Date: Fri, 15 Jun 2001 13:22:44 -0400 From: Bill Moran <wmoran@iowna.com> To: Eric Rescorla <ekr@rtfm.com> Cc: freebsd-questions@FreeBSD.ORG Subject: Re: Justification for using FreeBSD Message-ID: <3B2A4464.B709F6FD@iowna.com> References: <200106151635.JAA27943@romeo.rtfm.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Eric Rescorla wrote: > > > Eric Rescorla wrote: > > > > > > > I beg to differ. > > > > DHCP standards say that a machine that reconnects requests that it be > > > > reassigned the IP it had prior to disconnecting. If the IP is still > > > > available, the server should reassign the same IP. Thus, if you're only > > > > disconnected for a few minutes, the chances are very good that you'll > > > > get the same IP and the transmission will complete. > > > My experience differs from you. I almost enver get the same IP > > > address and hence this doesn't work. > > > > And your point is? > > Regardless of whether or not you get the same IP, TCP is still a > > reliable protocol and aborting a connection just because a physical link > > goes down for a few minutes is still a violation of that protocol. > > Period. > > My point is that if you are in an environment where you aren't likely > to get the same IP address again then this violation of the protocol > actually doesn't make very much difference to you since you'd lose the > connection in any case. I agree with this statement, if you're in such an environment. > Sure, Windows violates the TCP specification. So what? Unices do this > all the time as well (Check out TCP/IP Illustrated 2, Appendix C for a > long list of the ways that 4.4 and Net/3 violated the requiements > RFCs.) I should review that. I'll probably never get around to it, but I should. > The metric isn't whether you're standards compliant, since noone is, No, the original question that prompted this post was: "In what ways are M$ products not standards compliant" That particular example was one of 3 I gave. > but rather whether you're standards compliant in a way that's really > inconvenient. I maintain that in most cases you won't get the same IP > and that therefore this particular instance of noncompliance isn't > very interesting. Every ISP I've ever subscribed to (or been involved with) will give you back the same IP more than half the time. If you get disconnected and dial back right away ... the probability goes up to 90% or better (since most DHCP servers use a "least recently used" algorithm to determine when to recycle IP addys) As far as inconvenient ... I found it very inconvenient to have phone company problems require me to start big downloads over and over again. In the end, my personal solution was to start using FreeBSD instead. Eventually, I got tired of the lack of speed and got cable modem service, but that's another story and another potential discussion. > This isn't to say that Windows's TCP/IP stack isn't lousy. It is. Hmmm ... I don't 100% agree on that broad statement. It's "acceptable" as far as I'm concerned. (Depending on which version you're speaking of, the older Win95 PPP implementations were so far below lousy that I won't attempt to describe them) > I just don't think that this particular complaint is very interesting. Umm ... OK. -- If a bird in the hand is worth two in the bush, then what can I get for two hands in the bush? To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3B2A4464.B709F6FD>