Nice. Bold/apocalyptic, clear white text on black back, a little heavy on text but not too much. I like it.

Is there going to be a catch-page relevant to Carney at BlueAmerica.Firedoglake.com, or just the donations page that's there now?

I think that any unsubstantiated smear against anyone posted here in UT should be taken down... Why allow it here?

Why indeed.

The integrated campaign makes sense, and it sounds good. I figured the print piece was the run-in, i.e. the piece where all questions would be answered. It's a little heavy on text by conventional print ad wisdom, but not too much.

If you were paying me as a consultant, I wouldn't advise you to cut it any, either; in the mid-90s there was a trend toward shorter and shorter ad copy, but later studies showed that ads with more copy (up to slightly less than this) held readers' attention longer (duh) and got similar responses, slightly better.

Catch-page: I know this drops tomorrow, but I think it would be good to have one.

Your call to action is plainly to phone Carney at the number in bold letters ... but those who want more information are going to go to the donations page and get confused (two messages = no messages) and neither call nor donate.

The easiest thing might just be an interstitial ad, or a want-to-know-more page with a repeat of call-to-action (call carney@) at the top and links to more articles about FISA/PAA/telecom immunity underneath(I guess we know where to get those).

Am I missing something?

Probably nothing important. In the last thread, there are about four pages of comments between bedtime and this morning (EST). Some folks refer to comments that they say are no longer there, and I saw some inline pastes from previous comments I hadn't seen, but it could be that I just missed 'em the first time. Could be other people did too.

Oops, GG responded, so nevermind my superflous comment.

Thanks for the heads-up about digby's post; I haven't set up my RSS reader so I missed basically anything she's written in the last week.

Main Core does sound a lot like the 'much worse' we've all been worried about, doesn't it? A couple things particularly raised my eyebrows, and not all of them relied on assessments about the current regime (meaning, there's no reason to breathe a sigh of relief come January):

1. "Continuity of Government" - This had me reaching for my tinfoil hat. There was a big push for Cheney back when he was SecDef to upgrade these COG contingency plans, and they instituted drills to practice the contingency plans. It does sound like a road map for martial law ... think now about Iran, massive demonstrations, the suspension of elections and so forth.

2.'Not really a database but a way of searching other databases' - Just who designs those algorithms, according to what kind of review, and how often are they revised?

Worse than this: let me explain via anecdote. A dozen or so years ago, I was principal for a consulting group that built a lot of data warehouses, decision support systems and data mining applications for corporate customers. All the data from generations of heterogenous applications had to be extracted, transformed and loaded from existing data stores into a new data warehouse to do any data mining. Often, the 'transformation' also meant lots of cleaning, which was an ugly process ... but at least it didn't have to do with sensitive data we're discussing here.

The problem you eventually encountered was not technological (though there were those), but ontological: the design of data base tables, even for simple things named 'order', or 'customer' say, ran into divergent, sometimes incompatible assumptions from their source data. 'Fixing' this involved trade-offs; who in the national security apparatus makes those decisions, at what level, and are they bound to be any better than the ones that produced all the false positives the FBI was complaining about?

3. "Social Network Analysis" - I learned how to do this in O&I school, many many moons ago. Very useful tool, but finely focused and not meant to scale, and I can't imagine applying 'predictive' AI techniques to it without an awful lot of error. Ironically, the problem might actually be not enough data to prevent significant errors ... especially in light of (2) above.

There are probably a lot of people out there thinking, I have nothing to hide, or, the govt already has a file on me but I know I've done nothing wrong ... but these are the nightmares they should be having.