Here is something that I've more or less written about before, but Iwant to reinforce by saying explicitly:

Asking users questions never increases security.

Never ever. Really.

(See SecurityChoiceProblem for a discussion of why.)

What this means is simple. Every time you design a system where part ofthe design is 'if something questionable happens, we will ask the userif they approve it', assume that at least half of your users will makethe wrong choice. Ask yourself what this does to

As I alluded to in a comment here ,one reason that we can't expect people to understand that 'low qualityencryption' is basically 'no encryption' is that browsers don't. In factthe issue shows why having a distinct and specially marked 'https' isultimately a terrible mistake; it has trained programmers and users thathttps means security, when in fact it doesn't.

What would have been much better is if browsers hid the differencebetween the raw protocol being used and instead showed

My perspective is that I have always had a two-fold interest inOpenSolaris. One part of my interest was actually in Solaris sourcecode (for which OpenSolaris was the closest available substitute),and the other part was an alternative to Solaris if Sun went crazy.

I never truly trusted OpenSolaris for two reasons. First, because Inever fully trusted the depth of Sun's commitment to open source; italways felt like a tactical move, instead of a heartfelt conviction,and tactical moves are subject to revocation

One of the things that comes up every so often is that the Fedora peoplewould kind of like more people to use Rawhide, their rolling developmentversion, so that it gets more testing and all of the problems don'tsuddenly come out of the woodwork when a new version of Fedora isreleased (or so I heard once upon a time; maybe they have enough testersnow). Although I've written about this before , I'verecently had a new modest idea on how to make a version

In my entry on the lifetime of our fileserver infrastructure , I said that 10G Ethernet wasn'tsomething that I expected to force hardware changes on us in the nextthree years. The reason for this is pretty simple: it is too expensivenow.

Well, sort of. What is actually important is what happens because 10GEthernet is so expensive, namely that it is vanishingly rare today.Rarity matters in several ways. First and most obviously, by defaultrarity means low or no demand for

Today, I was considering reconfiguring a program to stop supportingrelatively less secure SSL connections and found myself thinking the oldrefrain of 'well, maybe there's some client that only works with the oldstuff, isn't it better for them to have some encryption rather than noencryption?'

This is a superficially attractive thought, and it's certainly an easyrefrain to fall into. It's also wrong.

From the semi-mathematical security perspective, it sounds good; havingweak encryption defeats some

In the previous entry I noted thatone reason I was basically disinterested in Solaris Live Upgrade is thatit had hung up when I tested it several years ago (and quite a few patchlevels back). This may strike people as a rather harsh reaction to abug, to which I am going to say: absolutely, but it's the same reactionI have to bugs in any similar tool, regardless of who it's from or whatit runs on.

In order to make Solaris Live Upgrade worth

Solaris Live Upgrade is one of those things that I would like to love,but I can't seem to get interesting in, and here's why.

First, as I've written before ,we don't patch our Solaris machines . They're onlyaccessible internally and they're effectively black-box appliances;there's very little enthusiasm here for patching a working appliance.

When we do patch our Solaris machines, we never need to roll backpatches because we test them first in

There has been a little bit of commotion lately about Wayland perhaps replacing X ( see also ). I'm not really happy, andwhy goes like this:

• odds of a full Wayland environment having an X compatibilitylayer: decent but nowhere near sure.
• odds of Wayland having remote window support on initial deployment: low.
• odds of Wayland supporting remote X the way it's done today: basicallycertain if it has X support at all.
• odds of there being a Wayland to X translator, so you can