Re: Help with legal audit (yes - this is directed to boost!)

From: williamkempf_at_[hidden]
Date: 2001-11-29 13:23:47

--- In boost_at_y..., Sean Parent <sparent_at_a...> wrote:
> Having had it pointed out to me that this topic had been discussed 
recently,
> I went back and read all the messages on the topic and would like 
to make
> some concrete suggestions:
> 
> 1. For the purpose of making it easier to locate copyright notices. 
All
> copyright notices in files should be delimited with:
> 
> <copyright>
> </copyright>
> 
> These may appear anywhere on a line so they can be proceeded by 
comment
> characters or other necessary quoting mechanism.

I don't like this.  These are HTML like tags which will cause 
problems in HTML documentation and will just look silly in code 
comments.  Why not just require either the copyright symbol © or (C) 
(I realize the (C) form has no legal significance, but it also 
doesn't invalidate the copyright statement, is universally (?)
understood by readers and is easy to grep for).  You can then simply 
search for either of these in the files and find the relevant 
copyright notice.

> 2. For the purpose of distilling copyright notices to a unique 
list. Between
> the copyright delimiters there should not be any text which is not 
part of
> the notice. Items such as the file name, or change history should 
not appear
> within the delimiters.

White space is enough of a delimiter, IMHO.
 
> 3. For easy review. License files should be in an easy to read text 
format -
> preferably plain text but html, pdf, or rtf are acceptable.
> 
> 4. For locating license files - the name of the file should contain 
the word
> "license".

I'd go a step further and require a license.txt file for each 
library, UNLESS the library uses a (hypothetical) standard Boost 
license.  A directory tree search for license.txt would then reveal 
all libraries with unique licenses and would allow easy review by 
legal types.

Bill Kempf