Subj : Re: GNU Public Licences Revisited (again)
To   : comp.programming
From : Rob Thorpe
Date : Wed Aug 31 2005 07:06 am

Randy Howard wrote:
> Rob Thorpe wrote
> (in article
> <1125480091.914143.93630@z14g2000cwz.googlegroups.com>):
>
> > Randy Howard wrote:
>
> >> A compiler and linker is just as much a tool to
> >> produce object code as a hammer or saw is to a cabinet maker.
> >> The source code isn't what most programmers are paid for, they
> >> are paid for the finished product.
> >
> > A compiler is completely different to a hammer or saw, the hammer or
> > saw doesn't make the product almost incomprehensible, even to those
> > knowledgable in the subject.
>
> No, it is a tool that converts raw materials into a finished
> product.

I didn't say it wasn't, I said it makes the work the programmer has
done incomprehensible.

> That product is a working piece of software when
> executed, which is much more important than the encoding on the
> disk before it runs.
>
> Software written in scripting languages is of course trivial to
> understand.
>
> I would also point out that there was a time when I could "read"
> PDP-11 binaries and disassemble them in my head to macro-11
> syntax.  About all I remember of that today is the opcodes for a
> few instructions and the main addressing modes.  However, at the
> time, it was certainly doable, because I did it every day, all
> day long.  It's been many a moon since I've seen a running
> PDP-11, or a binary for it, so I won't pretend to still have
> that ability.

That's true, some old architectures and compilers (S/360 ?) were
understandable as machine code.  It's not really true now though.

> >>> I don't think that something that could be changed, it would be an
> >>> unnecessary imposition on peoples freedom to limit their ability to
> >>> keep secrets.
> >>
> >> But somehow, that only applies to everyone /other/ than those
> >> wishing to keep source code secret?
> >
> > No, it should apply to people writing programs too.  It would not be
> > civilised to demand that people give away their secrets.
>
> Ok.
>
> > What shouldn't apply is copyright, people should not be able to
> > arbitrarily protect with copyright what they distribute.
>
> But once they a buy a copy from this person, who they were just
> being so 'civilized' toward, they are free to make as many
> copies as they like and give them to all their friends?  How
> civil.  That doesn't seem to be very consistent with the above.

Would it be more civil if the person who bought the software was
prevented from selling it or giving it to someone else?  I don't think
so.

> I'm not 100% sure I know whether or not you think it's okay for
> people to not release their source code, or it isn't ok.  I'm
> not sure how it lines up with your comment about keeping secrets
> if they want to do so.

Keep it secret is OK in my book.  But once you distribute it those who
you distribute it to should have the same rights you had over it, that
is to see the source, distribute it and change it.

> BTW, copyright is not arbitrary, it's automatic.

It's application to object code is arbitrary, it was intended to apply
to the work of humans.

> >>> What I object to is people try to impose copyright on
> >>> the output of a *compiler* that they don't even understand.
> >>
> >> I don't follow you.
> >>
> > Oh come on :)
>
> Put it this way, Tollkien copyrighted the Silmarillion, which is
> basically unparseable.  People can write books in made up
> gibberish, and copyright them.  Or, they could write a book
> totally in code, a modern equivalent of the Hypnerotomachia
> Poliphili if you will, and copyright that.  It's actual meaning
> would likely be /far/ less comprehensible to you, even after
> decades of study, than the object code of a compiler.  Just
> because some copyrighted works are easy to understand, that does
> not mean that they all are as easy.  Also, I fail to see how the
> ease of comprehension makes the copyright more or less valid.
>
> For a real life example. pick up a copy of "Reimann's Zeta
> Function" by H.M. Edwards, and get a randomly selected math
> teacher to explain every page to you.  Better yet, try and get
> them to explain pages 152 and 153 alone in a convincing manner.
> Yet, I see this text "Copyright (C) Harold M. Edwards" inside.

These things aren't really relevant.  The mathematics textbook will be
understandable to someone else working in that field.  It is isn't then
it's useless.
If people can write works of fiction that are gibberish and make money
out of them then good luck to them.

Software is different in that it works by being run by a machine, so
even if it is incomprehendible it can still be useful.


> > Maybe, but today patents don't protect the inventor they only protect
> > the financial concern that owns the patent.
>
> Except in the cases where those are one and the same.  You're
> just annoyed that corporations have patent portfolios.

Not really, if they pay people for their work then that's OK.

> If you
> want to argue about redistribution of wealth, take it to a
> political newsgroup.  Individuals apply for and are granted
> patents all the time.

You can only do that if you have enough money to patent the idea in
several juristictions, which means you have to be independently
wealthy, or have backing.  If you have backing you give away your idea
anyway.  Practically speaking, you or I or any normal individual cannot
patent something and expect to make money from it.

> >>>> It's not incomprehensible to someone sufficiently skilled.
> >>>> Analyzing modern art and finding meaning in it is as opaque to
> >>>> me as reading even the source code, much less the binary of a
> >>>> large software project is to a performance artist.
> >>>
> >>> That's completely irrelevant, you're not an artist.  It's not that
> >>> important that things be understandable to those outside the field,
> >>> since they couldn't possibly build on them.  The problem is could an
> >>> artist analyze a peice of modern art?
> >>
> >> That would be a funny experiment.  Find 100 artists, get 100
> >> different opinions.  :-)
> >
> > Not a problem.  Find 100 art buyers get another 100 opinions.
>
> Your point above was that it wouldn't be understandable to me,
> but it would be to experts.  By your own admission, the experts
> couldn't figure it out with any kind of a quorum on the facts
> either, so it's an intractable problem.  So much for copyright
> on art I suppose if your theories were applied.  :-)

But if it's art, if anyone can understand it, they can immitate it.

> >> Artists work in clay, or wood, or pastels.  They sell statues,
> >> carvings and paintings though.  Get it?
> >
> > Yes, but the structure of the clay, wood or pastel is obvious in the
> > statue, carving or painting.
>
> The structure of a web browser is obvious to even relative
> novices to computer usage too.  You don't need to examine object
> code to determine what software does.  Running it is quite often
> sufficient.  Again, I fail to see how this 'degree of
> difficulty' impinges on the ability to copyright something.

Is it practical to design software by reverse engineering?

> >>> Programming is different, the intention is only to make something that
> >>> works.
> >>
> >> If that were completely true, there would be a lot less broken
> >> software out there.
> >
> > Are you saying the intention is to make something artistic?
>
> No, I was being flippant about whether or not programmers as a
> whole can be seriously described as having the sole intention of
> making something that works.  Many of them don't appear to care
> if it works one way or the other, judging by the junk available
> on the market today.

Fair enough.  Programmers of proprietry code have little to motivate
them, they spend most of their time immitating features present in
other proprietry software.  It's all such a waste of time.

> > Programming is best analogised to science not art, because no one
> > really cares much how aesthetic it is.
>
> I'm pretty much at the opposite end of this from you.
> Programming isn't a science, it's much too lacking in concrete
> knowledge. Look at the dozens, if not hundreds of books on
> various aspects of it that start out "The Art of ... <insert
> geek words here".  I can't recall anyone writing a book entitled
> "The Science of All Night Coding Sessions".  Can you?

Good idea though.

I didn't mean the programming is a science, I meant that it's outputs
are very like the outputs of science.  Programs like scientific
conclusions (and unlike physical things) are easily copied.  They are
not tied up in aesthetics like works of art.  Lastly, the process of
programming is accumulative, as knowledge is accumulative.

> > It just has to work, so
> > building directly on the work of others is not tasteless.
>
> Ah, but the /art/ is in making it work when it doesn't want to.
> :-)

That would be true, true of science also.

> >> You are arguing that programmers should have a right to
> >> plagiarism.  Stunning.
> >
> > Yes.
>
> Well, I have to give you brownie points for admitting it.  I'm
> disappointed in the answer, but at least you don't run away from
> it.
>
> >> Since it is easier to steal a plot from a book than to steal a
> >> function from a piece of hidden source code, you think there is
> >> a problem?
> >
> > Just as scientists are allowed to build on the work of their
> > predecessors, so programmers should be allowed to build on the work of
> > their predecessors.
>
> Because some scientists choose to publish their work, for that
> purpose.  Some scientists /do not/ publish their work, and keep
> it as a trade secret as long as possible.  Some scientists
> actually put out false information on purpose in the hopes of
> leading competitors astray.

Unfortuntely all this is true.

> Anyway, what you are really advocating is forced confiscation of
> someone else's source code against their will?

Oh no, if you want to keep it, then keep it.  But if you're going to
distribute it it isn't right that you're allowed to limit the rights of
those you distribute it to.

> > At present they're not because of arbitrary restrictions put in their
> > way.
>
> They are not arbitrary.  If I tell you, "hey, this is my code,
> I've been working on it for the last 5 years, and you can't have
> it", and you decide you are going to take it from me anyway, you
> better bring a weapon and hope you know how to use it better
> than I can use mine.  That has nothing to do with copyright law,
> it's outright theft.

Why did you spend 5 years writing it.  If software were free you could
probably spend a couple of weeks modifying someone elses program to do
what you need.

> >> Why should a programmer have a right to use source that I spent
> >> weeks or months working on, just so he can feel he has it as
> >> easy as an author too lazy to dream up his own murder plot line?
> >
> > So the scientist is allowed to benefit from his/her predecessors, the
> > artist or writer can benefit from his/her predecessors but the
> > programmer cannot.
>
> *sigh*  There is no law that a 'scientist' has to release his
> work.  He is free to set fire to it if he wants to.

Yes, but if he does release it, then any other scientist can build on
it.

>  And sorry,
> programmers are /not/ scientists, there are some scientists that
> also know how to program, but you have to pay attention to the
> venn diagram on that one.  It's not what you think it is.

I agree, it's the end products that are similar.

> > This is why programming remains in the dark ages.
>
> That sounds so catchy, but it just isn't reality.  Scientists
> have been studying chemistry for thousands of years.  "Computer
> Science" has been around for a blink of an eye in comparison.  I
> would expect it to be woefully incomplete at this point.  It's
> rather staggering how far it has progressed in so brief a time.
> In another field, in this phase of infancy, they'd be burning
> you at the stake for something like thinking up quicksort
> instead of agreeing with the notion that bubblesort is nirvana
> about this far into the timeline.

I think computer hardware has certainly gone a long way, which has
helped the software.  It's not a complete disaster certainly, but a
great many people spend their lives solving solved problems.

> >> It would however, be totally unfair to the original author,
> >> unless that person /wanted/ to give it up for such acts.  You
> >> might as well demand that chemists give up their formulas for
> >> snake oil, so that other snake oil companies won't have to start
> >> from scratch.  What rational basis is there for this need to
> >> make others have it easier off the fruit of others' labor?
> >
> > I'm asking for nothing much different than other disciplines enjoy.
> > If you sell snake oil I can buy some of yours, analyze it and steal the
> > recipe.
>
> And you can do the same thing to a program.  "Analyze" does not
> mean "give me the formula for that snake oil, right now, it's my
> right to steal it from you so I can advance my own 'science'".
> You must be arguing for access to the source code, because you
> have already stipulated that you find object code to hard to
> read.

Exactly, giving others access to the object code is not really helpful.

> >> Strange, Office is going XML reportedly, despite still retaining
> >> copyright protection on their software.  How can that be?
> >
> > They're not doing it in a way that's useful to anyone else.
>
> Explain.

I think others have posted about this.

> > There is no point in those companies that have an effective monopoly,
> > like microsoft, from trying to build better products, it's just a waste
> > of their money.
>
> M$ sure spends a lot of money on R&D for a company that sees no
> point in it.  I wish they'd spend less, it would make it easier
> for some other companies to catch up with them.

They don't apply it much though.  Many of their products haven't
changed significantly in years.  They use it to break into new markets
and extend their dominance.

.