Re: What makes software copyrightable anyway?
Raul Miller wrote:
>On 5/11/05, Humberto Massa <humberto.massa@almg.gov.br> wrote:
>
>>Raul Miller wrote:
>> >On 5/11/05, Humberto Massa <humberto.massa@almg.gov.br> wrote:
>> >>We have a "brocardo" (legal axiom) in our doctrine: "He who can
>> >>do more, can do less" (horrid translation to "quem pode mais,
>> >>pode menos" ["Quién puede más, puede menos" in Spanish]). So,
>> >>if the binary is the result of an automated,
>> >>non-easily-reversible process over the source, and if you grant
>> >>me the right to distribute the source (ie, I can do more), you
>> >>are implicitly granting me the right to distribute the binary
>> >>(ie, I can do less). Now, you can in the conditions to your
>> >>grant, explicit that if I am distributing it in another form,
>> >>then I must distribute it in the source form (or do at least
>> >>one of the things section 3 enumerates, in the case of the
>> >>GPL).
>> >
>> >Ok, this doesn't sound like the source is equivalent to the
>> >binary but that the binary is protected the same way the source
>> >is. In other words, you probably have a special legal category
>> >for "derivative which is not creative enough to be granted
>> >copyright protection which is distinct from the original".
>>
>>Not really. The work is the content, not the form. Suppose two
>>(dead-tree) copies of my book "How to Die...", one being a
>>hardcover, and the other begin a paperback. Suppose further that
>>the imagery in the cover, etc are identical in both books.
>
>
>This doesn't sound like an application of that legal principle.
>
>You don't get a paperback from a hardback, or a hardback from a
>paperback. You produce each independently.
Intelectually, no. You don't produce either, intelectually. You just
print them, physically. What you do produce, intelectually, is the
content of both (which is the same in casu, and you produced only
one time).
>
>This legal principle seems more analogous to: If the hardback is
>protected by copyright then pages ripped out of the hardback are
>protected by copyright. The pages ripped out are not equivalent to
>the hardback, but they carry the same protections.
>
That is your mistake: it's not the pages that carry protection, it's
the words and illustrations on the pages (as in abstract,
intelectual entities) that carry protection.
See, in 9610/98, art 7, caput, we have ''Intelectual works protected
[by this Law] are the creations of the spirit, expressed by any
means and fixated in any medium or support, tangible or intangible,
known or to be invented in the futures, as for example:'' (hideous
translation and braces are mine). This is AFAIK boilerplate text
copied directly from the Berne text (I may be wrong in this account,
tho).
>>I am not convinced of this, either: the word "transformation" was
>>in the quotation of 17USC101 WRT derivations. And no case law
>>presented in those threads lead me to believe that automated
>>transformations are to be regarded as derivative works.
>
>
>Well... the GPL does clarify what it means by a "work based on the
>Program" in a fashion which I believe does incorporate transformed
>works. However, you've asserted that this would not be the case
>under Brazilian law.
>
Yes, I believe that the "namely" part does not carry any weight in
any law, but I am pretty sure of it under Brazilian Law.
>If that's so, it might very well be the case that under Brazilian
>law there are anthologies which are not considered, under law, to
>be derivatives.
>
Yes. Anthologies are not derivatives under Brazilian Law. They carry
copyright protection, too, but they are definitively not
derivatives.
>
>In this situation, for the GPL to grant you license to make copies,
>you would have to rely on the GPL's grant to make verbatim copies
>of the program. If the transformations needed to build the program
>are not considered verbatim under Brazilian law, then the GPL fails
>to grant license to make copies in those cases.
No, because of the principle that if you can do copies of the
source, and aggregate it with other stuff, then you can apply
automated transformations on it. He who can do more (copy the
source, redistribute), can do less (copy the binary, redistribute)
as long as the conditions are met.
>
>If I sincerely belived this to be the case, I'd consider it a bug
>in the GPL and would report the details of why this is likely to be
>the case to the FSF.
I do. And I consider it a bug in the GPL (as opposed to the stated
intent of it expressed in the FSF GPL FAQ that it should prohibit
linking non-GPLd material). But I also think it is a good bug.
The only way to correct it would be to specify exactly what
operations you can or not do (like static linking, dynamic linking,
remote linking via DCOM/CORBA/SOAP/your preffered RPC interface,
etc) and this is just Not Good.
Keep in mind that if you put any thing that could be construed as an
additional restriction in the GPLv3, it is automatically
incompatible with the GPLv2, and this is a problem WRT very large
GPLv2'd projects without copyright assignment (like Linux).
>
>Of course, if the transfomed copies are considered to be verbatim
>copies under Brazilian law, then you have no problem. Or, if
>permission to distribute verbatim copies always implies permission
>to distribute anthologies which includes some part of those
>verbatim copies, then you have no problem.
I have to think better, and maybe to research some more, because you
used the word "verbatim" in this paragraph 50% more than it is used
in the entire text of the GPL. I'll get back to you.
>
>>Maybe I was more clear now?
>
>
>I think so. Thanks,
>
--
HTH,
Massa
Raul Miller wrote:
>On 5/11/05, Humberto Massa <humberto.massa@almg.gov.br> wrote:
>
>>Raul Miller wrote:
>> >On 5/11/05, Humberto Massa <humberto.massa@almg.gov.br> wrote:
>> >>We have a "brocardo" (legal axiom) in our doctrine: "He who can
>> >>do more, can do less" (horrid translation to "quem pode mais,
>> >>pode menos" ["Quién puede más, puede menos" in Spanish]). So,
>> >>if the binary is the result of an automated,
>> >>non-easily-reversible process over the source, and if you grant
>> >>me the right to distribute the source (ie, I can do more), you
>> >>are implicitly granting me the right to distribute the binary
>> >>(ie, I can do less). Now, you can in the conditions to your
>> >>grant, explicit that if I am distributing it in another form,
>> >>then I must distribute it in the source form (or do at least
>> >>one of the things section 3 enumerates, in the case of the
>> >>GPL).
>> >
>> >Ok, this doesn't sound like the source is equivalent to the
>> >binary but that the binary is protected the same way the source
>> >is. In other words, you probably have a special legal category
>> >for "derivative which is not creative enough to be granted
>> >copyright protection which is distinct from the original".
>>
>>Not really. The work is the content, not the form. Suppose two
>>(dead-tree) copies of my book "How to Die...", one being a
>>hardcover, and the other begin a paperback. Suppose further that
>>the imagery in the cover, etc are identical in both books.
>
>
>This doesn't sound like an application of that legal principle.
>
>You don't get a paperback from a hardback, or a hardback from a
>paperback. You produce each independently.
Intelectually, no. You don't produce either, intelectually. You just
print them, physically. What you do produce, intelectually, is the
content of both (which is the same in casu, and you produced only
one time).
>
>This legal principle seems more analogous to: If the hardback is
>protected by copyright then pages ripped out of the hardback are
>protected by copyright. The pages ripped out are not equivalent to
>the hardback, but they carry the same protections.
>
That is your mistake: it's not the pages that carry protection, it's
the words and illustrations on the pages (as in abstract,
intelectual entities) that carry protection.
See, in 9610/98, art 7, caput, we have ''Intelectual works protected
[by this Law] are the creations of the spirit, expressed by any
means and fixated in any medium or support, tangible or intangible,
known or to be invented in the futures, as for example:'' (hideous
translation and braces are mine). This is AFAIK boilerplate text
copied directly from the Berne text (I may be wrong in this account,
tho).
>>I am not convinced of this, either: the word "transformation" was
>>in the quotation of 17USC101 WRT derivations. And no case law
>>presented in those threads lead me to believe that automated
>>transformations are to be regarded as derivative works.
>
>
>Well... the GPL does clarify what it means by a "work based on the
>Program" in a fashion which I believe does incorporate transformed
>works. However, you've asserted that this would not be the case
>under Brazilian law.
>
Yes, I believe that the "namely" part does not carry any weight in
any law, but I am pretty sure of it under Brazilian Law.
>If that's so, it might very well be the case that under Brazilian
>law there are anthologies which are not considered, under law, to
>be derivatives.
>
Yes. Anthologies are not derivatives under Brazilian Law. They carry
copyright protection, too, but they are definitively not
derivatives.
>
>In this situation, for the GPL to grant you license to make copies,
>you would have to rely on the GPL's grant to make verbatim copies
>of the program. If the transformations needed to build the program
>are not considered verbatim under Brazilian law, then the GPL fails
>to grant license to make copies in those cases.
No, because of the principle that if you can do copies of the
source, and aggregate it with other stuff, then you can apply
automated transformations on it. He who can do more (copy the
source, redistribute), can do less (copy the binary, redistribute)
as long as the conditions are met.
>
>If I sincerely belived this to be the case, I'd consider it a bug
>in the GPL and would report the details of why this is likely to be
>the case to the FSF.
I do. And I consider it a bug in the GPL (as opposed to the stated
intent of it expressed in the FSF GPL FAQ that it should prohibit
linking non-GPLd material). But I also think it is a good bug.
The only way to correct it would be to specify exactly what
operations you can or not do (like static linking, dynamic linking,
remote linking via DCOM/CORBA/SOAP/your preffered RPC interface,
etc) and this is just Not Good.
Keep in mind that if you put any thing that could be construed as an
additional restriction in the GPLv3, it is automatically
incompatible with the GPLv2, and this is a problem WRT very large
GPLv2'd projects without copyright assignment (like Linux).
>
>Of course, if the transfomed copies are considered to be verbatim
>copies under Brazilian law, then you have no problem. Or, if
>permission to distribute verbatim copies always implies permission
>to distribute anthologies which includes some part of those
>verbatim copies, then you have no problem.
I have to think better, and maybe to research some more, because you
used the word "verbatim" in this paragraph 50% more than it is used
in the entire text of the GPL. I'll get back to you.
>
>>Maybe I was more clear now?
>
>
>I think so. Thanks,
>
--
HTH,
Massa
Reply to: