expression Problem[1] both FP and OOP languages have evolved a common pattern:

In FP the pattern is called "Final Tagless" [2]
In OOP the pattern is called "Object Algebra" [3]

These are used as a possible foundation of a paradigm called "Feature Oriented Programming" [4]. There are only toy languages for this at the moment AFAIK including one I am developing.

I can expand on this topic more, but I'd prefer to stay aligned with terminology at this point and try to draw out more from you on what you mean when you say "misplaced meta-programming"

[1] <http://homepages.inf.ed.ac.uk/wadler/papers/expression/expression.txt>
Not mentioned in the original description, but an additional requirement is not need whole program analysis for a solution. 
[2] <http://okmij.org/ftp/tagless-final/JFP.pdf>
[3] <https://www.cs.utexas.edu/~wcook/Drafts/2012/ecoop2012.pdf>
[4] <https://en.wikipedia.org/wiki/Feature-oriented_programming>

[toc] | [prev] | [next] | [standalone]

#37766

On Monday, 11 November 2019 19:54:48 UTC+1, Michael Haufe (TNO)  wrote:
> On Monday, November 11, 2019 at 10:51:41 AM UTC-6, Julio Di Egidio wrote:
> > 
> > "Political" issues aside, from a strictly technical point of view I would
> > maintain that object oriented programming (meaning the whole idea) in its
> > essence is *misplaced meta-programming*!
> > 
> > Thoughts?
> 
> I would say that BOTH the Functional and Object-Oriented language paradigms
> are obsolete

I would say that they both keep evolving: but I'll check your links, thanks...

> In regards to OOAD as a whole, I wouldn't want to make the same claims and I
> think it should be discussed separately.

But that is totally integral to what I am talking about: it's your perspective
strictly on the details of a formal language that is not even in question...

BTW, "meta-programming" as in... meta-programming: do you mean you don't know
what that is, or you just annoyed that I write down two liners instead of just
posting links?  ;)

Julio

[toc] | [prev] | [next] | [standalone]

#37773

On Tuesday, November 12, 2019 at 9:09:06 AM UTC-6, Julio Di Egidio wrote:
> On Monday, 11 November 2019 19:54:48 UTC+1, Michael Haufe (TNO)  wrote:
>  [...]
> > I would say that BOTH the Functional and Object-Oriented language paradigms
> > are obsolete
> 
> I would say that they both keep evolving: but I'll check your links, thanks...

I could expand on this in detail but it would get... verbose.
Probably better as a unified blog post for presentation though
I suspect a couple already exist on the interwebs. If someone
has a specific gap in their knowledge here I'll try to elucidate
as needed.

> > In regards to OOAD as a whole, I wouldn't want to make the same
> > claims and I think it should be discussed separately.
> 
> But that is totally integral to what I am talking about: it's your perspective
> strictly on the details of a formal language that is not even in question...

If we can't agree on a subset of it, what good would it be to talk about 
the whole? I suspect it would just devolve into over-generalizations.

> BTW, "meta-programming" as in... meta-programming

Tautologies are true but useless. 

> : do you mean you don't know what that is, or you just
> annoyed that I write down two liners instead of just
> posting links?  ;)

Annoyed that you didn't post links. I won't assume what your
definition is in this context. What does "meta-programming"
mean in terms of a full-cycle methodology like OOAD?

Which "meta-programming" approach? 

A Meta-Object Protocol?
Scheme style macros?
C++ style static templates?
other?

As paraphrased by Keean Schupke on Wittgenstein: "philosophy only exists due to the confusion of meaning caused when the same word is used to mean different things in more than one language-game."

So let's not play games.

[toc] | [prev] | [next] | [standalone]

#37780

Michael Haufe (TNO) wrote:

> On Wednesday, November 13, 2019 at 7:45:59 AM UTC-6, Julio Di Egidio
> wrote:
>> On Tuesday, 12 November 2019 21:30:24 UTC+1, Michael Haufe (TNO)  wrote:
>> > I could expand on this in detail but it would get... verbose.
>> You should learn to read first, as I have just warned you that is NOT in
>> question.
> 
> This isn't only about you...

Why are you still feeding this troll?  Don’t you have better things to do?
 
-- 
PointedEars
FAQ: <http://PointedEars.de/faq> | <http://PointedEars.de/es-matrix>
<https://github.com/PointedEars> | <http://PointedEars.de/wsvn/>
Twitter: @PointedEars2 | Please do not cc me./Bitte keine Kopien per E-Mail.

[toc] | [prev] | [next] | [standalone]

#37781

On Thursday, 14 November 2019 19:29:17 UTC+1, Thomas 'PointedEars' Lahn  wrote:
> Michael Haufe (TNO) wrote:
> > On Wednesday, November 13, 2019 at 7:45:59 AM UTC-6, Julio Di Egidio
> > wrote:
> >> On Tuesday, 12 November 2019 21:30:24 UTC+1, Michael Haufe (TNO)  wrote:
> >> > I could expand on this in detail but it would get... verbose.
> >> You should learn to read first, as I have just warned you that is NOT in
> >> question.
> > 
> > This isn't only about you...
> 
> Why are you still feeding this troll?

The abusive fraudulent whiner is back, what fun...

Please do not feed that troll.

*Plonk*

Julio

[toc] | [prev] | [next] | [standalone]

#37793

On Thursday, November 14, 2019 at 12:29:17 PM UTC-6, Thomas 'PointedEars' Lahn wrote:

> Why are you still feeding this troll?  Don’t you have better things to do?

- I have a long commute for the near-term
- The topic is relevant for something I'm working on
- Why do you care?
- He's engaged in good faith thus far
- I think every person who engages in this group is a masochist

[toc] | [prev] | [next] | [standalone]

#37795

Michael Haufe (TNO) wrote:

> On Thursday, November 14, 2019 at 12:29:17 PM UTC-6, Thomas 'PointedEars'
> Lahn wrote:
>> Why are you still feeding this troll?  Don’t you have better things to
>> do?
> 
> - I have a long commute for the near-term
> - The topic is relevant for something I'm working on

ACK.

> - Why do you care?

You already know.

> - He's engaged in good faith thus far

ROTFL.

> - I think every person who engages in this group is a masochist

Apparently.

-- 
PointedEars
FAQ: <http://PointedEars.de/faq> | <http://PointedEars.de/es-matrix>
<https://github.com/PointedEars> | <http://PointedEars.de/wsvn/>
Twitter: @PointedEars2 | Please do not cc me./Bitte keine Kopien per E-Mail.

[toc] | [prev] | [next] | [standalone]

#37897

On Tuesday, 12 November 2019 21:30:24 UTC+1, Michael Haufe (TNO)  wrote:
> On Tuesday, November 12, 2019 at 9:09:06 AM UTC-6, Julio Di Egidio wrote:
> > On Monday, 11 November 2019 19:54:48 UTC+1, Michael Haufe (TNO)  wrote:
> >  [...]
> > > I would say that BOTH the Functional and Object-Oriented language paradigms
> > > are obsolete
> > 
> > I would say that they both keep evolving: but I'll check your links, thanks...
> 
> I could expand on this in detail but it would get... verbose.

You should learn to read first, as I have just warned you that is NOT in
question.  That said, the problem seems rather the opposite, that you cannot
synthesize.

> Probably better as a unified blog post for presentation though
> I suspect a couple already exist on the interwebs. If someone
> has a specific gap in their knowledge here I'll try to elucidate
> as needed.

In a separate thread one would indeed hope.

> > > In regards to OOAD as a whole, I wouldn't want to make the same
> > > claims and I think it should be discussed separately.
> > 
> > But that is totally integral to what I am talking about: it's your perspective
> > strictly on the details of a formal language that is not even in question...
> 
> If we can't agree on a subset of it,

Agree on the very reductions and inversions I keep denouncing as fraudulent
and deadly?  Get real (and a bit more sincere): engineering is relevant to
production, not mathematics.  We *use* mathematics, *among other things*,
just as the economist does, or the physicist.  And the computer scientist who
misses that point is not even very competent in his own discipline: s/he better
meditate on Knuth's most famous disclaimer.

> what good would it be to talk about 
> the whole? I suspect it would just devolve into over-generalizations.
> 
> > BTW, "meta-programming" as in... meta-programming
> 
> Tautologies are true but useless. 

Just as useless as your question was.

> > : do you mean you don't know what that is, or you just
> > annoyed that I write down two liners instead of just
> > posting links?  ;)
> 
> Annoyed that you didn't post links.

Links to what?  The OP is simply a hint on something to think about, I am
not questioning any established definition or any some such, just forwarding
the hypothesis/hunch that OO in inception boils down to misplaced meta-programming: a specific case of reinventing the wheel and reinventing it wrong.
And if that rings no bells to you, in a sense or the other, so be it, I won't
hold a grudge for that, on the contrary.

> I won't assume what your
> definition is in this context. What does "meta-programming"
> mean in terms of a full-cycle methodology like OOAD?

Nothing much, meta-programming is a construction pattern that one can use
irrespective of languages and platforms: indeed a pre-processor or generator
of any sort does it, we are not limited by specific language features.

> Which "meta-programming" approach? 
> 
> A Meta-Object Protocol?
> Scheme style macros?
> C++ style static templates?
> other?

As said, meta-programming is a construction pattern, not a tool or feature.
I wouldn't call it an "approach" either, if not in the sense that it implies
a (pre-existing) higher-level approach to architecting our code.

> As paraphrased by Keean Schupke on Wittgenstein:

Paraphrasing "on" Wittgenstein is not even meaningful.

> "philosophy only exists due to the confusion of meaning
> caused when the same word is used to mean
> different things in more than one language-game."

Just like food only exists in order to nourish.

> So let's not play games.

Let's play an interesting game for a change...

Julio

[toc] | [prev] | [next] | [standalone]

#37898

On Wednesday, 13 November 2019 14:45:59 UTC+1, Julio Di Egidio  wrote:
> On Tuesday, 12 November 2019 21:30:24 UTC+1, Michael Haufe (TNO)  wrote:
<snip>
> > I won't assume what your
> > definition is in this context. What does "meta-programming"
> > mean in terms of a full-cycle methodology like OOAD?
> 
> Nothing much, meta-programming is a construction pattern that one can use
> irrespective of languages and platforms: indeed a pre-processor or generator
> of any sort does it, we are not limited by specific language features.
> 
> > Which "meta-programming" approach? 
> > 
> > A Meta-Object Protocol?
> > Scheme style macros?
> > C++ style static templates?
> > other?
> 
> As said, meta-programming is a construction pattern, not a tool or feature.

Indeed, I must thank you for bringing that up, as that kind of "literalism" just
taken to its extreme is exactly of the same kind that I mean by OO is misplaced
meta-programming.  "Instantiating the pattern"...

Julio

[toc] | [prev] | [next] | [standalone]

#37901

On Wednesday, November 13, 2019 at 7:53:34 AM UTC-6, Julio Di Egidio wrote:
> 
> Indeed, I must thank you for bringing that up, as that kind of "literalism" just
> taken to its extreme is exactly of the same kind that I mean by OO is misplaced
> meta-programming.  "Instantiating the pattern"...

Words matter, that's why I'm willing to take and possibly waste time getting on the same page before I agree or disagree.

[toc] | [prev] | [next] | [standalone]

#37900

On Wednesday, November 13, 2019 at 7:45:59 AM UTC-6, Julio Di Egidio wrote:
> On Tuesday, 12 November 2019 21:30:24 UTC+1, Michael Haufe (TNO)  wrote:
> >
> > I could expand on this in detail but it would get... verbose.
> 
> You should learn to read first, as I have just warned you that is NOT in
> question.

This isn't only about you...

>  That said, the problem seems rather the opposite, that you cannot
> synthesize.

"seems"? Why?
 
> > If we can't agree on a subset of it,
> 
> Agree on the very reductions and inversions I keep denouncing as fraudulent
> and deadly?  

Don't just denounce: demonstrate. Now's the time.

> Get real (and a bit more sincere): engineering is relevant to
> production, not mathematics.  We *use* mathematics, *among other things*,
> just as the economist does, or the physicist.  

No shit. What's your point? 

> And the computer scientist who
> misses that point is not even very competent in his own discipline: s/he better
> meditate on Knuth's most famous disclaimer.

I'm still not seeing the relevance. Bring it home.

> Links to what?  The OP is simply a hint on something to think about, I am
> not questioning any established definition or any some such, just forwarding
> the hypothesis/hunch that OO in inception boils down to misplaced meta-programming: a specific case of reinventing the wheel and reinventing it wrong.
> And if that rings no bells to you, in a sense or the other, so be it, I won't
> hold a grudge for that, on the contrary.

I'm not saying you *are* questioning definitions, and I accept that this is a hypothesis of yours; fine. At this stage in the conversation I want us to establish our domain of discourse and the terms we're using otherwise we'll just continue to talk past each other.

> > I won't assume what your
> > definition is in this context. What does "meta-programming"
> > mean in terms of a full-cycle methodology like OOAD?
> 
> Nothing much, meta-programming is a construction pattern that one can use
> irrespective of languages and platforms: indeed a pre-processor or generator
> of any sort does it, we are not limited by specific language features.
> 
> > Which "meta-programming" approach? 
> > 
> > A Meta-Object Protocol?
> > Scheme style macros?
> > C++ style static templates?
> > other?
> 
> As said, meta-programming is a construction pattern, not a tool or feature.
> I wouldn't call it an "approach" either, if not in the sense that it implies
> a (pre-existing) higher-level approach to architecting our code.

ok, so if I understand you correctly you are treating them like macros (either static or dynamic)?

[toc] | [prev] | [next] | [standalone]

#37777

On Wednesday, November 13, 2019 at 12:46:48 PM UTC-6, John G Harris wrote:
> 
> Which objective are you two aiming for : large programs that work, or
> programming languages that are more expressive ?

I don't think these are mutually exclusive. I'm not completely sure what Julio is against yet though hence my restricted response to the language paradigm. 

I think we're going into Software Engineering methodology and that Julio hypothesizes that using "OO" for this (OOAD) is an overreach but he'd have to elaborate or correct me if this is not true. 

[toc] | [prev] | [next] | [standalone]

#37779

On Wednesday, 13 November 2019 19:46:48 UTC+1, John G Harris  wrote:
> On Mon, 11 Nov 2019 10:54:40 -0800 (PST), "Michael Haufe (TNO)"
> <tno@thenewobjective.com> wrote:
> >On Monday, November 11, 2019 at 10:51:41 AM UTC-6, Julio Di Egidio wrote:
> >> 
> >> "Political" issues aside, from a strictly technical point of view I would
> >> maintain that object oriented programming (meaning the whole idea) in its
> >> essence is *misplaced meta-programming*!
> >> 
> >> Thoughts?
> >
> > I would say that BOTH the Functional and Object-Oriented language paradigms are obsolete.
> > In regards to OOAD as a whole, I wouldn't want to make the same claims and I think it should be discussed separately. 
>   <snip>
> 
> Which objective are you two aiming for : large programs that work, or
> programming languages that are more expressive ?

Great question, thank you.  In a first instance, I too would notice that these
are not mutually exclusive, on the contrary: still it is interesting to see what
very different roles they play and at which very different levels, plus it might
help clarify where I am coming from.

Programs that work and keep working, large or otherwise, is essentially the
primary goal of *production* (again, all political issues aside).  This can be
made precise with technical (engineering) definitions: quality is the degree of
correspondence of a product to its requirements, and competitiveness is a total
measure of how good production is doing, taking into account quality vs costs.
The exact formula can be seen in a blog post of mine where I also mention a
peculiarity of software production, and only of software production, with the
relation of quality to costs, missing which is yet another recipe for guaranteed
disaster: <http://architectando.blogspot.com/2013/08/competitiveness-in-production.html>

On the other hand, as for programming languages, relative to the production
problem these are "just" tools, so, sure, we can imagine the sorry situation of
the plumber who forgets his toolbox home, or who takes a sloppy one, but, for
how important even the tools are, it is never the tools that make the art...
Indeed, just to say, I have built successful distributed (LAN plus WAN) systems
that have been in operation for more than a decade with stuff like MS FoxPro and
PC Anywhere on MS DOS 5 before even Ethernet existed, and I can confirm it was
painful, yet I could have done it in plain BASIC and raw COM if that was all
that was available, it just would have been even more painful, but then read
costs...

As for "expressiveness" of languages, while you might be hitting the nail in
the head relative to the issue with OO that I am raising ("expressing" what
and where), I won't reply to you to the point: disentangling what you just
said would be laborious :) and maybe not that critical.

Julio

[toc] | [prev] | [next] | [standalone]

#37794

On Thursday, November 14, 2019 at 11:44:33 AM UTC-6, Julio Di Egidio wrote:
> [...]
> <http://architectando.blogspot.com/2013/08/competitiveness-in-production.html>
> [...]

I noticed through here you have a featured link to:

<http://www.geocities.com/tablizer/oopbad.htm>

Archived at
<https://web.archive.org/web/20140705032244/http://www.geocities.com/tablizer/oopbad.htm>

Is that the main foundation of your opinion on OO in general still?

I haven't read through it yet, but I suspect it will answer some questions I didn't think to ask you.

[toc] | [prev] | [next] | [standalone]

#37804

On Fri, 15 Nov 2019 11:16:49 -0800 (PST), "Michael Haufe (TNO)"
<tno@thenewobjective.com> wrote:

  <snip>
>I noticed through here you have a featured link to:
>
><http://www.geocities.com/tablizer/oopbad.htm>
>
>Archived at
><https://web.archive.org/web/20140705032244/http://www.geocities.com/tablizer/oopbad.htm>

The authors show why it is unwise to predict the future in too much
detail : 
  "I will eat a week's pay if OOP is still in vogue in 2015."
:-)


They say 
	"I agree that there may be areas where OOP does reasonably well. My
  main complaint is about using OO for so-called 'business modeling'."

They seem to have missed an important point. In C++, at least, you can
create an object type when it is convenient, and sometimes it is very
very convenient, but you don't have to make *everything* an object.
This use of convenient objects applies to any kind of program,
including business systems.

Once a type is created you can think of the objects instead of the
petty details inside the object. You can think of woods instead of
trees.

  John

[toc] | [prev] | [next] | [standalone]

#37902

On Mon, 11 Nov 2019 10:54:40 -0800 (PST), "Michael Haufe (TNO)"
<tno@thenewobjective.com> wrote:

>On Monday, November 11, 2019 at 10:51:41 AM UTC-6, Julio Di Egidio wrote:
>> 
>> "Political" issues aside, from a strictly technical point of view I would
>> maintain that object oriented programming (meaning the whole idea) in its
>> essence is *misplaced meta-programming*!
>> 
>> Thoughts?
>
>I would say that BOTH the Functional and Object-Oriented language paradigms are obsolete. In regards to OOAD as a whole, I wouldn't want to make the same claims and I think it should be discussed separately. 
  <snip>

Which objective are you two aiming for : large programs that work, or
programming languages that are more expressive ?

  John

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.javascript

csiph-web