Weblogs Forum - Hidden Variables

Articles |
News |
Weblogs |
Books |
Forums

Artima Forums | Articles | Weblogs | Java Answers | News

Sponsored Link •

Weblogs Forum
Hidden Variables

68 replies on 5 pages. Most recent reply: Sep 11, 2006 2:28 AM by Miguel Montenegro

Welcome Guest
Sign In

Back to Topic List

Reply to this Topic

Search Forum

Threaded View


Previous Topic		Next Topic

Flat View: This topic has 68 replies on 5 pages [ « | 1 2 3 4 5 | » ]

Achilleas Margaritis

Posts: 674
Nickname: achilleas
Registered: Feb, 2005

Re: Hidden Variables

Posted: Jul 12, 2006 1:16 AM

> I really don't know what you mean here. One of the great
> discoveries of Stephen Hawking was that black holes emit
> energy and the mechanism for this depends absolutely on
> quantum mechanics. Nuclear weapons are also a result of
> quantum mechanics. The way we understand the sun depends
> on quantum mechanics. Quntum mechanics is by far the most
> important advancement of science in the 20th century.

Agreed, but humanity already has experience of the situation where a theory explains what we see, and after a few years a new theory that is a superset of the previous theory comes along and explains even more things.

>
> You don't really seem to have a good feel for the basics
> of quantum theory. A really good book for this is call
> "Quantum, A Guide for the Perplexed". I studied quatnum
> mechanics breifly but this book was what really cleared up
> the basic concepts for me.

Wow, thanks a lot for remotely understanding if I have a good feel of quantum physics. I really appreciate that :-).

>
> One of the things that is very hard to accept for most
> people is that quantum mechanics is not just a trick and
> that our natural understanding of how the universe works
> is not complete.

But that is exactly what I am saying: our natural understanding of how the universe works is not complete...and therefore we can not rule out the possibility that the randomness we see is a result of the works of another system hiding under what we know so far.

Achilleas Margaritis

Posts: 674
Nickname: achilleas
Registered: Feb, 2005

Re: Hidden Variables

Posted: Jul 12, 2006 1:26 AM

> I think we are missing the point completely. When you look
> at the big picture, the real problem is that our so called
> software engineering discipline is not a proper
> engineering discipline. More specifically, we don't have a
> proper component model for software.

Additionally, we don't have a proper way to write code that writes code. Well, that technique exists in certain languages, but it has not caught up yet with mainstream languages.

For example: many (if not all) GUI applications use dialogs. We code each dialog with much pain, whereas a simple meta-function that was fed the kind of data that we would like to edit and produced the right code would be more than enough.

Can we say to the computer: "edit a number that ranges from 3 to 100 and store it in variable 'foo'" in one sentence? nope, we can't. This simple thing alone takes too much code to write (create the widgets, arrange the widgets, setup the callbacks, setup the validation, setup the persistence etc).

Can we say to the computer: "my persistent data is a list of foo named 'foos' and a map of bar to foo named 'bar', and they are persisted in computer 'a' with username 'b' and password 'c'". Nope, we can't. We have to setup a whole mechanism of mapping databases to objects, open connections, close connections, have connection pooling, write SQL, etc.

All these things would propably be possible with code that writes code...then the level of abstraction would greatly increase and therefore the possibility of errors would be minimized.

Achilleas Margaritis

Posts: 674
Nickname: achilleas
Registered: Feb, 2005

Re: Hidden Variables

Posted: Jul 12, 2006 1:32 AM

> The solution is to drop determinism entirely and find out
> at every moment what works and what doesn't.

I partially agree with that. The problem is not that determinism is an illusion...it is not. It is just us that we are not able to calculate the outcome. For example, take a French billiard table with two balls only: we can easily predict their trajectories, when they will collide etc. Add another ball, and the problem becomes very difficult, if not impossible, to predict. But the outcome is deterministic nevertheless because it is defined by physical laws.

I wonder if it would be possible to make a compiler which 'pseudo-runs' a program in order to see where it fails and if the implicit or explicit contracts hold. By 'pseudo-run' I mean to skip the side effects and pretend some things happened in certain ways...it would be helpful to do so because, as you say, it is difficult to predict the future since it contains so many variables.

disney

Posts: 35
Nickname: juggler
Registered: Jan, 2003

Re: Hidden Variables

Posted: Jul 12, 2006 2:58 AM

An interesting take on one of the oldest topics in software, Bruce, and maybe still the most important?

Those who would like to move software into the arena of predictable and deterministic science and engineering have been trying for a long time, but have so far come up with nothing. My imminent redundancy has been predicted by such people for years, but my skills are still very much in demand.

That software is a craft is a perspective that works for me, and those I work with. Who can ask more than that?

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: Hidden Variables

Posted: Jul 12, 2006 6:11 AM

> > Does your computer not contain trasnistors? I suspect
> it
> > contains many millions. Transistors are a direct
> result
> > of the study of quantum mechanics.
>
> I thought transistors were a result of the science of
> electronics. Please show me a link where the study of
> quantum mechanics led to the invention of the transistor.

Well, you can't you understand how transistors work without quantum mechanics.

http://www.pbs.org/transistor/science/info/quantum.html

"Quanta, quanta everywhere

This idea that particles could only contain lumps of energy in certain sizes moved into other areas of physics as well. Over the next decade, Niels Bohr pulled it into his description of how an atom worked. He said that electrons traveling around a nucleus couldn't have arbitrarily small or arbitrarily large amounts of energy, they could only have multiples of a standard "quantum" of energy.

Eventually scientists realized this explained why some materials are conductors of electricity and some aren't -- since atoms with differing energy electron orbits conduct electricity differently. This understanding was crucial to building a transistor, since the crystal at its core is made by mixing materials with varying amounts of conductivity."

> Ok, we don't know what causes the randomness, I agree with
> that. But exactly because of that, we can not say that
> randomness has no explanation and that nature is random by
> itself.
>
> We simply do not know what causes randomness.

Unless we can measure what causes it, it can't be understood with science.

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: Hidden Variables

Posted: Jul 12, 2006 6:23 AM

> > I really don't know what you mean here. One of the
> great
> > discoveries of Stephen Hawking was that black holes
> emit
> > energy and the mechanism for this depends absolutely on
> > quantum mechanics. Nuclear weapons are also a result
> of
> > quantum mechanics. The way we understand the sun
> depends
> > on quantum mechanics. Quntum mechanics is by far the
> most
> > important advancement of science in the 20th century.
>
> Agreed, but humanity already has experience of the
> situation where a theory explains what we see, and after a
> few years a new theory that is a superset of the previous
> theory comes along and explains even more things.

But they don't invalidate the original theory (if it was good) they qualify it. Newtonian physics is supercedede by Einsteins theories but people sill use Newton's laws much more often than Einstein's theories.

> >
> > You don't really seem to have a good feel for the
> basics
> > of quantum theory. A really good book for this is call
> > "Quantum, A Guide for the Perplexed". I studied
> quatnum
> > mechanics breifly but this book was what really cleared
> up
> > the basic concepts for me.
>
> Wow, thanks a lot for remotely understanding if I have a
> good feel of quantum physics. I really appreciate that
> :-).

I don't want to offend you but a lot of the things you have written are plainly wrong and can be easily cleared up with a little more knowledge.

Quantum mechanics may be superceded by another theory but neither you nor I are qualified to make any hypotheis about it. Your assertion that there are hidden variables appears to come solely from an emotional response. There is no eveidence for it. This is the mistake that Einstein made. He let his feelings cloud his judgement and spent many of the last years of his life on a wild goose chase.

> > One of the things that is very hard to accept for most
> > people is that quantum mechanics is not just a trick
> and
> > that our natural understanding of how the universe
> works
> > is not complete.
>
> But that is exactly what I am saying: our natural
> understanding of how the universe works is not
> complete...and therefore we can not rule out the
> possibility that the randomness we see is a result of the
> works of another system hiding under what we know so far.

We can't rule it in either. There's no evidence for it. We can't rule out the existence of tree-spirits, gnomes, fairies or anal-probing aliens either.

Isaac Gouy

Posts: 527
Nickname: igouy
Registered: Jul, 2003

Re: Hidden Variables

Posted: Jul 12, 2006 8:44 AM

James Watson wrote
Newtonian physics is supercedede by Einsteins theories but people sill use Newton's laws much more often than Einstein's theories.
Could that be because for the problems and measurement scales those people are dealing with Einstein's theories don't provide a practical improvement on Newtonian physics?

-snip-
We can't rule it in either. There's no evidence for it. We can't rule out the existence of tree-spirits, gnomes, fairies or anal-probing aliens either.

Which brings together those old notions of good reasoning
- absence of evidence is not evidence of absence
- an inability to disprove does not prove

what can be asserted without evidence can also be dismissed without evidence

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: Hidden Variables

Posted: Jul 12, 2006 8:56 AM

> James Watson wrote
> Newtonian physics is supercedede by Einsteins theories
> but people sill use Newton's laws much more often than
> Einstein's theories.
> Could that be because for the problems and measurement
> scales those people are dealing with Einstein's theories
> don't provide a practical improvement on Newtonian
> physics?

Precisely. Most of our experiences occur in the range of speeds and forces where Eisteinian physics is so close to Newtonian physics that we cannot detect the difference which is why Newton wasn't 'wrong'. Even the widely known experiment using a solar eclipse that 'proved' Einstein's threoroes has been shown to lack the precision required to test the result it purported to show.

Bruce Eckel

Posts: 875
Nickname: beckel
Registered: Jun, 2003

Re: Hidden Variables

Posted: Jul 12, 2006 11:05 AM

> Re the componentization of software, I think there are
> economic and physical reasons why componentization will
> never be a panacea. And, we can't say that it hasn't been
> attempted; it's been attempted countless times.

And yet componentization *helps*. As does automatic code generation, when that's useful, and a framework, when that's appropriate, and very often object-orientation, etc. etc.

The problem with the deterministic mindset is that it wants to extrapolate "helps" to "the solution for all problems." Thus, if a little static type checking is helpful, the logical conclusion is that our problems are caused by not enough static type checking. The "structured revolution" was based on the idea that the problem was insufficient documentation to describe your system, and the solution was to create documents for everything. You can usually find a single defining principle -- and underlying assumption -- for just about any of these "revolutions."

Bruce Eckel

Posts: 875
Nickname: beckel
Registered: Jun, 2003

Re: Hidden Variables

Posted: Jul 12, 2006 11:09 AM

> I wonder if it would be possible to make a compiler which
> 'pseudo-runs' a program in order to see where it fails and
> if the implicit or explicit contracts hold. By
> 'pseudo-run' I mean to skip the side effects and pretend
> some things happened in certain ways...it would be helpful
> to do so because, as you say, it is difficult to predict
> the future since it contains so many variables.

A compiler is a translator. It converts a program from one form to another. However, your concept of 'pseudo-run' is an approximation, which is basically what we do with testing -- in particular what you describe sounds mostly like testing with mock objects.

Bruce Eckel

Posts: 875
Nickname: beckel
Registered: Jun, 2003

Re: Hidden Variables

Posted: Jul 12, 2006 11:15 AM

> > therefore we can not rule out the
> > possibility that the randomness we see is a result of
> the
> > works of another system hiding under what we know so
> far.
>
> We can't rule it in either. There's no evidence for it.
> We can't rule out the existence of tree-spirits, gnomes,
> , fairies or anal-probing aliens either.

Actually, I think it's Bell's Theorem -- with attendant experimental evidence -- that proves that there is no hidden variable. We'd have to check with a more up-to-date physicist than I, but I do know that there has been experimental evidence in the last 30 years that actually does rule out another system hiding under what we know, adding in the deterministic factor.

Michael Feathers

Posts: 448
Nickname: mfeathers
Registered: Jul, 2003

Re: Hidden Variables

Posted: Jul 12, 2006 11:17 AM

> > Re the componentization of software, I think there are
> > economic and physical reasons why componentization will
> > never be a panacea. And, we can't say that it hasn't
> been
> > attempted; it's been attempted countless times.
>
> And yet componentization *helps*. As does automatic code
> generation, when that's useful, and a framework, when
> that's appropriate, and very often object-orientation,
> etc. etc.
>
> The problem with the deterministic mindset is that it
> wants to extrapolate "helps" to "the solution for all
> problems."

I agree. I think that in the case of software componentization we've seen some benefits, but we can't hold up hardware componentization as the standard to strive for; it's a siren song.

Isaac Gouy

Posts: 527
Nickname: igouy
Registered: Jul, 2003

Re: Hidden Variables

Posted: Jul 12, 2006 11:32 AM

Bruce Eckel wrote
-snip-
> Thus, if a little static type checking is
> helpful, the logical conclusion is that our problems are
> caused by not enough static type checking.

I can see that someone might suggest that more static type checking would be more helpful, I don't see why you think someone would suggest that static type checking has anything to do with invalid requirements.

-snip-
> The "structured revolution" was based on the idea that the
> problem was insufficient documentation to describe your
> system, and the solution was to create documents for
> everything.

It's a long time since I read the 1972 "Structured Programming" O.-J. Dahl, E. Dijkstra, and C. Hoare - but I don't think it had much to say about insufficient documentation.

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: Hidden Variables

Posted: Jul 12, 2006 11:43 AM

> > > therefore we can not rule out the
> > > possibility that the randomness we see is a result of
> > the
> > > works of another system hiding under what we know so
> > far.
> >
> > We can't rule it in either. There's no evidence for
> it.
> > We can't rule out the existence of tree-spirits,
> gnomes,
> > , fairies or anal-probing aliens either.
>
> Actually, I think it's Bell's Theorem -- with attendant
> experimental evidence -- that proves that there is no
> hidden variable. We'd have to check with a more up-to-date
> physicist than I, but I do know that there has been
> experimental evidence in the last 30 years that actually
> does rule out another system hiding under what we know,
> adding in the deterministic factor.

From reading the introduction to this:

http://www.ncsu.edu/felder-public/kenny/papers/bell.html

It appears that this theorem is about locality and not the uncertainity principle.

I don't think there's anyway to prove that there is no undetectable system below what we can detect. It's kind of like arguing whether there are other universes. By definition we can never know about another universe because if we could it would be part of our universe. Proving that something we cannot possibly know about doesn't exist is a pretty tricky proposition. Saying that we know what doesn't exist in a realm that is undetectable is like saying "I can't know anything about what is inside this box but I know what is not inside it." I'm not going to say that's an outright contradiction but it seems to be one.

James Watson

Posts: 2024
Nickname: watson
Registered: Sep, 2005

Re: Hidden Variables

Posted: Jul 12, 2006 11:50 AM

> > > Re the componentization of software, I think there
> are
> > > economic and physical reasons why componentization
> will
> > > never be a panacea. And, we can't say that it hasn't
> > been
> > > attempted; it's been attempted countless times.
> >
> > And yet componentization *helps*. As does automatic
> code
> > generation, when that's useful, and a framework, when
> > that's appropriate, and very often object-orientation,
> > etc. etc.
> >
> > The problem with the deterministic mindset is that it
> > wants to extrapolate "helps" to "the solution for all
> > problems."
>
> I agree. I think that in the case of software
> componentization we've seen some benefits, but we can't
> hold up hardware componentization as the standard to
> strive for; it's a siren song.

I think it's interesting to consider that even in manufacturing (an old industry) componentization is not absolute. For example, two car maker's parts are not interchangeable and many parts are not interchangeable between two models made by the same car maker. It's conceivable that cars could be manufactured almost solely from a standard set of components but they are not. The parts are custom designed for each maker and also for each model. If you look at how good software is made currently, it's basically the same thing. You design resuable components for your team, project or company. The idea that software development will become like putting lego blocks together was never a sound theory IMO.

Flat View: This topic has 68 replies on 5 pages [ « | 1 2 3 4 5 | » ]

Previous Topic

Next Topic