One would assume that the software was not downloaded automatically. A
person, somewhere, decided that he/she wanted the software to begin with.
Presumably, this person is bright enough to know what the expected behavior
of the software is. This person, then, would not reject a web server on the
basis of the halting problem. Instead, this person might run other tests;
like, say, a program that performs side-effect detection (in my opinion,
more people are interesting in whether a given program will reformat their
hard drive than in whether the program will terminate or not. After all,
'kill -9' is fairly easy to type in. But 'give_me_back_my_data' doesn't
have a man page, as far as I can recall). The last thing we want is for our
computers to, willy-nilly, download random software and run it. Want we
want are tools to allow the HUMAN who downloadrf a program to be able to
determine if there are logic bombs hidden inside it. Open source is great,
but really, who has the time? And then look at all those Microsoft binaries
that I so *desperately* want to run...
--Guy
-----Original Message-----
From: Mike Schiraldi [mailto:[EMAIL PROTECTED]]
Sent: Friday, May 11, 2001 3:12 PM
To: [EMAIL PROTECTED]
Subject: Re: [freenet-tech] Off-Topic: Ideas for a FreeNet Process
> Time is a very good indicator of failure.
> If things did not time out the whole internet would be deadlocked in a
day.
Okay, let's say the timeout was one hour. If i send you a craftily-disguised
endless loop, and you run it for one hour, you can say with some certainly
that it's an endless loop, right? Sound good so far ... you kill the
process, and free up CPU cycles for other, useful processes.
But what is a useful process?
Your original example was to run a web server over this. So a nice innocent
person sends you a web server to run. Now, what happens after an hour? Your
computer says, "Gee, this process has been running for an hour; it must be
an endless loop; let's kill it."
--
Mike Schiraldi
Verisign Applied Research
_______________________________________________
freenet-tech mailing list
[EMAIL PROTECTED]
http://lists.freenetproject.org/mailman/listinfo/tech
_______________________________________________
freenet-tech mailing list
[EMAIL PROTECTED]
http://lists.freenetproject.org/mailman/listinfo/tech
