Date: Wed, 25 Sep 1996 15:02:23 -0500 (CDT) From: Joe Greco <jgreco@brasil.moneng.mei.com> To: terry@lambert.org (Terry Lambert) Cc: wollman@lcs.mit.edu, terry@lambert.org, jhs@FreeBSD.org, current@FreeBSD.org, commercial@FreeBSD.org Subject: Re: Licensing Software Message-ID: <199609252002.PAA08640@brasil.moneng.mei.com> In-Reply-To: <199609251759.KAA06301@phaeton.artisoft.com> from "Terry Lambert" at Sep 25, 96 10:59:23 am
next in thread | previous in thread | raw e-mail | index | archive | help
> > No, do not use any form of network address for machine > > identification. This sort of idiocy is one of the most significant > > barriers to the IPv6 transition and to flexible renumbering in > > general. Network addresses are just that, ADDRESSES. They are > > subject to change at will, without notice, and without any recourse on > > the part of the licensee. (Some of the other proposals for IPng would > > have separated the addressing and identification completely, but > > unfortunately this did not happen.) > > The return from the hostid function is supposed to be a 32 bit value. > > One of the biggest barriers to IPv6 transition? I don't think so; > why would it be a barrier? One of the biggest barriers to IPv6 is the people who are going to cling to their IPv4 IP addresses for reasons such as legacy license servers. (I don't know if that is true or not, but I suspect it is more true than not)... I think that was what he was sayin'. > Flexible renumbering in general? Yes, I'll admit it's a barrier > to flexible renumbering. Under what circumstances would you want > to allow a license host to "flexibly renumber"? To hide the > licenses from Billy-Bob? It makes no sense. When the customer is assigned an address block out of an ISP's CIDR block and wants to change ISP's because the old ISP is out of business? There is a definite need to be able to flexibly renumber. > If you are talking about renumbering for some other scheme than > transient connectivity and/or temporary address assignment (neither > of which should apply to a license hosting host), I'd suggest allowing > renumbering under software control of fixed installations is silly > and unnecessary and is the reason we have non-numeric machine "names" > associated with the tuples in the first place: so we can move the > tag around without moving the tuple. > > > > DEC's license manager does not identify machines at all. You can in > > fact just copy /var/adm/lmf/ldb from one machine to another and it > > will work just fine. (We can legally do this because we have a site > > license.) > > But if you didn't, you could still illegally do it. The point is > that you want a software barrier to illegal activity. > > > > If PCs had some sort of IEEE 802.x address burned into them > > as a sort of serial number, you might be able to do this, but they > > don't, so you can't do this, either. > > I don't know about your machine, but mine has one: > > # dmesg | grep de0 > de0 at pci0 dev 6 function 0: DC21040 [10Mb/s] pass 2.3 > de0: Ethernet address 00:80:48:e8:1b:b1 > ------------------------------***************** > de0: enabling 10baseT/UTP port > de0: interrupting at irq 11 > de0 <Digital DC21040 Ethernet> rev 35 int a irq 11 on pci0:6 > de0: DC21040 [10Mb/s] pass 2.3 Ethernet address 00:80:48:e8:1b:b1 > --------------------------------------------------------***************** > de0: enabling 10baseT/UTP port > > The problem with using this value is that it is larger than 32 bits. Of course, when you switch Ethernet cards, you are screwed. Although I will tend to think that's a better solution than IP address :-) The PC isn't suited to this. It has no hardware to do it. And even on machines where there is hardware to do it, node locked licenses suck. ... JG
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199609252002.PAA08640>