BT et C

Thursday, September 29, 2005

HE

Well none of you have seen anything on this topic for awhile -- you may even need to consult the "Most Important Abbreviations" in the bottom-right to find out what "HE" is. There's a story behind that. Over the course of June and July I dropped about $700 by playing medium-stakes limit hold'em, and playing it badly. The bonus referred to in my last poker post didn't happen (basically there's a time-limit, and I got punched, so I couldn't clear the bonus). And I withdrew a few hundred to pay bills.

I played a bit more limit HE in August and September, netting $104. Had ups and downs at Richard's game, didn't track that precisely but probably -$50 or so.

The new leaf I turned over yesterday is that I'm back in the no-limit game, partly because I'm helping a friend write a book on NL hold'em. And results were great: +$89 over about 1100 hands of $50 buyin no-limit at Absolute Poker. That's 32.3 BB/100 -- which rocks.

I gotta get off the game and do the dayjob for a couple days, so September's final balance is:
Richard's game: +$44
Limit HE: $92
NL: $139
Bonuses: $50

Total: $324

Will be back up on cloud nine pretty soon at this rate.

Wednesday, September 28, 2005

a WikFisk

There's this problem I have, and today I encountered the epitome. Seems like Luke has the same problem from time to time:

I go to my favorite resource, the wikipedia, and decide to get current on definition/discussion of SOA. Here's what I get (begin fisk):

"In computing, the term Service-Oriented Architecture (SOA) expresses a software architectural concept that defines ..."
Warning bells going off already. Why?
A term doesn't express a concept. People express concepts. Using terms. Terms are supposed to just flat-out mean something. What a term means is called its definition. So when you ask for the definition of X, you should be told what it means, e.g.:

Matt: what does SOA mean?
Wik: "SOA means ... " [stuff that it means]

Well, I won't edit the wikipedia there ... they might not understand; I'll do it here:

"In computing, SOA means ..." (and picking up where we left off) "...the use of services to support the requirements of software users."

See I left "in computing" there, in case someone happens to search "service-oriented architecture" without realizing that it's kind-of a computer term.

"Supporting the requirements of software users", if you read carefully, means ...

"being software"

Don't believe me? I stipulate that "supporting" means "fulfilling" (i.e. fulfilling the requirements). Rickety buildings and people with low self-esteem need "support". "Fulfilling the requirements of software users" means "doing stuff with software". QED.

To paraphrase, then, we've learned that service-oriented architecture means "orienting software around services" or maybe "using services to do stuff with software". Yeah. Then

"In a SOA environment, nodes on a network make resources available to other participants in the network as independent services that the participants access in a standardized way"

Questions for the audience:
1) What's an SOA environment? I comprehend metaphors, but you should use one at a time: architecture (principles/etc. by which a thing gets built) or environment (a place where actions occur). This sentence, then describes an SOE.

I swear I'm not just being obtuse. The mind recoils from this kind of metaphor-overload. And if you let this slide, it's not long before you get "SOA solutions" (instead of SO Solutions) then "Enterprise SOA solutions". Then you get people who "deliver Enterprise SOA solutions". You're being expected to fuse the metaphors of service, architecture, problem-solving, delivering, and of course everyone's favorite "the enterprise" (really digging kuro5hin lately)

2) What has this sentence added? You have to dig a bit, but there really is some new info here:
-"services" refers to resources on network nodes.
-there are standards that govern access to these resources.

...so far I've basically got: "In computing, SOA means using standard methods to access services/resources in a network"

Well, I did in fact get a bit of meat out of that, but it was too much work, so I looked elsewhere for further clarification, rather than reading the whole article.

Reynolds is much closer to the mark. Note that he yanks out the ubiquitous "supports" in "the architectural style that supports loosely coupled services" so that we get "architectural style that encourages the creation of loosely ..." This is correct; was that so hard? Note that the architecture metaphor is truly explanatory here: this architectural style encourages these construction methods.

Like so many things in IT, there really is a pony in that pile, but as one of Reynolds's commentors pointed out it often turns out to be an old pony:

"tech marketeers recently 'discovered' this long-known concept, then promptly smothered it in a bunch of buzzwords. That's what they do. Remind [programmers] it is no different than their recent 'discovery' of the simple decorator pattern, which they now shill as AOP. I expect you will get a big sigh of relief out of them.

"Skip the hype. Simply explain that SOA means nothing more than separating business functions into routines, just as they have always done."

Thursday, September 15, 2005

Cool

Try searching for "MASOTY"

Wednesday, September 14, 2005

-NC

Kur5hin makes an interesting argument against using the non-commercial (-NC) license for free content. I suspect that this is the license individual artists gravitate toward, out of a fear that allowing commercial uses would make it possible for another party to monetize their content and grab away all the revenue.

I'll admit that's my first impulse with regard to my content...esp. when I saw that's what most of these guys did.

The first point is very well taken; my personal use of wikipedia continues to skyrocket (I'd estimate I visit 15-20 articles per day, even if some people think I'm stupid for doing so) and if your content can't be weaved into their system it's going to be a serious benefit lost.

And as for ye olde argument that other guys will be making money that I should be making off your product, here are the relevant grafs:


The moment you choose any Creative Commons license, you choose to give away your work. Any market built around content which is available for free must either rely on goodwill or ignorance.

The potential to benefit financially from mere distribution is therefore quite small. Where it exists due to a predominance of old media, it is likely to disappear rapidly. The people who are likely to be hurt by an -NC license are not large corporations, but small publications like weblogs, advertising-funded radio stations, or local newspapers.


Eventually, he suggests the obvious step of using the Share-Alike license (bascially a GPL for creative commons content). The other alternative would, of course, be normal rights-restricted copyright.

Personally, I'm undecided. One thing that came up, and seemed smart, is a -NC license that reverts to Share-Alike after 5 years (though I'd be inclined to make it 2 or 3). Maybe I'll ask the Comfort Stand guys for input. Dang, that's a snazzy outfit. (Go to Legal Torrents to get their 10-disc sampler.

Monday, September 12, 2005

You know what we need?

Longer copyright terms

"The justification from an economic perspective is absolutely baseless."
Lawrence Lessig

And yet ... something tells me it'll go through.

Friday, September 09, 2005

BSE & BS

In today's post, Bovine Spongiform Encephalopathy, also called "mad cow disease", becomes a starting-point for an exploration into political reality-versus-perception issues.

I think if you asked your average down-to-earth, common-sense oriented conservative what's up with government regulation of business, consumer safety legislation, etc. you'd be told that regulation in modest amounts is defensible, but that it's just gone way too damn far in this country.

Your super-conservatives (the "radical right" (:) would, of course, deny that first part, and yearn for total deregulation of everything. But let's stick with the boring, middle-of-the-road mainstream position for awhile.

The basic idea is that companies (e.g. beef packers) will invest what they can in making their meat safe. They are in fairly constant conflict with bleeding-heart liberals who want them, increasingly, to cut into their profits and do more testing, etc. that the consumer has not explicitly asked for in the market.

And the small-government position is that there isn't any need for all these programs to crawl over the companies' facilities with a microscope, because the information about which companies have the safest meat will get around in the marketplace, and the companies that make the safest meat will reap the benefits, etc. This all makes perfect sense.

Now think for a minute about that picture of things ... how reasonable it is, and how -- with only minor adjustments -- most people would agree that yeah that's pretty-much what's up. The government's too big, and trying to over-regulate stuff is at least part of the problem.

Into this wheel I'm going to throw a spoke:

Two companies in America would like to test ALL of their cattle for BSE and provide this reassurance to their customers. But, alas, they cannot.

How come? Is it technically difficult? No -- the BSE test is pretty straightforward.

Can't afford it? On the contrary, one of them has already invested in the necessary equipment and is just rarin' to go. Largely because much of their (former) business went to Japan, which banned US-imported beef until all animals are tested.

You've got a company in a "free" market that wants to spend its own money to make extra-sure its products are safe, and a bunch of customers who are cool with footing the bill. So what's up?

The USDA forbids the company to do these tests.

Let's read it again: The USDA forbids a private company from testing for BSE at its own expense. The testing equipment is required to sit there, unused.

One of the companies, Creekstone, is losing about $280,000 a week because of unavailability of the Japanese market. "That the USDA is standing in our way makes no sense," Fielding (COO of Creekstone) said. "Their position flies in the face of the basic rule of business -- that the customer is always right, and our job is to meet their demands." (src)

Well, Mr. Fielding is right of course. This post is about what does and does not make sense, or more-to-the-point, whether common sense has anything to do with What Actually Is. Because this picture, of the free market asking for X, and a business deciding to provide X, and the government saying "no way", does not fit in neatly with the common-sense (I would posit: reductionist) view of the situation presented above.

Let us propose three possible ways of organizing a market for beef:
1) "Japan style" - require that 100% of beef sold come from cows tested for BSE.
2) "free market style" - let people know which packages of beef have been tested (i.e. with a label or something) and they can pay extra for these packages, or take their chances. "Americans are willing to fund a higher level of reassurance. A January poll by the Consumers Union showed that 95% of adults would pay 10 cents more a pound for tested beef. Testing every slaughtered cow would cost about six cents per pound." (src)
3) "US style" - make everyone take their chances, even if they don't want to and are willing to shell out cash so that they don't have to. "The U.S. Department of Agriculture (news - web sites) (USDA) currently does not allow such private testing for mad cow disease. And it claims that a new government testing system it approved this month is perfectly adequate. " (IBID)

This leads us to two questions:
1) Why do we have a US style market?
2) Why do we think we have a free market style market?

Well, if you looked through those linked pages enough to solve the mystery, you found yourself right back at common sense: it is because large corporations are not willing to compete fairly. (Bet you're going to have a heart attack and die from that surprise). The "big four" agri/beef companies have put a lot of pressure on the USDA to disallow any private testing. That is to say, it would cost them a Lot of Money to test all their cows, and they don't want to. But they can't very well have their beef sitting on the shelf next to a package of some other company's beef that says "tested for BSE". That would be a #2 style market, and we all know what another word for "number two" is.

Interlude: yes, this is more about beef than anyone wanted to know, but too bad. Japan apparently backed down from the 100% testing requirement, but as of March still hadn't reopened the market. I bring this in only to show that I'm aware of it; it is, however, completely beside the point, viz. my right as a Beefy Entrepreneur to test for Neisseria Meningitidis if I want to and/or if I think my customers want me to. Even though it only affects humans.

Well, I shoulda made the interlude into an Epilogue, 'cause I think that's just about it. I'm not of the opinion that all American industries are under-regulated, just that things are not as simple as they seem. And, emphatically, I believe this:

The influence of major corporations over our governmental institutions is a much larger problem vis-a-vis the free-ness of the market than the (I daresay weak) regulation of industries and standards by government. Much larger.

In short I think the whole business-versus-government thing is a myth, and that it was basically foisted upon an unthinking populace by business-and-government, which are in fact Very Good Buddies.

ESR

Best Blog Entry Ever

Thursday, September 08, 2005

Wirelessless

Okay, now I've done it about 4 times. And I ^$%^@ing hate it. This failure came after a big and exciting "yum update" which (theoretically) got me the latest versions of 300-something packages. Well, I haven't looked at the actualy yum script but I think it's like this:

--------------------------------
#include

install 150 packages
break wireless networking
install the other 150 packages

--------------------------------
okay, I'm kidding. Was pissed but am able to run wlan0 just by selecting the old kernel in startup. But shucks I want to use the new kernel someday. Ndiswrapper people, any chance for a "it just works" type system? even -- gasp! -- a GUI has been suggested.

Wednesday, September 07, 2005

Just to keep current...

Mostly this is because I haven't posted anything in awhile ... explanations for that later. Here's one of the simplest, clearest, and most cogent paragraphs I've read in weeks:

"This book centers in what we call a "propaganda model," an analytical framework that attempts to explain the performance of the U.S. media in terms of the basic institutional structures and relationships within which they operate. It is our view that, among their other functions, the media serve, and propagandize on behalf of, the powerful societal interests that control and finance them. The representatives of these interests have important agendas and principles that they want to advance, and they are well positioned to shape and constrain media policy. This is normally not accomplished by crude intervention, but by the selection of right-thinking personnel and by the editors' and working journalists' internalization of priorities and definitions of news-worthiness that conform to the institution's policy."
Manufacturing Consent, Herman, E. and Chomsky, N.