Matthew's Weblog

Another gem via Bruce Schneier… a mathematical analysis of the NSA data-mining of our (yes yours and mine) phone calls and internet access. The AT&T stuff that got folks in an uproar.

Surely the NSA folks are smarter than this…

Here’s a good article (through Schneier) that feels a lot like my recent search for a lost document.  I spent hours trying to find the soft copy and never did find it, though I could find a version I had printed out within 10 minutes.  With that experience, I would not doubt that it would take months to try and build a terrorism case… and sometimes be completely fruitless.

Technorati Tags: terrorism, security, data+overload

An interesting update from the halls of Choicepoint – after regulators slapped them with a big fine, it sounds like they’re moving in the right direction.  Now if only it wouldn’t take a catastrophic embarassment to make companies do the right thing… or if consumers could get excluded from the database upon request.

An insightful piece into the Agency Who Cried Wolf.

via Schneier

Technorati Tags: security, nsa, wiretapping

Yes, it would be hard to configure and extremely personal, getting things wrong for everyone else.  However, if the configuration time paid back, I’d set up an intelligent filter in a moment.  When do I get a chance?

Technorati Tags: personal+technology

I’m impressed with the technology and ingenuity of recombinant ideas in new photo-sharing sites that TR discussed, but once again I don’t want to have my stuff controlled by someone else.  I’m very interested in using facial recognition for good instead of evil, such as the auto-indexing of pictures that Riya is doing.  Including a blog by the CEO where he actually apologizes for the company making mistakes… acknowledgement of the reality of life, what a concept!

I want to be able to ensure my pictures are archived, backed up, indexed, never altered, and able to be hidden.  I’m not sure a hosted service will give me such assurances, even though I’m willing to pay for that.  I’m currently paying in my time and hardware, and I’m not ignorant enough to count that as free.  Maybe if there was a way to back up the entire image database associated meta information offline…

Now, let’s hope Gallery puts it in version 3?

I don’t trust anyone to have all of my personal details – and I know it is somewhat naive of me to think I can hide them – but the issues I keep having with ideas of Universal Authentication are that I want to control my own data.  I don’t want any specific repository, I don’t want the software deciding what information is appropriate to share, I don’t want to have conflicts with other users when we have the same information… I want to be the one deciding what to distribute.

I even – gasp – lie to some sites.  The LA Times (free login required) thinks I live in Kansas.  How will the Shibboleth system handle my lies?

SecurityFocus examined a court’s reaction to one of the laptop-with-personal-data losses and suggested we should have strict liability for data breaches and my question is why don’t we already?  The court’s argument that no one protects data so no one has to is silly in my mind. An average “reasonable man” may not encrypt data, but you don’t hire them – or shouldn’t – to run data security.  You hire experts.  Every expert I have talked to (admittedly a small sample of only dozens) says: “Encrypt.”

So let’s use a “reasonable expert” standard instead.  Any lawmakers feel game to write that one up?

So I’ve found a clear description of the reason I hate the idea of hosted email… I don’t control it. TR did an article on why it’s scary to use Gmail for your email.

Just think about it before you put your information out there…

So TR wrote about the danger of Google’s ‘Search Across Computers’ function since it uploads files to a server you don’t control. Makes you wonder whose company data is finding its way up there, unknown to the appropriate internal watchdogs.

The EFF suggests Google take two steps – find a different way to implement the searches, and reform the law. Both are good ideas – it’s too bad they undermine them by suggesting an inappropriate solution (peer-to-peer) rather than letting Google figure it out.