Re: Is the BitKeeper network protocol documented?

[David Schwartz]

On 20 Jan 2003 09:28:35 -0500, Dana Lacoste wrote:

>On Sun, 2003-01-19 at 20:05, David Schwartz wrote:

>>    Don't blame me. The GPL just says the "preferred" form and 
>>leaves
>>us
>>to wonder. As I understand it, however, you cannot ship binaries of
>>a
>>GPL'd project unless you can distribute the source code in the
>>"preferred form .. for making modifications to it".

>The GPL specifically allows for multiple methods of accessing the
>'preferred form' including FTP, including the source in the
>distribution, etc.  BitKeeper is nothing more than another method
>to access that 'preferred source'.

	I think you're entirely dropping the context. If the development of 
a project is centered around a version control system, then that 
version control system contains metainformation that is useful when 
you're making modifications.

	In this case, the raw source code, less the change history and check 
in comments, would not actually be the preferred form of the source 
code for the purpose of making modifications. This has nothing to do 
with how you get the information but what information you get.

>Please stop this.  You're looking kind of silly here.

	Only because you are misrepresenting my argument.

	Let me give you a hypothetical. There's a program and you have to 
make some changes to it. Would you prefer to have the raw source code 
or the source code with change history and commit comments? I'm not 
talking about how you get either set of information, I'm talking 
about what information you get.

	Checking code out of a repository is as much an act of obfuscation 
as stripping comments.

>PS: nobody said 'IANAL' yet.  meaning you're just a noisy peanut
>gallery

	Any lawyer who claimed he or she could predict how a court would 
interpret this clause of the GPL is lying to you. That is why it is 
essentially impossible to be sure you comply with this.

	DS