View  Info
 

Podcast 075

Revision #2, 11/23/2009 1:56 PM
User
Tags: (None)

Previous Next 

Podcast 075

Revision #6, 12/3/2009 5:40 AM
79.193.204.64
Tags: (None)

Previous Next 

Spolsky: edit me!

Atwood: edit me!

Limoncelli: edit me!

(1:06)

Spolsky: Our guest today is Tom Limoncelli, who is the writer of The Practice of System Network Administration and also Time Management for Systems Administrators, recently out from O'Reilly. Right? Is that a recent book?? I haven't... I'm so behind on my reading. (laughter)

Limoncelli: The second edition of the Systems Network Administation book is new -- it's about a year old. The Time Management book is about four years old.

Spolsky: Right. That's sort of an old classic... older book. Well, not that you're that old, but that book is like... the first one came out in...?

Limoncelli: 2001

Spolsky: ... and so there's a new second edition that's out, so if you have the first edition, it's all wrong. Buy the second edition.

Limoncelli: Absolutely.  We made so many changes to the second edition.  First of all, all of the anecdotes that were about Solaris we now say they are about Linux.  You know, appeal to the young crowd.

Spolsky: Right.  Does the young crowd even know what Unix is?  If you said that, would it just not make any sense?

Atwood: So, Joel, you should also mention that Tom wrote The Complete April Fools' Day RFCs.

Spolsky: Ah!  Oh yeah, look at that.  Here it is on his website.

Atwood: Well, he's co-author, but.  That's more of a collection, right?  Did you actually write April Fools' day RFCs?

Limoncelli: I have never written April Fools' day RFCs.  It is a collection of the RFCs that are available for free, but the book has the advantage that Peter Salus (my co-editor) and I added some interviews and historic background to some of them.  And actually, there are three forwards to the book, because we asked three famous internet luminaries in parallel, figuring that all three of them would say no, but all three of them said yes.  So we published all three.

Spolsky: Does that include the famous "TCP/IP over carrier pigeon"?

Limoncelli: Yes.  And pictures of its first implementation by some people in Europe.

Spolsky: Oh yeah, that was much later, long after the RFC, I think was the first implementation.

Limoncelli: Yes, almost 20 years later.

Spolsky: It takes a whole new generation of computer science students to actually try some of these crazy protocols.  So that sort of disqualifies it as an April Fools' day RFC if it was implemented (laughter)

Limoncelli: Well, somebody recently published an RFC where they implemented IPv6 over social networks, and the first implementation is a Facebook plugin or app and you configure your little IPv6 router and every time you click "update" it sends out packets and you eventually build up a little IPv6 network.  So it's for modern times, a modern protocol.  Modern times is social networking, not carrier pigeons, and its a modern protocol.

Spolsky: And this is interesting...why?

Limoncelli: Because geeks are bored, and this is what we do.

Spolsky: Yeah.  There are a lot of geeks implementing one thing in terms of another thing, that's so interesting.

Limoncelli: Yes, like IP over DNS, and then DNS over IP over DNS.

Spolsky: Right.  IP over DNS is actually kindof a good way to get through various firewalls that may be blocking you.  Like, you're in a hotel room and you haven't paid for the internet access, but DNS is still working for some reason.

Limoncelli: Yes.  And recently some malware has been using DNS's signalling mechanism.  So the key loggers that they're finding out there are reporting back your keystrokes over DNS.  They're querying for special URLs that mean "this person just pressed the A key; this person just pressed the B key."  And then they do another query that means "should I start attacking?"

Spolsky: Wow.

Limoncelli: Yeah, scary stuff.

Atwood: Well, the April Fools' Day RFCs go way, way back.  In fact, I'm just looking here and the first one was 1973: "Arpawocky".

Spolsky: edit me!

Limoncelli: edit me!

Atwood: edit me!

Spolsky: edit me!

Limoncelli: edit me!

Atwood: edit me!

Spolsky: edit me!

Limoncelli: edit me!

Atwood: edit me!

The demonstration with the outline mode and the mouse that one of you was referring to was performed by Doug Engelbart and a team from SRI (not Xerox PARC) in 1968: 

http://en.wikipedia.org/wiki/The_Mother_of_All_Demos

-- Alison Chaiken