Linux-Advocacy Digest #324, Volume #34 Tue, 8 May 01 12:13:05 EDT
Contents:
Re: Richard Stallman what a tosser, and lies about free software (T. Max Devlin)
Re: Richard Stallman what a tosser, and lies about free software (T. Max Devlin)
Re: Richard Stallman what a tosser, and lies about free software (T. Max Devlin)
Re: Richard Stallman what a tosser, and lies about free software (T. Max Devlin)
Re: Richard Stallman what a tosser, and lies about free software (T. Max Devlin)
Re: Richard Stallman what a tosser, and lies about free software (T. Max Devlin)
Re: Richard Stallman what a tosser, and lies about free software (T. Max Devlin)
Re: Richard Stallman what a tosser, and lies about free software (T. Max Devlin)
----------------------------------------------------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss,comp.os.ms-windows.advocacy
Subject: Re: Richard Stallman what a tosser, and lies about free software
Reply-To: [EMAIL PROTECTED]
Date: Tue, 08 May 2001 16:03:54 GMT
Said Ayende Rahien in comp.os.linux.advocacy on Mon, 7 May 2001 00:00:53
>"T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
>news:[EMAIL PROTECTED]...
>> Said Ayende Rahien in comp.os.linux.advocacy on Sun, 6 May 2001 12:15:23
>> >"Les Mikesell" <[EMAIL PROTECTED]> wrote in message
>> >news:Pg3J6.9817$[EMAIL PROTECTED]...
>> >>
>> >> "T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
>> >> news:[EMAIL PROTECTED]...
>> >>
>> >> > >An API is not complete without the documentation of what its
>function
>> >> does.
>> >> >
>> >> > You mean the library won't work if a programmer makes a function call
>> >> > unless the function is documented?
>> >>
>> >> He means that the program must be expected to break when you upgrade
>> >> the library to the next version if you use anything beyond what is
>> >> documented. If you link statically that might not be a problem.
>> >
>> >Not the library is expected to break if what is uses have been changed or
>> >remove.
>> >Otherwise, it can go on happily with any number of updates.
>>
>> "Can" != "will".
>
>Can == Will in computer world.
>This is a deterministic enviroment, T. Max.
No, it is a conditional environment; "can" != "will".
--
T. Max Devlin
*** The best way to convince another is
to state your case moderately and
accurately. - Benjamin Franklin ***
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss,comp.os.ms-windows.advocacy
Subject: Re: Richard Stallman what a tosser, and lies about free software
Reply-To: [EMAIL PROTECTED]
Date: Tue, 08 May 2001 16:03:55 GMT
Said Ayende Rahien in comp.os.linux.advocacy on Mon, 7 May 2001 00:09:18
>"T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
>news:[EMAIL PROTECTED]...
>> Said Les Mikesell in comp.os.linux.advocacy on Sun, 06 May 2001 03:10:39
>> >"T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
>> >news:[EMAIL PROTECTED]...
>> >
>> >> >An API is not complete without the documentation of what its function
>> >does.
>> >>
>> >> You mean the library won't work if a programmer makes a function call
>> >> unless the function is documented?
>> >
>> >He means that the program must be expected to break when you upgrade
>> >the library to the next version if you use anything beyond what is
>> >documented. If you link statically that might not be a problem.
>>
>> No, he meant that the program must be expected to break unless you have
>> complete documentation for the library to begin with. Have you been
>> following the discussion?
>
>No, it certainly isn't expected to break if I don't have a full
>documentation of the library that it uses.
Don't people notice when they make their statements unfalsifiable? Or
do they simply not realize how completely it refutes their own point?
The "if", Ayende, is the whole point of the matter. It makes your
statement a useless tautology, since there is no way of knowing, save
some person's personal bias, which is 'breaking the contract', and which
is 'not having full documentation'. Get it?
>I only need documentation for the part that I use.
You only need documentation if you don't already know what you're using.
IOW, the documentation is "the API", or rather as close as you can get
to the platonic object you're referring to with 'API'. What the term
actually means is "library", but due to the ambiguity of the verb "use",
a different term was needed for "use" of a library that doesn't violate
copyright law or license agreements as some "uses" do. The "use" of a
library which cannot infringe is metaphorically redirected to the API,
and this mythical thing called an API is then exempted from copyright
protection. The "use" of a library which could still be an infringement
of copyright is the "use" covered by licensing agreements. This is
further discombobulated, though, by the use of trade secret agreements
as EULAs.
Yet another flaky extraction of the quite complete and consistent
understanding of the abstraction of copyright. Copyright 2001, T. Max
Devlin.
>One of many examples, Accent Express, a word processor that allows you to
>write in one of 30 languages no matter what your OS support.
>I've a version that was built before 9x was invented.
>It run perfectly well on ME.
That is not an example of anything, really. If you believe it refutes
my point, it means you still don't understand my point. Ask questions;
I'll be happy to discuss it.
--
T. Max Devlin
*** The best way to convince another is
to state your case moderately and
accurately. - Benjamin Franklin ***
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss,comp.os.ms-windows.advocacy
Subject: Re: Richard Stallman what a tosser, and lies about free software
Reply-To: [EMAIL PROTECTED]
Date: Tue, 08 May 2001 16:03:56 GMT
Said Ayende Rahien in comp.os.linux.advocacy on Mon, 7 May 2001 00:20:27
>"T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
[...]
>A contract has physical, legal, and monetary existance.
Sure, but that's three different existences, not three characteristics
of a metaphysical object. (Rhetorically, of course, it could be
either/or. I would just like for you to pick one and stick with it
throughout the discussion, and I prefer not to bother with metaphysics.)
>So does an API.
>API spesification contain its documentation. An undocumented API is not part
>of the contract.
Which, the physical, monetary, or legal part of the contract?
>You can equal it to a verbal contract, versus a signed contract.
More metaphysics.
>A verbal contract has little existance, it can be upheld, or not. A real
>contract is one that has existance.
Yet it, too, can be upheld, or not. "Existence" is obviously not what
makes the difference, then.
>> So, like the cook reading the
>> recipe, the idea of the API as "a contract" might work just fine for a
>> rough technical translation of what an API is to the non-technical, but
>> I don't think you should take the analogy as seriously as you seem to
>> do.
>
>No, it's exactly like a contract, which is why I used this analogy.
>A contract spesified the duties and obligations of each party, and what they
>get in exchange of those obligations. An API does the same.
So does a dictionary. Why not call it a dictionary? So does a law.
Why not call it a law? It is *analogous* to a contract, Ayende; that is
why you use that analogy. It cannot possible be "exactly like a
contract", because it is not a contract, it is an API.
>> Or at least recognize that "contracts" are mental constructs, as much as
>> specifications written on paper. An API is a platonic object that
>> represents nothing more than the library (or any other library which
>> supports the API).
>
>No, an API is not a platonic object, it is real, it has existance, it has
>affect.
So does "science". Yet it is not a physical object. All abstractions
are platonic objects. Some have no physical existence at all, like
"truth". Some have a physical extraction, like "law" and "API" and
"agreement". Either the API *is* the documentation/specification of the
API, or the API does not have a physical existence, and is a platonic
object. A handy abstraction for describing something that pointedly
*doesn't* exist in the physical world. It is a "contract" between two
programmers that do not have any binding legal agreements (contracts)
between them.
>A book represent nothing more than an idea, is it a platonic object too?
"Represent" is not something that physical objects do, in and of
themselves. That is just a reference you put on them from outside. A
book is a thing, a physical object. The concept of "a book" is a
platonic object, ultimately, but only because all words, even labels for
physical things, are concepts which are separate from those physical
things. Their "existence" in the physical world is a bit more
complicated than simple string theory physics can explain.
>A house blue prints?
Is an API a blue print? Blue prints are protected intellectual
property, I will remind you.
--
T. Max Devlin
*** The best way to convince another is
to state your case moderately and
accurately. - Benjamin Franklin ***
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss,comp.os.ms-windows.advocacy
Subject: Re: Richard Stallman what a tosser, and lies about free software
Reply-To: [EMAIL PROTECTED]
Date: Tue, 08 May 2001 16:03:57 GMT
Said Les Mikesell in comp.os.linux.advocacy on Mon, 07 May 2001 01:03:58
>"T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
>news:[EMAIL PROTECTED]...
>
>> If an API is "like a contract", then it is simply the documentation of
>> the agreement. Not "the agreement" itself.
>
>The API is the agreement - by definition. If you don't understand
>that, there is no point in arguing about what it is.
If I am required to understand anything "by definition", then it is not
worth understanding. It is something to be memorized, not understood.
Certainly the point would not be worth arguing about with someone who
has simply memorized it and considers the concept thus inviolate, but
whether an API "is" the agreement is not as easily hand-waved as you
would like it to be, Les.
That being said, I agree with your definition; it does match my
understanding. An API is every bit as ephemeral as "an agreement",
regardless of how well documented, or how often implemented, it is.
Thus, if there is only one library which presents an API, you have no
option but to agree to that author's terms, or write another library for
that API. Claiming you have the right to use the library, simply
because you have the right to "use the API" (enter into an agreement
with any library author), is bogus. That would be as if I could walk
into a store and steal anything I want, claiming that since I'd be able
to take it if I paid for it, they can't prevent me from taking it.
--
T. Max Devlin
*** The best way to convince another is
to state your case moderately and
accurately. - Benjamin Franklin ***
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss,comp.os.ms-windows.advocacy
Subject: Re: Richard Stallman what a tosser, and lies about free software
Reply-To: [EMAIL PROTECTED]
Date: Tue, 08 May 2001 16:03:59 GMT
Said Les Mikesell in comp.os.linux.advocacy on Sun, 06 May 2001 20:26:49
>"T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
>news:[EMAIL PROTECTED]...
>> >> >
>> >> >The GPL does not allow for any non-GPL'd part, making no distinction
>> >> >about between component types.
>> >>
>> >> Yet when I point out that neither does copyright, you seem to miss the
>> >> point. Why is that?
>> >
>> >Because copyright allows you to obtain the right to use two different
>> >things separately, under different terms, and use them together as
>> >is the case when you obtain a program that uses functions from
>> >a dynamic library or source code that you link yourself to a library
>> >which you are allowed to use. The FSF wants to claim that your
>> >right to use one component is somehow affected by the other. Unless
>> >you copy something that you don't already have the right to copy there
>> >is no basis in copyright law to support that claim.
>>
>> You're switching who and what you are referring to with "use" too often
>> for me to make sense of that. If you base your work of authorship on
>> someone else's work of authorship, it is *derivative*. Is that somehow
>> *too* simple for you to understand?
>
>Not too simple, just untrue in the programming case. Code calling
>functions
>from another library is not a derivative of that library any more than a
>playlist
>is a derivative of the songs mentioned. [...]
Okay, I can live with that. Code calling functions from a library are
not a derivative of the library. That doesn't prevent the program from
being derivative of the library, according to the FSF's position. You
disagree with that position, and you have made the position very
comprehensible. I certainly consider it a reasonable position, when you
state it like that. I still consider it mistaken, based on a
metaphysical concept of copyright which denies the reality of the
conflict between software and copyright which makes it a defensible (and
thus legally correct) position. Just because no literal copying is
going on does not prevent infringement from occurring.
--
T. Max Devlin
*** The best way to convince another is
to state your case moderately and
accurately. - Benjamin Franklin ***
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss,comp.os.ms-windows.advocacy
Subject: Re: Richard Stallman what a tosser, and lies about free software
Reply-To: [EMAIL PROTECTED]
Date: Tue, 08 May 2001 16:04:00 GMT
Said Austin Ziegler in comp.os.linux.advocacy on Mon, 7 May 2001
>On Mon, 7 May 2001, T. Max Devlin wrote:
>> Said Roberto Alsina in comp.os.linux.advocacy on 6 May 2001 15:32:49
>>> Maxie:
>> [...]
>>>>And what is then is "the API itself", but a description of the API?
>>> That�s like saying a paperback of "The Great Gatsby" is a description
>>> of "The Great Gatsby". It makes no sense.
>> I don't see why. It seems to me that a copy of "The Great Gatsby" would
>> be a rather ideal and precise description of "The Great Gatsby". Now
>> ask yourself "is it a description of the intellectual property?"
>
>If you think that, then you're more deluded than I thought. A copy of
>"The Great Gatsby" is a copy of "The Great Gatsby", not a description
>of same. A description is one level removed from the thing itself and
>is "about" the thing.
It certainly isn't the description you would expect to get if you asked
merely for a description. Yet it is the most complete description you
could possibly have. Your observation about descriptions being "one
level removed from the thing" is understandable, but I would point out
that a "label" is also "one level removed from the thing". It is
obvious that some labels would be more or less descriptive than others.
My point is that you are right that any label or description is not the
thing itself.
Yet, if someone were to ask me, "What is the Great Gatsby?", I can think
of no more complete answer I could give him than to hand him the book.
My point here is that there is a rhetorical level which your statement
is true, but that is mere philosophy. In the real world, it becomes
pedantry. The API is nothing but its description.
>This is where I disagree a bit with Ayende: I believe that an API can
>be an API without documentation, but that it's not necessarily usable
>without that documentation. If the items in the API aren't documented,
>however, they're not part of the agreed-upon "contract" between the API
>and the program.
But now this draws a distinction between what "is" the API, and the
contract the API presents. This falsifies the idea that the API is the
contract. But, then, we know the API is nothing but "its contract";
this clearly shows that APIs are metaphysical objects, they don't really
"exist" the way real things do. Real things would include contracts,
descriptions, and books. What really does exist is libraries, and
programs, and whether they can interoperate.
--
T. Max Devlin
*** The best way to convince another is
to state your case moderately and
accurately. - Benjamin Franklin ***
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss,comp.os.ms-windows.advocacy
Subject: Re: Richard Stallman what a tosser, and lies about free software
Reply-To: [EMAIL PROTECTED]
Date: Tue, 08 May 2001 16:04:03 GMT
Said Roberto Alsina in comp.os.linux.advocacy on 7 May 2001 13:45:16
>On Mon, 07 May 2001 05:36:17 GMT, T. Max Devlin <[EMAIL PROTECTED]> wrote:
>>Said Roberto Alsina in comp.os.linux.advocacy on 6 May 2001 15:32:49
>>>On Sat, 05 May 2001 03:26:40 GMT, T. Max Devlin <[EMAIL PROTECTED]> wrote:
>> [...]
>>>>And what is then is "the API itself", but a description of the API?
>>>
>>>That�s like saying a paperback of "The Great Gatsby" is a description
>>>of "The Great Gatsby". It makes no sense.
>>
>>I don't see why. It seems to me that a copy of "The Great Gatsby" would
>>be a rather ideal and precise description of "The Great Gatsby". Now
>>ask yourself "is it a description of the intellectual property?"
>
>Well, it may seem to you, but it does not seem to me, and it probably
>does not seem to copyright law.
If you believe your position is correct, why screw it up by claiming to
speak for "copyright law"?
>For example, descriptions of "The Great
>Gatsby" (let's call it TGG) are not covered by TGG's copyright.
>That's why literary criticism can exist.
Just as an API is not covered by a library's copyright, right? By
analogy, then, it seems obvious why using the API is not always but can
be infringement on the copyright of a library, in the very restricted
case that the FSF specifies. Some descriptions are just literary
criticism; some are attempts to use technology to outwit copyright, by
replicating the book.
JD brought up the idea of Cliff's Notes, of course. Cliff's Notes are
limited to books which are in the public domain, I think. This robs us
of the chance to see how the courts would answer the question of whether
or not they are substantial enough to infringe. I could be mistaken
about that; does anyone have any details?
[...]
>>Yes, that was my point. Therefore, your concept of "the API" is flawed
>>and useless, because it can only be compared to itself. It is either
>>the program or the library; there is no API.
>
>Because something can not be wrong it is flawed and useless?
Yes, all an unfalsifiable claims are false. Flawed and useless. Thus,
if an API cannot be wrong (but only the library implementation or the
documentation) then it is a mythical abstraction. Handy for
understanding the relationship between objects, not an object itself.
>I know
>it would make for bad Usenet debate, but things that are correct
>and can not be wrong are quite useful.
Nothing that "can not be wrong" could ever be considered correct, and
are never useful, either logically or in "the real world". This was
proved by Karl Popper decades ago. It is the fundamental premise of
science: the theories are falsifiable, and only temporarily considered
"correct". It is the math, not the theories explaining why the math is
at it is, that is useful.
>>"There is no spoon." The API can morph its form as easily as the spoon
>>in the movie. Its existence is just as much fiction, then.
>
>A book can morph as easily as the API. Yet the book exists.
I've never seen a book morph. You can use the phrase "the book" as an
abstraction equivalent to "the API", yes. The difference is, the book
IS a real physical object, which cannot morph. The API is not.
>>>>>An API can not be "wrong" like a rock can not be "wrong".
>>>>
>>>>If someone throws a chunk of concrete into a pile under study by
>>>>geologists, they will most surely say that rock is "wrong", don't you
>>>>think? Rocks are physical things; APIs are just platonic objects.
>>>
>>>No, I don�t think they would say the rock is "wrong".
>>>APIs exist materially. They can not be platonic objects.
>>
>>No, APIs don't exist materially, that is my point.
>
>Well, blatant assertion is harly a point. APIs exist. I have several
>copies of them in my desk.
Pointless rhetoric is hardly the point. Your "blatant assertion" that
API's "exist" is noted, as is your failed attempt to prove the point
empirically. The specification of an API (documentation of the API)
cannot be "the API". If it were, APIs would be covered by copyright,
for one thing, because a document describing an API is.
>> They are "a contract
>>for services", a metaphoric (metaphysic, i.e. platonic) objects only.
>>What exist materially (we will say) are libraries and programs, and the
>>necessity to determine if one is derivative of the intellectual property
>>of the other.
>
>Declaring the immateriality of an object is a risky trick.
Being able to claim that metaphysical objects don't exist (yet still
using words to refer to them) is itself the risky trick. It is called
"abstraction", and it is the whole "trick" of language. It is, I think,
what distinguishes "programming languages" from "natural language".
Programming languages are just a complicated form of math; they do not
and can not communicate meaning, as a real language can, because they
are not capable of 'encoding' an abstraction.
>>>>>An API can be inadequate, though, and it can not fulfill its purpose.
>>>>>
>>>>>An implementation of the API can be wrong, because its "platonic object"
>>>>>is the API.
>>>>
>>>>There are no platonic objects in this universe, Roberto, and this is the
>>>>only universe that exists.
>>>
>>>That is why I used quotes. The API is an object that describes what
>>>the implementation should be. The use of "platonic object" was an
>>>analogy.
>>
>>No, you got it backwards. The API is not an object; objects don't
>>"describe things" unless they are platonic objects, i.e., abstractions.
>
>Quoting "Riding the Iron Rooster": "Peking is a sooty city, littered
>with high cranes shaped like inverted Ls, looking down.". "Riding
>the Iron Rooster" is an object. That is a description.
"Riding the Iron Rooster" is a title. Only the book is an object, and
you have obviously not presented that here (it is impossible to transmit
physical objects through the Internet.) And that was a quote, possible
an extract from the book (if chosen for some reason other than to
communicate the meaning of the book itself), or possibly an abstraction
of the book, if you honestly meant to use only that quote to somehow
describe the entire book. Neither is a description of the book.
If you meant that calling it an object is a description, then I don't
understand the point of the quote. Perhaps you meant to illustrate the
difference between an abstraction (the quote) and a description (that
the book is an object). If so, I agree with your point, but I would
point out that simply saying that "Riding the Iron Rooster" is an object
is not a very good description. It is not very concise, since there are
many things that are objects but aren't that book. It is not very
comprehensive, because there are many things which could describe the
book other than simply the fact that it is an object. And finally it is
rhetorically questionable, because the title not a physical object.
>> [...]
>>>An implementation of an API is something that performs the functions
>>>and shows the behaviour required by the API.
>>
>>Documentation of an API fulfills this definition, as far as I am
>>concerned, since the behavior of a correct implementation cannot be
>>different from a correct logical analysis of how it should behave.
>
>That makes no sense.
Then try reading it again, only slower. It made sense when I wrote it,
and it still makes sense now, as far as I can tell. I know its very
complex, but it is, after all, a complex issue, isn't it?
>A logical analysis doesn't behave in any particular
>way. How do you call a function of a logical analysis?
"It", Roberto. The problem you're having is the word "it". The logical
analysis doesn't behave any certain way; *it* behaves a certain way,
according to logical analysis. Get it?
--
T. Max Devlin
*** The best way to convince another is
to state your case moderately and
accurately. - Benjamin Franklin ***
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss,comp.os.ms-windows.advocacy
Subject: Re: Richard Stallman what a tosser, and lies about free software
Reply-To: [EMAIL PROTECTED]
Date: Tue, 08 May 2001 16:04:05 GMT
Said Austin Ziegler in comp.os.linux.advocacy on Mon, 7 May 2001
>On Mon, 7 May 2001, T. Max Devlin wrote:
>> Said Roberto Alsina in comp.os.linux.advocacy on 6 May 2001 15:40:09
>>> T. Max Devlin:
>>>>Said Roberto Alsina in comp.os.linux.advocacy on 4 May 2001 14:24:23
>>>>>T. Max Devlin:
>> [...]
>>>>>What can I say, if you say a tautology is not true, you are by
>>>>>definition wrong.
>>>>That doesn't mean there's no difference between a statement that is true
>>>>and a statement that is unfalsifiable.
>>> Being unfalsifiable and being true can be two faces of the same
>>> statement. The above statement is true. The rest I don�t care.
>> That statement is false; all unfalsifiable statements are, by
>> definition, untrue. This was proven by the philosopher of science Karl
>> Popper several decades ago.
>
>That's not how I recall this being stated. I believe that Goedel stated
>it better, on the whole. While not all unfalsifiable statements are
>axiomatic, axioms are unfalsifiable but true by definition. They may
>not be true in every system, but they true.
I cannot conceive of a useful definition for the word "true" that does
not include being true in every system. I overstate the case when I say
that a statement that is unfalsifiable is untrue, granted. In truth, a
statement that is unfalsifiable is neither true nor untrue. An axiom is
simply useless for actually explaining anything, although you can spend
a great deal of time failing to explain things using axioms.
--
T. Max Devlin
*** The best way to convince another is
to state your case moderately and
accurately. - Benjamin Franklin ***
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list by posting to comp.os.linux.advocacy.
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Advocacy Digest
******************************
