I've been considering writing a fully asynchronous ATA over Ethernettarget driver (I'm not entirely happy with the current one and itsperformance). If Linux asynchronous IO worked on sockets (specificallyraw network sockets), there is a nice simple design where you just setup a pool of buffers and then cycle each one through a little statemachine (network in to disk IO to network out).

(This works especially great for AOE because the protocol is both basedon raw packets and completely unordered,

In an aside in a recent entry , I wrote:

With sufficient cleverness, you can construct a version of capture() that is passed store 's name and puts the result in it directly,without needing you to make it an array and use store[0] .

I'm wrong. In Python, no outside power can change a function's localvariables (more precisely, no outside power can change the namebindings , doing the equivalent of ' localvar =something '). While you can

First I'll show you the symptoms, then I'll explain what I did to shootmyself in the foot:

# zpool create tank01 c0t38d0# zpool create tank02 c0t39d0# zpool replace tank01 c0t38d0 c0t42d0

(Time passes, the resilver finishes, and zpool status tank01 shows nouse of c0t38d0.)

# zpool attach

There's a lot of times where a function wants to return more than aplain boolean; for example, validation routines often want to returnsome sort of explanation of the validation error (if there is one).At the same time, the most natural way to use the function is oftenin the flow of a conditional:

if blah(o):    ....elif not validates(o):    # whoops, lost the error reason

There's a variety of traditional answers to this,

C's !! double negation idiom is a way of compressing all non-zerovalues of an expression down into a 1. There's a number of things thatyou can use to get the same effect in Python:

• if you know that your result is always zero or positive, you can use cmp() . I'm ambivalent about whether this is obscure usage; you'reat least making it clear what you're doing, although probably not why.

  (If your result can be

The number of bits of information in a password are a function of thealphabet that the password is drawn from and how many characters longit is. The formula is:

nbits = ceil(log2(len(alphabet)) * nchars)

So what does that mean? Let's take the case of 8 character longUnix passwords, and do a table:

Sort of presented in the traditional illustrated format:

; sdig ns ibc.com.au.ns1.ibc.com.au.ns2.ibc.com.au.; dig cname ibc.com.au. @ns1.ibc.com.au.[...];; flags: qr aa; QUERY: 1, ANSWER: 1, [...][...];; ANSWER SECTION:ibc.com.au. IN SOA ns1

Here is a useful safety tip if you are constructing HMAC signatures for structured messages,things that you want to parse later: put a format version number intoeither the key or the message.

If you want messages in the old format to fail signature checks so thatthe rest of your code doesn't have to deal with them, put the versionnumber into the key. Put it into the message if you want your code togracefully deal with several generations of message formats; each willstill verify, but the

This week, we:

• got 9,123 messages from 254 different IP addresses.
• handled 17,076 sessions from 1,364 different IP addresses.
• received 264,864 connections from at least 70,143 different IPaddresses.
• hit a highwater of 12 connections being checked at once.

Volume has dropped compared to last week ,including total messages, which surprises me a bit. As we can see inthe per-day table