In the last entry I wrote that some full diskencryption schemes could cause total data loss on occasions other thanyou forgetting your keys (or having the entire disk die). A commentatorthen suggested dealing with the key loss problem this way:

Many systems support multiple keys (e.g. dm-crypt). Just keep onesealed in a safe.

Unfortunately, this is a great way to increase your chance of total dataloss (in one of the ways that I was thinking of). In fact

Via Hacker News I saw that theEFF's suggested New Year's resolution is full disk encryption onevery computer you own .When I saw this I did something halfway between wincing and grindingmy teeth. The issue is that full disk encryption has a significantdownside, and because of this downside choosing full disk encryptionis making a much deeper choice than you might think.

The downside is simple: the failure mode for meaningful full diskencryption is catastrophic data loss . If you forget your passwordit is game over

In reaction to my entry on Python 3 dropping the comparison functionfor sorting , some people may feel that asorting order that is neither simple field-based nor based on a computed'distance' (the two cases easily handled by a key function) isunrealistic. As it happens I can give you a great example of a sortorder that cannot be handled in any other way: software package versionson Linux systems.

For simplicity (and because I know RPM best), I'm going to talk about

To summarize aggressively, one of the quiet long-running disputesin the OO test driven community is between classical TDD, where youuse real or stub support classes, and mockist TDD, where you usebehavior-based mock objects. My guide on this is primarily MartinFowler (via Jake Goulding ).Jake Goulding summarizes the difference as stubs assert state whilemocks assert messages (or method calls). I'm mostly a classicalistas far as testing goes for no greater reason than I generally findit easier,

A whole lot of things have been written this year about better versionsof SSL designed to get away from the many practical weaknesses of thecurrent CA-based SSL model (you know, the one where CAs have terriblebusiness practices, get compromised, and get leaned on by governments).Unfortunately, I don't think that any of them are likely to catch onany time soon, because of what is essentially the inverse of the falsepositives problem .

The reality is that forged SSL certificates are in general a

One of the changes that Python 3 has made is that, to quote the documentation :

builtin.sorted() and list.sort() no longer accept the cmp argument providing a comparison function. Use the key argumentinstead. [...]

I feel unreasonably annoyed about this change. At least on the surfacethere's no obvious reason why; basically all of the uses of a comparisonfunction I've ever used are to pick a specific field out, and that'shandled much better

There's a vague mythology that spammers take holidays off, or at leastsome spammers do. I decided I was curious enough to see if this wastrue, at least as far as our own data goes, so I looked at our filteringsystems's statistics. This actually gets into interesting issues,because Christmas fell on a weekend this year plus you'd expect somedegree of day to day fluctuation.

(For bonus complication, a month ago includes American Thanksgiving,which spammers might also partially take off

There are any number of blog packages (and blog design templates)that don't make the date of entries very obvious; it's not in theURL, it's not at the top of the page, and perhaps it's hiding insmall type down at the bottom of the page. I've come to feel thatthis is a significant mistake, because of a combination of realblog usability and how blogs are written .

How blogs are written means that you probably never go back to

I've talked about Python 3 from the perspective of a Unix sysadmin and Python 3 from the perspective of someone withexisting Python 3 code ; now it's time for thefinal viewpoint, that of someone writing new code.

There are a bunch of practical difficulties with this, things likehaving Python 3 installed on machines and third party modules beingported to Python 3, but they're either gone or going away (and most ofwhat I write doesn't depend on third party modules). Ignoring