From the Wikipedia entry on "Normative":

Standards documents

In standards terminology, "normative" means "considered to be a prescriptive part of the standard". It characterises that part of the standard which describes what ought (see philosophy above) to be done within the application of that standard. It is implicit that application of that standard will result in a valuable outcome (ibid.). For example, many standards have an introduction, preface, or summary that is considered non-normative, as well as a main body that is considered normative. "Compliance" is defined as "complies with the normative sections of the standard"; an object that complies with the normative sections but not the non-normative sections of a standard is still considered to be in compliance.

The English version of the XML Schema Primer (2nd Edition) is the only normative version!

"The media" gets accused of political or cultural bias on a regular basis, especially during an election season. I often hear this phrased as "the liberal media bias," as a general statement, but also especially applied to the New York Times, and I've also heard "the conservative media bias" especially when the Media under discussion is Fox News.

What if we wanted to make a factual examination of the claim of media bias for a single media organization or a group of them? Ought we break apart a single news organization into its functional units in case there are different biases? Do we count editorial/opinion pieces or exclude them? And who among all of us has the authority to meaningfully verify or contradict this conventional wisdom for both liberals and conservatives and people who don't like pigeonholing themselves within those terms -- that is to say, whose assessment would we all accept?

This issue seems like a good example of a cultural catch-22. Lets assume the best of each other: we all want honest politicians who exercise good judgment. But in order to assess the honesty and good judgment of a politician, we rely on facts and the luminaries who we trust. Unfortunately, we don't even trust the same sources for *facts* and we certainly don't trust the same editorial/opinion column writers. One oft-proposed way to get past this traffic jam is the idea of independent organizations doing fact-checking and etc. But any such organization seems to be immediately accused of bias when it consistently flags one side as misrepresenting the truth; I'm sure some of the organizations really are showing their own underlying bias, but others may simply be revealing the truth that one candidate or side of a political argument is factually incorrect either through misunderstanding or lying.

Who benefits from this set-up? Do we? How can we get past it?

what would life be without too much excitement?

I've had a couple ideas lately about better ways the real world could work if we find ways to add some functionality. A completely virtual world could work this way, but I'm interested in thinking about how the physical world can be made to work differently or hybridized with interfaces typically associated with virtual reality, like HUDs.

Two ideas:
1. Calorie counter - food items in your visual field display their individual calorie count. Information is exact for the item. Like a hit-point counter. Somewhere in your field of vision you see a cumulative total of your calories for today, either counting up or down depending on preference. During cooking, you can set a "this trip" sort of counter for items you add to a dish. This could be modified to include fat, trans fat, etc. You would be able to customize it and turn it on and off.

Use case: I am munching on some almonds before lunch. I have no idea how many almonds I've had or how many calories total. I am not on a diet, but I try to stick to around 1500 calories a day. Sure would be handy for planning the rest of my day to know how many calories I have left at the end of the almonds.

Implementation hurdles: user friendly heads-up display; management of visual clutter in heads-up display; collecting caloric information from objects in physical world (differentiating food/non-food items, calculating calories for individual items).

2. Frequency/status sort for clothing - wouldn't it be great if you could keep all your clothes in two giant heaps, clean and dirty, but always find the exact pieces you were looking for or browse your collection to find exactly the type of gear for a particular occasion? I'd like to be able to sort and instantly resort my clothes in the following ways and combinations of ways: frequency that I wear garment (something like popularity), status (clean/dirty), location (storage/closet/parents'/floor/car), color, pattern, material, outfit (groups of clothes I've pretagged as belonging to a named set, an outfit -- same garment can participate in more than one set), garment (pants/skirt/shorts/blouse), coverage (lowerbody, upperbody, whole body, neck, etc.), sleeve length, skirt length, cut (v-neck, boat-neck, t-neck, mandarin collar, relaxed fit, bootcut), manufacturer, occasion (formal/informal, etc). I don't want a report about what I own, I want the selected items to levitate in front of me on command.

Use case: I'm going out to a trendy bar. A rare thing for me. I know I've got some trendy bar clothes somewhere. I can picture them, but I have no idea where they are. It's a complicated outfit involving a special bra, 3 shirts, a skirt, tights, a hat, a scarf, and a belt. I need to find them all, right now. I ask for this outfit and it flies to me, or I get alerts that some pieces are dirty. If something is dirty I can look for substitutes along the lines of: find clean v-neck white shirts.

Implementation hurdles: associating appropriate metadata with clothing either on-the-fly or in a more permanent way; tracking location and status; making clothing fly, but only when it is in a set of search results, forcing the flying clothes to sort so that most relevant items come to the front.

open when relevant to your life

For several months now, I have been documenting and studying work flow and business processes. A few days ago, I put together a draft problem statement for the executive who commissioned this work in which I defined a problem as an area where a lot of staff time is being spent to do something that could be done more efficiently or not done at all. I listed the problems: "We spent X amount of time doing Y per year."

The executive came back with a challenge [in my words]: rephrase these problem statements in terms of ultimate causes rather than proximate causes.

Some former Google employees started a new search engine company called cuil (http://www.cuil.com). Based on a review in NYT, I decided to check it out. The review said it has a novel way of clustering results  that's very slick.

I decided to check it out. I did a classic search engine test search and one I made up on the fly, comparing the Google results and the cuil results for "apple" and then "ipod shuffle usb".

Apple is a fun search term to test because the results need to be heavily clustered for disambiguation. Google is not that good at displaying these types of results. It turns out cuil isn't either -- cuil's results for apple look like a web portal. They do some clustering, but the clusters take back-stage to what looks like a screen-full of mini-articles, mostly about Apple computers. Google's display for "apple" isn't great, but at least it is still obvious that you're looking at a search engine and not www.apple.com. Both companies could use to do more work on the presentation. I didn't feel like wading through a bazillion stories about apple computers and I didn't really have a target page in mind, so I stopped there.

"ipod shuffle usb" produced a predictable list of commercial offerings on google, mostly relating to buying an ipod shuffle usb cable. The Same search term produced no results on cuil. That's disturbing since there are clearly thousands of pages that have the phrase "ipod shuffle usb" cable as an exact phrase and surely many more than contain those three words in or out of proximity. I was puzzled by this so I tried just "usb" and also got no results.

cuil may yet be very interesting, but I'd say it's far from ready for prime time.

Our Senate just voted to give retroactive immunity to corporations who broke the law at the behest of the US gov. by illegally spying on their customers though the government had no obtained warrants. The legislative branch has now blessed these companies with immunity from customer lawsuits. I saw: they can make them immune, but they can't stop us from boycotting them. Except, it's not all that easy to live a normal modern life and boycott these companies; even if you refuse to become or remain a customer, your data passes through their networks. Exciting, huh?

I'm not very pleased right now. I'm especially unpleased with Senator Obama. He used to be opposed to this bill, but he voted in favor. I have to decide if this means he's lost my vote in November. I'm certainly not going to donate any additional funds. The only thing more ridiculous than Congress passing to permit warrantless domestic surveillance is Congress passing a law to forgive companies who did it when it was illegal. I did have some sympathy with a proposal in Congress to give the telecos immunity by transferring liability to the government itself; that, at least, didn't seem like passing the buck or denying the wrong-doing.

And in case you're wondering -- I don't think that terrorism is LESS dangerous than an unfettered police state. I think an unfettered police state IS terrorism.

I can't do polls since I don't pay LJ, but...

Lucky Kitchen:
good, bad, mediocre?

Break my household tie!