The issue here is his particular job in intelligence. An extramarital affair, heavy drug use, or anything of the like is a job liability (not just a political liability) in public policy because it opens an opportunity for blackmail. That's the first problem. The second problem is that even if nobody finds out, you still have no idea what he's telling his mistress, or when they'll break up and she'll start talking. We can presume that whatever level of commitment she has in the relationship, it's probably not as high an investment as, say, his wife has in their marriage. Eventually, it will end.

Furthermore, since this whole thing is also supposed to remain a secret, that also minimizes the amount of overt protection he can afford his mistress. (This would be more of an issue, say, during the height of Cold War, when kidnapping an intelligence chief's mistress for interrogation might one day be a tempting enough target for an enemy agency. Still, it's a possibility.) There are a whole slew of operational issues built into the secrecy of this that make mistresses a bad idea for anyone in intel, with the reasons becoming more important the higher up the chain of command you go.

So now he's come clean. Doesn't that short-circuit the danger of a secret mistress? Sort of, but now you have the inherent personnel problem: it's hard to tell your operational agents about the dangers of secret affairs when you're doing it yourself.

Then you have the underlying issue of character: if he can't remain loyal to a marriage, why should we assume he can remain loyal to his country. I know that sounds like a leap. It is. But it's still the sort of question that needs to be asked. Secret societies -- even extremely popular ones, like the Masons -- have small secrets like handshakes, passwords, and rituals for a reason: if you can't trust a man with a trivial secret like a handshake, you sure as hell can't trust him with a big, juicy secret. Discipline has to be developed, and lack of discipline anywhere is a bad sign in the long run. Hell, military intelligence frowns on anyone who has more than two drinks per meal as being risky.

Personally, I'd be fore ditching time zones, as it makes inter-zone business a lot easier. Never again would you have to pipe up to remind your boss' boss' boss (and all of his peers on the line) that his proposed, weekly 8 AM teleconference (New York time) means you have to be at the office at 5 AM (California time). Instead, HR can just gloss over it with company-wide policies like "inter-deptartment meetings will be scheduled between common time 16 (PST 8 AM) and 22 (EST 5 PM)."

It would also make international business easier, since it's immediately clear from email time stamps, hours of business on web sites, etc. when a supplier, customer, or branch is open. Unfortunately, these sort of things are all rare enough -- either rare events in an average person's life, or common in the lives of relatively rare people -- that most people don't see a huge benefit.

I'd be in favor of keeping some version of the 12 hour AM/PM designation, just because it could help disguise dateline issues that arise with the new system. (That is, the fact that a group of people all have to go into work after sun-up "late" Tuesday and comes home before sundown "early" Wednesday.)

As for DST itself, there's really no point to it. If you work according to the sun, you go to work whenever it's up, regardless of time. If you don't, you don't really care when the sun is up, because virtually no office, shop, or business depends on natural light anymore. Ever since artificial lighting became the norm, the scheme only "benefits" people of the latter group who support the former group, such as banks. And we all know banks strive to maximize their availability to customers, right? Right?

The savings from coordination are completely lost when you factor in the cost of adjusting everyone to a new time. Hotels, airports, and such are especially affected. Then there's the fact that in some regions you actually want people working when it's darker, because air conditioning costs much more to run than lights. Abolish the system and never worry about it again.

Maybe next time you'll vote early and get one of the plentiful "I Voted Early" stickers (with 200% more smugness).

TX - Austin

Careful, they might just jump ahead to directly buying Joss Whedon.

So I wasn't the only one to have that reaction.

First, note that "we" (am I to understand this to mean Americans? Texans?) did not sign such a document. We signed a document in good faith that we as a nation abide by. We are a federal republic, and the powers of the federal government that made that agreement are limited, and the document reflects that in the boilerplate. Believe it or not, in this country, officials can't use a treaty to bypass the defining legal framework of our nation.

They are not barred. But they do have to keep 100' feet away. (For those not familiar with feet, that's about 30 meters.) That's the law. The law that the treaty respects. Because it would be illegal for the treaty not to. I'm sorry you live a country either small enough or tightly-governed enough to not have a concept of federalism. But here in the U.S., both state and federal laws are applicable.

I don't see any articles about America. I see an article about Texas. And where Texas is concerned, we tend to say, "thank you for coming; don't let the door hit you on the ass on the way out."

First, a question: why should foreign nationals working on behalf of an international organization have more access to proceedings than United States citizens - whether civilian, or state or federal authorities -- are currently allowed by law?

Second, there is no such thing as "technically legal". There is legal and there is illegal. It is illegal for the observers to be within 100'. It is illegal for pretty much anyone to be within 100', except for voters and designated, trained administrators.

Third, let's look at the sections quoted by dinifinity above:

And they are. The extent permitted by Texas State law is "so long as you keep 100' away, just like everyone else doing exit polls, campaigning for specific candidates/propositions/constitutional ammendments, and and anyone else who might influence the election by mere proximity.

The United states is endeavoring to the extent they are able. As a matter of Constitutional law, there' not much more they can do. Federal election laws do not provide for strong federal oversight of state elections. Nor should they as a matter of federalism, since one would expect the federal government to have more power to coerce voters and influence state elections than any one state has of coercing voters to influence national elections. The issue the OSCE complains of in the linked document amounts to saying, "the United States is not organized like other countries, and that's a nuisance for us from a regulatory perspective." It would be simpler to enforce uniform requirements if the U.S. were like, say, England or France, with a strong central government and provincial governments in all cases subservient to that central government. Then they could apply nation-wide sanctions to effect a national change. But it's not, and they can't.

The OSCE could always try to sue in federal court if they feel the law is in error. So far, they have not done so. So far, this comment from Abbot is little different from the AG pre-emptivly reminding any group to obey the law, and there will be no special treatment. No different that a protestor rally.

Of course, we ARE following our own laws. The issue in this case is that an international group is complaining about our laws. (What? Our neighbors have an issue with how we do things in our house? Well, let me just put on my "give a fuck" boots and walk over to talk it out....)

And the last election? Barack Obama. He won pretty handily....
Wait, you're not still talking about the Bush/Gore election, are you? That was in 2000. Twelve years ago.

The main thing you have to remember about aggressive traders is that they're actually both smarter and dumber than you'd expect. That is, they're smart enough to recognize that most of their money is not made by spotting winners or losers early enough to get on the winning team. No, most of their money is stolen in fits by outracing other investors when things suddenly change. If we're lucky, they usually have a counterpart somewhere who is responsible for shepherding a reserve of cash, slowly built up by investing in solid companies as they build, so that the life and death of the aggressive portfolio is not also the life and death of the company.

The aggressive traders know a lot of their job comes down to timing, that the value they gain and trade is temporary, and that eventually the whole thing will melt down around them. Eventually, they will be the slow guy getting beat by faster guys. The large scale and small scale objectives are similar: get in on the rising edge, get yours, and get out before the whole thing goes to hell. Collapse is not an "if", it's a "when". The first thing they look for is always "when do I pull out?"

Everyone expects reports after the bell. That way, there's time to actually read and reflect, and everyone starts on a similar footing when trading resumes in the morning. Just as importantly, everyone knows and expects that they'll start on a similar footing in the morning.

If it were released during the trading day, there'd be pressure to analyse the document (and I use the phrase loosely here) as quickly as possible, so you can sell while it's still high or buy when it's still low, before most people have had a chance to process the new information. Most of the time, this means jumping on a single factor and reacting strongly.

Of course, then other people wouldn't actually need read the document. They would just see the line trending, say, up and then figure that someone who can analyse better and more quickly than they has seen a value increase and is now buying. So they would buy. And why not? As long as they're on the rising edge, and can recognize a peak/plateau, they can sell at the peak and still make money. So this compressed window leads to panicked decisions based on incomplete information which is multiplied across the market. Very disruptive.

Now, imagine if the report were not only released during trading, but _unexpectedly_ so. Not only would you have information, you would have information that the majority of actors don't have. You would have an advantage over them, one that will evaporate in a matter of minutes or hours. Once the trading halted for the day, the advantage would be lost. So they would move even more quickly and panicked than if they had been expecting the report during trading (which, of course, no one was).

The phenomenon you describe -- trying to profit off of the correction when the initial trend is proven to be based on incorrect assumptions -- would then drag the trading artificially in the opposite direction. It's like kicking and oscillator. And, of course, there's no reason that a smaller group of investors couldn't capitalize on the over-correction, and another group on the re-correction, and so on. Maybe the price "rings" for a long, long time before it settles to a more representative value. Maybe it gets so low or high that non-linear effects ("buy at ..."/"sell at ..." directives) come into play and either dampen or excite the oscillation further. Maybe the stock just bottoms out -- that is to say, the investors buying or selling lose enough money at once that they can't make call, even though the stock they hold may have value.

It's hard to say. But considering that it's all an artifact of traders trying to capitalize on the stupidity of other traders, and not at all a matter of the real price of the stock, it sounds like the kind of thing you want to discourage as much as possible.

On a related note: based on the chaos caused by automated trading routines of late, I think we can expect more limits and delays on trading to be mandated in the future.

An alternative interpretation:
Jimmy's girlfriend Sally takes his badge to class and drops off his homework. Jimmy is truant, but the school is getting state money, so at an administrative level, who cares? The teacher notices he's absent (though apparently can't notice Sally walking past the scanner twice a day. There's homework for him, but it's pretty poor since he never attends and his class discussion grade is, predictably, zero. While Jimmy is truant, he takes up a life of crime. After he gets caught on camera, he points to the school's RFID system as an alibi. Not surprisingly, the cops don't believe it, the administrators say they have no real faith in it, the teacher confirms he was out and never comes, and they can probably figure out that Sally was scanning the badge for him. Or, the cops just use common sense, say "we don't believe you" and arrest him based on real evidence a hell of a lot stronger than an ID chip location.

Jimmy is now in jail. Sally doesn't scan his badge anymore. The school no longer receives $30 a day for Jimmy.... but it did receive about $3000 more than they would have off of a truant kid like Jimmy than if they hadn't had the RFID system.

So really what is the down side to this system for anyone involved as compared to what happens now?

That was my first thought as well. With the increase in On-Demand content, they're already running pretty low. (Ever come home and find a "this channel is temporarily unavailable" message?) It's also a good reason to start phasing out non-HD boxes (provided your provider doesn't still charge more for an HD box). Eventually getting rid of the low-res channels and carrying only the HD channels would free up a little more bandwidth.

...email servers and clients pretty much handle the technical side already. All you need is a new "social" interface.

This about it. A social network needs first and foremost a list of contacts, their unique identifiers, and lists that partition your big "everyone I know list" into smaller lists like "friends" or "coworkers" or "SPAM/blocked/ex-friends/people I know but just hate". The address book is also the most basic, not-strictly-necessary feature of any email client.

You would like to be able to push data (updates, tweets) to everyone who matters instantaneously, or in a very quick, timely manner. This is the main point of email. A social network website just stores your mailing lists and fills in the "to" field for you.

In a distributed version of such a network, there are additional complications and benefits. You have to have background processes to poll other servers (nodes) to fetch data, to make sure that all archives stay in agreement and don't lose data, that there are fail-over and reconciliation mechanisms for when communication is not possible (there may or may not be new data that I'm missing). This isn't trivial to implement, but it's also not foreign ground. It's not too different from what a news group client does, with a little torrent-like dynamic peer management. Newsgroup readers are generally built and bundled with the software that had the most interface and back-end similarities to it... the email a client. You would have a lot of data to collect from new friends, but the fact that you actually know each poster means that the more of the data pulled will be relevant to you than it was back in the newsgroup days.

You would like to be reminded or actively informed of certain information (birthdays, events). Calendars are built into every modern mail system, as is the ability to invite/require people at meetings and events.

You would like to play games and compare scores with people you specifically know. All of Facebook's games are flash-based (run on the local machine anyway) with some state information (scoreboard) tied to a third party server. Other than the fact this is a browser job more than a mail client job, this is already mundane, and nothing would change on a new system except for better visibility into the API, and control over what servers you connect to and what data you release. You could store a small, cookie-like fie for each game which friends could compare to their own to dynamically generate a "my friends only" scoreboard for them to compare to, if you for some reason don't want to expose your friend list to a particular game. In other words, games are "least facebook-y" aspect of facebook.

You would like to be able to set up "public" pages not tied to any person (groups, events). To continue the email metaphor, this is just a mass email chain with a specific subject line. The network makes sure that reminders are enforced, people don't "fall off" the chain (the only valid reply to a group-style message is "reply all"), and you have a body of data (history) that you want to be available to people who join later. The last bit produces some overhead, as the group is essentially a "pseudo-friend", whose friend list is identical to the member list. In a distributed system, multiple nodes will have to have to responsibility of maintaining this data, so that it's not lost if some large number of nodes decide to drop it simultaneously -- for example, if every such node is actually a user running his own server, and all of them leave the group simultaneously. This is not trivial, but is also not impossible. It will take some basic management (no more members = no more group) and perhaps some interface changes ("This event is two years old. Can we delete this stuff yet? )" or "do you want to archive this event to your local machine permanently?") but it can be done.

Furthermore, everyone today has an email client. Each of those is tied to a server that receives and stores data even when the client is not connected. So long as each message is below a certain size, has a header, and the total mail is below another certain size, most of these mailboxes don't care _what_ the data being mailed says or means, making them a good fallback for data intended for your friend if his node is temporarily knocked offline. Plus, it gives you a way to stay in touch with people who _aren't_ part of the network, leaving the door open to bring them in later. If you get the social network client integrated into the mail clients, eventually setting up your identity on the network becomes no more painful for newbies than setting up a new email address.

There are other technical challenges to a distributed network, such as managing public/private keys to validate data on a public network, the need for a system of discardable "addresses" (so you can de-friended people), and keeping encrypted mappings between "addresses" and real key for user data on what is essentially a public system. These will take people familiar with at security to design and shake-out.

But the sheer possibility of making it work? That's a pretty small question. We know it can work, because we've made pretty much every piece work before, on a large scale, 15 to 20 years ago. We know people will come, because people actually do want off facebook. None of the "problems" with a new system would be unique to it, or worse than problems we already face on similar media.

I'm going to reply to this because even though it's satire, it's the closest major post to what I actually came to say.

Rather than blaming the dehumanizing effects of the interface, the disinhibiting effect of anonymity, or the fact that there's still no good way to punch someone over the internet, has anyone considered that the web just throws you together with people you hate a lot more often than real life does?

In waking life, people meet at places where they have to be (school or work), or through networks of people they already associate with (family and friends), and only somewhat rarely in situations where they meet people about whom they know nothing and to whom they have no social ties (bars). Even if you go someplace with a lot of strangers who all came together for some external cause (say, a concert) you never insert yourself into strangers' conversations, or listen in on them.

Online, you interact almost exclusively with people you know nothing about and only "met" online. You may share a common interest -- technology, a certain game, a fondness for goofy cat pictures -- that brought you into the same "space", but that's about it. And the medium is such that you HAVE to listen in on these people's conversation. You have to participate in their conversation to participate in your own. There's only one conversation. That's why you came to the page/board/forum.

In real life, if you realized you were talking to a person you hate, who is so annoying and antithetical to your own opinions that you just have no reason to keep listening to him, you would excuse yourself (or tell him off), then walk away. And then you're done. You never need interact with that person again, at least not beyond that across-the-room glare that says "oh, you... you're here. You stay over there, I'll stay over here." Online, you can ignore him or tell him off, but he doesn't leave the conversation. He's still there. Because it's still his conversation too. If you want to talk about news for nerds, Warhammer, or lolcats on a given board, he will continue to be there, and you will continue to piss each other off. Even if the board design lets you "ignore" him (that is, preemptively hide his posts), it may riddle your conversation with holes.

Really, your only choices are to leave, to be a little miserable every time he shows up in "your" conversation, or to tell him off so badly that he leaves once and for all. And this doesn't just go for one person. It goes for several people, on any sufficiently large site you may visit. And I think that contributes a lot to the combative nature of online discussions.

To say nothing of the fact that the version as a number doesn't matter. The purpose of the version is to distinguish between different version of the same product so you know what's compatible, what broke, where to start debugging, etc. Most major OS releases don't even come close to being "the same product" from a user perspective, and the other factors are all issues that developers care about and end users pretty much shouldn't have to.

For things like Windows and OSX, all the differentiation that matters to developers comes from long strings representing the most recent build/service pack. For customers buying software packages -- who can expect the software to work reasonably no matter when or if they got the latest upgrade to a component they've never heard of -- you only need one distinguishing name. Why _would_ you choose to use a number? Why not just a year? After all, you didn't drive to work today in your Ford Four Door Sedan v56.2.3.

Of all the axes to grind, I can't for the life of me figure out why the submitter would care about _this_.