> Not comparable. Weights and measures have statutory force; web
> specifications don't.
Just out of curiosity, what statute is that?
I worked for several years ago as a Quality Engineer and then Supervisor and I know that the ISO 9000 quality system standard requires companies to have measurement systems traceable to NIST, but I am not aware of any law that requires this.
If there is a statute that governs metrology, then I can easily construct another example.
> The standards define the technical meaning of the events that occur. They
> don't (and don't purport to) define the legal meaning. These are two
> separate semantic domains, which can inform one another but are not
> equivalent.
Contracts and licenses are written in whatever language most suits the negotiators. The law gives force to the meaning of the words as used.
I could grant you a licence to my copyrighted works in Klingon if I so chose. I would expect that if a dispute over the meaning of the licence arose that a judge would proceed by finding the official Klingon dictionary, determining as a finding of fact what the meaning of the terms we used was, and then deciding as a matter of law how to apply the Copyright statutes when two people had agreed as we had. Change Klingon to HTTP and this is exactly what I am advocating.
> Because the http request and response headers are not about an agreement
> between people, but a transaction between two pieces of software - a server
> and a user agent. It can be argued that they are evidence of an agreement
> between people, but not to the degree that a signed contract is.
That is really, really weak. People use software all the time to facilitate agreements. Their intent gets translated into 1's and 0's and eventually pops out the other end with the same meaning it had when it was put in. If along the way, it is stored in a capturable form, that's a record of their intent. Billions of dollars change hand every day based on contracts formed by one party's software talking to another party's software. You just invalidated most of the banking industry. If I use a bot to bid on ebay are you going to tell me I'm not contractually obligated to pay? In copyright land its black letter law: Title 17 now explicitly recognizes that access licences can be coded in technological measures.
> It will never get that far; nor has such a legal principle been tested. It
> is, I submit, part of a "common understanding", and I agree that it's
> almost certainly not worth pursuing. But people do think in terms of their
> "website". As you know, you will find nothing about websites in the W3C
> specifications. Domains have meaning only in terms of the DNS/URI
> specification. And as far as I know the W3C specifications take as implicit
> another piece of information which would be germane to this, the file
> permissions which typically come from the server operating system.
It's not part of my common understanding. A "web site" is the union of all named resources exposed whose URI's begin with a common domain name. Many (most) websites do not have a natural starting place. That's why it is called a **WEB** site, because there is no "top".
You identified an important point about the OS permissions. The web server runs with the authority of some OS account. Often this is a special account that exists for no other purpose. It cannot serve out any file that it cannot access. By putting a file in the right place with the correct permissions, the copyright owner commits an act that authorizes the serving of those documents by the software.
In fact, the OS "read" flag, according to the POSIX specs grants authority to copy the file. If you want to reject my HTTP argument, apply the same reasoning to the following case. You and I are users on the same multiuser system, say at a univeristy or our mutual ISP. I create a file with my unpublished novel in it. I mark that file as world readable and place it conspicuously in a public directory. I don't know you and have never negotiated anything with you. I examine the public directory and see the file, note that the world was granted read permission, and copy the file to my own directory, where I later read it a few times, and forget about it. Have I commited copyright infringement against you? The HTTP game is exactly the same with different technical details.
This archive was generated by hypermail 2.2.0 : Mon Mar 26 2007 - 00:35:49 GMT