I was trying to preview my markup for this post, the preview function refused to work with a blank subject line, insulting me with "cat got your tongue?" so I'm posting under my working title.
Math software shows up here fairly regularly. I keep taking notes, but never get around to using any. The R statistical package also gets frequent good mention, which I understand is accessible from within Sage.
And what's more, error handling needs to be part of the specification, because bad error handling creates road-blocks towards progressive adoption. How many good intentions toward standards have been thwarted by inscrutable error messages received on the bumpy road to enlightenment? As I recall it, the C standard explicitly makes no mention of error messages. This was not such a big deal as most syntactic errors were easily resolved. Then the C++ standard continued this tradition of not specifying error diagnostics while adding the template mechanism to the language. In my opinion this was a grossly negligent self-inflicted wound on the C++ community, and provided unnecessarily ammunition to the denigrationists who view C++ as a pox on humanity.
So yes, when a standard involves parsing down through multiple layers of syntax, it does matter to specify sane error responses. Unless you want your language/protocol to become known as the fat kid with pimples and a contagious itch.
Actually, one of the main challenges with natural language is that we humans perform so badly to begin with. Half the time we neither say what we mean, nor mean what we say. But it hardly matters: far more than half the time, the person (or people) listening hear either what they expected to hear, or what they wanted to hear, or they already knew they would disagree with whatever you were about to say before you even opened your mouth.
Sometimes it does matter. However, by the time you design a linguistic study to isolate the human gift for parsing grammar, the experimental task is about as "natural" as writing a law exam.
I think the contribution of grammar to early human language is way overstated. You don't need much grammar to handle everyday events, such as determining how to dress for dinner when the report from the field comes back "mammoth tusk hunter" or "hunter spear mammoth": in the former case (x3!) you'll be polishing your nose bone.
Where word order begins to matter is parsing the daily scuttlebutt. Did Adam tell Carol about Bob and Eve, or was it Eve telling Adam what she overheard between Bob and Carol? It's not easy keeping the cheaters distinct from the cheated upon. Plus Adam has to remember when to look surprised when Eve tells him something he learned from Carol just the other day. Not keeping your past/present/future and your cheatee/cheaters straight was a certain recipe for not sleeping on the warm side of the fire pit.
Later on, the grammar we acquired to parse who's zooming who became useful for digesting the BBQ assembly manual, but of course, that remains an evolutionary work in progress.
Maybe when children of the current MySpace generation reach the age to pop the big question ("What's an iPod?") and we've given up the fight to prevent our every indiscretion and peccadillo from being publicly archived for all posterity, we'll actually need a natural language interface to really drill down into the zettaflood of who said what to whom and who first posted it online and whether revenge was sweet.
Gaa! My mouse hit a crusty, and I blimped submit instead of preview. Let's try that one again.
The network cloud won't engulf 90% of computing, maybe 30-40%.
For some definition of 90%. I think it will break down according to the 80-20 rule: the 20% of applications that produce 80% of the value will remain on the desktop, while the 80% of applications that produce 20% of the value will migrate to the network. You outsource a large slice of your IT hassle, but only lose 20% of your activity for the duration of a network outage, which are fairly rare events if you have a good ISP. If Google apps takes off, it will probably drive demand for a more reliable last mile, and even small companies will not have much reluctance to pay for this if their IT costs are significantly reduced.
It really doesn't make a lot of sense to have your in-house IT people supporting the 80% that produces 20% of your value.
No doubt the NSA is quite pleased to see people yammering on about encryption as if that mattered. On a practical level, the most important data stream is traffic analysis: who is talking to whom. Traffic analysis forms a graph. What's a graph worth? Tough question. Perhaps ask Google.
On this graph, what does the NSA know about the various nodes and links? They know who uses crypto, and when. They know who you talk to and in what order, the duration, and the interval between communications. They also know if you are talking to a Tor node, or any other node suspected of laundering one or more end-points.
Think you have a secret Tor node. Well, these nodes are fairly easy to detect by... traffic analysis... unless the node is almost completely private. A private steganographic Tor node is your best cover, but it will severely compromise your bandwidth and latency. We've all got one of those in our back pockets. Neither would I wish to be apprehended in possession of that particular coke spoon: it has Gitmo express written all over it.
For any node or link where the probers that be have the slightest suspicion, they can also determine in the majority of all cases what protocols you are running. Even if you tunnel your anonymity through SSL, packet timing profiles is likely to leak significant information about the protocols employed. Even if you leak no timing information, you distinguish your SSL segment from every other protocol which does.
With commercial software, it is almost impossible to know if the NSA hasn't found some clever way to leak key bits through the "random" number generator. What to do? Obtain hard core anon steg crypto from an open source project? Don't be seen doing it. That will flag your packets upward for years to come. Maybe through your functioning personal anon steg Tor server? That poses two problems: first of all, you don't have one yet. And second, even if your recently purchased four-digit Slashdot ID from the Dread Pirate Roberts included a secret anon steg Tor treasure map, your anon steg Tor server is severely bandwidth constrained (supposing you wish to continue flying under the radar much longer).
Even supposing, how well do you actually choose your password? It's a virtual certainty the NSA maintains a list of the billion most popular passwords in every language of the world, plus the one million most popular mnemonic devices (including all forms of keyboard mambo), and all the most popular transformations of the former against the later. Think your mnemonic device is unique? Guess again. Not unless it is almost as hard to remember as the password it replaces. Have you ever used any portion of that password in a context less strongly encrypted (such as any Microsoft or Apple program ever written?) The elephant never forgets. Remembering is cheap, cracking costs money.
Should you actually prevail and manage to conduct your electronic communications immune to any form of analysis yet mentioned, congratulations, you have now achieved exclusive membership among the hardened targets of the world, where brute force and white vans finally make good economic sense.
Those little chocolate mints that showed up on your pillowcase the other day are the sign of a job well done.
The mass of a vehicle makes very little difference to highway mileage in non-mountainous terrain. It does matter for start/stop traffic, such as quick trips back to the 200,000 sq ft hardware store in the not-so-local mall because you bought the wrong bolt.
The reason heavy vehicles have much poorer MPG ratings is that the engine is usually sized to achieve inside-lane passing speed while pulling a 10% grade with several passengers and as much again in cargo. Excess displacement kills at low load. If you can't drop the hammer and pass two Winnebagos in 500m while cresting into Boulder the studly man-mobile is worthless to 90% of blue collar America. That's a lot of displacement for fetching the bolt you should have bought in the first place.
First, if this new scheme works, it will likely lead to Google ranking their own knols up and Wikipedia pages down, which could severely reduce Wikipedia's reach. I think there is a conflict of interest brewing here, unless the Google knols are displayed among the ad results.
As for article quality on Wikipedia, I think Fermi estimation applies: you get 80% quality by throwing many bits of text into the air to see what sticks, but rarely much above that. If you think about it, the original Nupedia approach generated quality at an appalling pace. It's never been obvious to me why quality suddenly becomes easier after you have erected 2 million shacks in various stages of dilapidation. Most the shacks have something useful inside, few are pretty to look at.
It not surprising, either, that many of the people who thrived at Wikipedia in the barn raising stage are leaving the project as it meanders sideways when confronted with a much harder problem. One could argue that Wikipedia is the most successful knock-up ever.
I've never ceased to be amused by the number of people who stumble upon Wikipedia in relation to pop culture, then learn about the vastness and rapid growth and conclude "well, if everyone is doing this, it must be good" only to examine the actual articles and find obvious flaws abounding. Anyone who uses MySpace as their basis step (if everyone is doing it, it must be good) was due for a rude eye opening to begin with.
When Bill Cameron, a wonderful CBC news personality, was dying of cancer he wrote, in an award-winning article for the Walrus, "Hair loss. If you want to wound a television performer, take his hair away." Here's a man who devoted his life to the service of truth and poetry, who in his dying days referenced his hair as among his chief instruments.
I'm never sure when people comment on quality whether the underlying question is not actually "does it have a good hair?" The woman pictured in the picture beside the Google knol has a thick head of hair. What could be better than that: a Wikipedia with good hair?
I think Wikipedia begins to stumble the moment it triggers the human grooming reflex among its contributors. At given point in time, 99% of the articles on Wikipedia are having a bad hair day. So far, it can be anything but consistently well groomed.
Nicely drawn stereotypes. You've earned yourself an A+ in Spin City 101. But to quote the immortal Marge Gunderson, "I'm not sure I agree with you a hundred percent on your police work, there, Lou." If you inspect the actual diffs, the writer pulled off the dread two-handed such/rather brain suck:
Such a link was never suggested by President Bush or the Bush administration as a justification for the invasion; rather, that such a relationship existed at all is seen as compelling.
Is anyone here factoring into all this the "cycle cost" of the batteries themselves? Most bottled electricity technologies (excepting supercaps) have a finite number of charge cycles before an expensive replacement/refurbish is required. It can depend on strange parameters such as the depth of the discharge, or relative frequency of deep discharge cycles.
How will it be handled when the power company cycles someone's fancy new battery pack into the recycle bin in under a month because a controller glitch caused them to change the charge/discharge direction twice an hour in one obscure subdivision?
Hard to believe he would self-publish that story. Did that really happen, or is it the self-promotional fantasy of a man who knows he'll never get laid?
Slashdot degrades itself when it runs stories by Dvorak. It can't look good from any perspective when half the regular membership is tagging a story submission as troll. I'll show you how it works. Watch me degrade myself by making references to Woody Allen. In fact, I did watch Sleeper the other night. (An interesting calculation: when Allen wakes up in the year 2173, Soon-Yi will be 203 years old. Gaaa! Given enough time, he'll prove us wrong yet.) Dvorak is the Howard Cosell of the IT industry, and that's probably paying him a complement he doesn't deserve.
I agree with another poster. It is not obvious for any software project that given an approach that is sometimes the right solution, that it benefits the indented audience to inflict support for that approach upon the feature set.
Typically when an application has narrow requirements that are so absolutely right you can't live without them, the application does not turn out to be suitable for RAD tools of any kind.
OTOH, if you can demonstrate that adding composite keys will *simplify* the design and implementation and use and documentation and performance and backward compatibility of Rails, then it should of course be added.
Indeed it's a sad day when slashdot submissions attempt to think.
What strikes me is that so many parties were infringing upon the patent, and yet you need very few organisations to file an opposition. Why are not more patents opposed?" http://en.wikipedia.org/wiki/Bystander_effect
You are promoting reproducibility far beyond its practical reality. If an analysis program is shown to have severe defects, any paper whose analysis is tied to that package will come into disrepute, until either A) the analysis of the data presented is repeated with a different program (assuming the raw data remains available); or B) the whole experiment is repeated (possibly also bigger and better) and the conclusion is corroborated by independent research.
It would be nice to think that every research paper out there archives it's original data for all eternity, so that the analysis can be repeated by anyone who cares to do so at any point in the future, but this is not the world we presently live in.
In many cases, the raw data has privacy implications, and is quite a hassle to reliably and safely archive.
Science, in any case, does not depend upon reproducible analysis; it depends upon reproducible experimental protocols.
I am a huge believer that open source should set the precedent, where the data allows this, that the original data *is* archived, along with the tools of its analysis, so that past mistakes *can* be investigated at any point down the road. This doesn't come so much from the culture of science as you claim. It comes from the culture of openness and long-term accountability which tools like Sage promote.
Besides, in most cases it's a fiction that a calculation can be reproduced (by practical effort) at some distant point in the future. Platforms change, code experiences bit rot, essential libraries mysteriously vanish. TeX, in particular, is notable for bucking this trend, and not without a great deal of insularity in its design, which is the polar opposite of what Sage has set out to accomplish.
You lost me at could be shown. The whole point of the nature/nurture debate is that it can't be shown.
At the risk of firing an arrow so far over your head it passes into an alternate universe, this can be quite readily demonstrated from the vantage point of cryptographic theory. This page describes a hash function which fails to achieve uniform cryptographic avalanche.
What this means is that correlations remain between which input bit is modified and the distribution of the output bits. Consider that some of these input bits are nature, other input bits are nurture, and if you like, set aside some of the bits as hidden state variables for the subject concerned.
There are correlates for some restricted input subsets, but few general correlates over the full range. The human organism is complex enough that nature and nurture interleave into partial avalanche. Too many sub-correlations are exposed to make it much use as a cryptographic function, but all the same far too intricate to think it could ever be shown that genes and outcomes exist in any fixed relationship.
This conceptual fallacy was also debunked in the context of Laffer curves. Check out the Neo-Laffer curve. It also serves as a good illustration of the typical complexity of any gene determining any definable outcome.
My cryptographic example merely shows that it is possible to draw a Laffer curve over multiple inputs and outputs with no correlation anywhere. Anyone here surprised that genetics can't be reduced to the smooth hump of a Reagan-era tax slogan?
Now, at some point some genes will be shown to have strong effects under broad conditions, but even then, not without a twenty page appendix of fine print covering assumptions and conditions. Conversely, many of our genes will prove to have such a baffling array of possible downstream effects, that soon the refrain from the peanut gallery will become "why do we bother, if nothing is ever proved?" Get used to it.
Curious logic there dude.
We adjust our system in such a way that their self-interest will work in our favor. I know there is a prevailing linguistic theory that us wear the white hats, and them wear the black hats, it's what every cartoon narrative teaches our young children. Nevertheless, I'd be interested to know what lever of power they don't already control. If you meant that we the people write the constitution, you haven't been following the trend in America lately toward legislation that honors the constitution with glowing phrases and a one finger salute.
Absolutely. When I program in C++ on top of a sane foundation, I have no expectation of leaking memory any more so than I would in any other language. What I might say, though, is that C++ is a little less tolerant of divided attention. I rarely program in C++ while monitoring IM sessions or RSS feeds. Maybe the proper definition of a scripting language is a language where you can program approximately between the lines while moaning about your weekend hangover to your buddies list.
Accident rate for helicopters in America is 1 per 100,000 flying hours (11.4 S.R.Hadden-years) and trending downward. This includes the significantly more dangerous reciprocating designs. Few helicopter pilots chat on IM while flying, and only a vanishing minority shows up to work within 48 hours of non-subsistence substance ingestion.
I think C++ partially gets the bad reputation because so many dabblers out there have applied the worse-is-better philosophy to rotor proliferation. It's easy to architect a system in C++ with seven rotors. Not recommended; however, there is nothing to stop you but your own good judgment. Which seems to be the core problem with C++.
Larry is right on the money with his distinction about what a language forces you to say. C++ can become terrifically cumbersome in this department, to the point where C++ is a bad language choice by default unless you know in advance what abstraction C++ brings to the table (with some work) that simplifies your problem domain.
For example, as a C++ programmer I know exactly what Larry was talking about in his dispatch-by-committee passage. You can easily achieve this in C++ (on a compile-time basis for the idiom I have in mind): create a template function that does nothing but dispatch according to any set of rules concerning the types supplied that you can devise (which can be supplied to the template as policy templates, if you wish to do so). These dispatch functions can become ugly creations, but manageable if you are careful and not chatting on IM at the same time.
Personally, I wouldn't code in any multi-dispatch language with a lingering hangover from the previous morning.
One more parting shot about "leakage". It's like Gere's "The gun, the gun, the gun" ditty in Chicago: the purpose of his obsession with the gun is to dim the mind of the jury on related concerns such as motive and opportunity. I don't think about leakage when I'm writing C++ code. I think about correctness, establishing my post-conditions, and not violating my preconditions. Obviously, this can't be achieved if you are programming on top of a hideous library where a complete statement of such is too complex (or too mysterious) to write down for any given routine you need to call. But then you can't write correct code by any metric on top of such a library, so it fails my comprehension that any serious programmer would brag about not leaking memory faced with such an abomination. Many a sailor has drowned going to sea in a ship that never leaked.
Actually, Slashdot banned 99.9% of the IP range from contributing to the front page, and left the rest of us with no ability other than entering into endless discussions on the talk page, none of which ever appear in Google's top ten search results for common search terms.
Yet one more data point in the universal requirement of introducing learned helplessness into any large-scale human interaction. Promise too much, you get a riot. Promise too little, the circumscribed become defenders of righteousness.
Even with the quick fingered (and occasionally abusive blocks) on Wikipedia, it remains 100 times more open than Slashdot. But since we never labored under the delusion that this was more than "news for nerds" social unrest on Slashdot has no ability to cure a slow news day on the Inquirer/Register (with their own sordid feuds to better help them condemn the feuds of others).
Lately the Inq/Reg are suffering from the hiatus in the Intel/AMD Nvidia/ATI sand squabbles. Eight cores. Yawn. Opteron errata. Yawn. The Everywhere Girl played out her string. Yawn. Fiorina and Cappellas are both gone. Now all they have left to scratch eyeballs is the Wackypedia. If AMD finally delivers a compelling product from their ATI ingestation, they might have eight weeks worth of IT news to report. And then what? I hate to think.
The fundamental problem at Wikipedia right now is that sock-puppets are conceptually impossible to patrol. In the old model, the problem was a fringe lunatic who needed to adjust his/her meds. In the old days, spam was likewise largely the result of isolated cowboys. The first round of defensive measures made it difficult to succeed at spam on a disorganized basis. Systems theory response? Shift to organized spam. The same will happen with sock puppets at Wikipedia, if interest groups perceive enough incentive to do so.
It would be a bit of work to pull off. You'd need several hundred identities in good standing, and enough "real" people to keep these identities in good standing with a diversity of edit histories. Maybe you can seed the pool by purchasing on the sly a long established identity from someone who has since fallen on hard times. You would probably need some orchestration to ensure that your puppets don't cross paths and congregate on the same issues too often.
I'm not saying I see the economic justification for anyone to go to this trouble, but if someone did, the existing sock-puppet defenses would not prove adequate to the task. I believe that this uncomfortable, naked fact that defense against orchestrated sock-puppets is presently unworkable contributes to the paranoia level of the beleaguered admins who teeter on the precipice of failure. I can recall a point in time when some of self-appointed spam-fighters got themselves into the same emotional state (and the same fights over abuse of blacklists erupting).
The crucial difference is that Wikipedia is centralized whereas spam was decentralized (although some of the largest players such as AOL, Yahoo, MSN, and Google could sample a large fraction of the effluent flow). Wikipedia needs to further exploit that centralization to find other mechanisms to accomplish the same end (preventing vested interests from banding together to shout down opposing perspectives) without having to resort to their shadowy order of paranoid puppet police.
The word "valid" is not a synonym for "prevail". Valid means worthy of respectful consideration by the other involved parties. An opinion such as "all opinions are equal (and deserve to prevail)" is unlikely to prevail because it doesn't lead to a workable outcome.
Intelligent design and evolution are less incompatible than most people think. What evolution teaches is that we have made a very large number or precise observations (across a broad swath of distinct disciplines ranging from astronomy to geology to biochemistry) that fit coherently within a 13-billion year story, with a handful of memorable moments in the theatrical trailer: 2.5Ga (toxigen), 500Ma (bilateria), 250Ma (first CGI extravaganza), 65Ma (second CGI extravaganza), 1.5Ma (habilis/erectus overlap), and 30ka (good riddance, hairy brute). Science digs up a lot of stuff that fits coherently within this narrative.
I.D. asserts that this could not have unfolded as described without guidance at the hand of an unexplained power, uh, the almighty basis step. Quite possible. It's quite possible the laws of probability were bent by a guiding force. Science can't readily dispute this. That whole period between 5Ma and 1.5Ma needs a little cleaning up before we tackle statistical plausibility of the narrative as a whole relative to any chosen cosmology.
It's also quite possible that the universe winked into existence at 8ka, with all the fossils embedded in granite and sandstone in their present configuration, and starlight of the released in space as-if it had been in transit for billions of years. Perhaps the almighty basis step likes to cover his/her/its creative seams.
What we need to teach in school is the coherence of the many observations we have gathered relative to the evolutionary narrative they fit within (thus far). There is no need whatsoever to assert that any of this actually happened. That's not a matter of science. Science describes, it does not explain. Evolution is a descriptive narrative, not a causative narrative.
I.D. is a causative narrative with little concern for unifying what sciences observes. It surprises me how little the I.D. movement cares about the handiwork of the almighty basis step. He/she/it apparently went to a lot of trouble to start things rolling with an entertaining yet exquisitely implausible back story.
We tend to forget that collusion, like any other business model, is subject to cost-benefit analysis. It's likely the cost of this spectrum will be high enough that the incentives will be aligned toward exploiting the resource for value, rather than choking it off to obtain collusive side-benefits.
I don't see the point of libertarian extrapolation of "free" markets. When you model the interactions of independent agents in game theory, the concept of "freedom" is nowhere to be found in the bare equations. Everyone makes their own choices and suffers the benefits and consequences accordingly.
What you can do is set up some rules of the game (cue sound of freedom pissing down the drain) so that the game is less likely to degenerate into fixed collusive alliances. Collusion is a fact of life in any multiparty competitive system. In much the same way that we attempt to prevent petty crime from amalgamating into organized crime, we also attempt to prevent petty (implicit) collusion for amalgamating into organized, institutional collusion. Certain small freedoms are sacrificed in the transmutation. The main freedom one gains in the exchange is the freedom to disassociate (aka "bugger off"). If two parties become entangled in an exploitative relationship which greatly benefits one party at the expense of the other party, the abused party can elect to disassociate, and enter into fresh relations that function on a more equitable footing.
Historically, there have been many social impediments to disassociation: slavery, marriage, citizenship, conscription, and debt. An emotional list. No accident there.
Debt it one of the smoother plays. A tin pot dictator accepts $100 million in loans from the World Bank, squirrels half of it away in Swiss bank accounts, then the (captive) citizenry toils for decades to repay this loan in support of the high 1st world standard of living. Sweet. When the abused citizenry moves to disassociate themselves from this corruption (which was of no original benefit to them in the first place) we label it an insurrection and bring out the tanks (the other half of the $100 million was well spent).
I don't see how an ideological extrapolation of freedom has much to contribute to this debate.
This has been understood in the industry for decades: in a given silicon process, power consumption fits roughly within the envelop V^2 * F, where F represents frequency. Given a process shrink, this relation might or might not hold true. For a long time it was a good rule of thumb, but then in the era of excessively high leakage current it did not hold true, more recently with better control over leakage, the relationship is again a good rule of thumb. The upshot is that, over two decades, almost every reduction in voltage for a given class of part corresponds to a significant increase in power efficiency.
What the article failed to explain is this long history of voltage serving as a proxy for power efficiency.
The other relationship is that a given part will usually demonstrate a relationship where lower frequencies are stable at lower voltages. If increasing the voltage by 20% allows you to overclock a processor from 2GHz to 3GHz, you can estimate your increased power draw as 1.2^2 * 1.5, about double where you started.
It's almost pointless to convert this measure into watts, as so many other variables change in tandem. The new part has different bandwidth, different latency, different leakage, different dynamic consumption. There's no simple number that gets you apples vs apples. Most of the time, however, voltage is fair proxy. Peak consumption figures are mostly worthless from an efficiency perspective, except for sizing your power and cooling requirements.
On a side note, I'm wondering when we hit the floor on practical CMOS voltage levels. Surely the band-gap will come into play in the near future, and then what? Does the efficiency graph suddenly develop a crimp and stagger forward on a much reduced slope? This happened with hard drives, where there was a period of accelerated capacity increase (PRML/GMR/pixie-dust era) only to return to the more sedate curve once again later on. It wasn't long ago that F hit thin air (due to thermal issues) and now F is increasing at half the rate it sustained for a least a decade prior.
Long ago apparently respectable sources used to proclaim "silicon will hit the brick wall at 0.1um". In turns out S-curves hardly ever play out that way. The curve begins to taper downward when the easy gains are exhausted. The phrase "peak oil" is another one of those conceptual nightmares, much like the chimeric brick-wall on photo lithography. It's not going to be a peak, is it? It's going to be a wavy plateau. On any particular graph, you can point to a "peak" (though none of the graphs will agree), it's just that there won't be a momentous Alderan-disturbance that ripples though planet earth as the precocious metaphor suggests. Much like the silicon people had to finally confess, driving F higher and higher as your primary performance metric (at the cost of absolute efficiency) makes about as much sense in the long run as a single-occupancy air-conditioned Hummer in rush hour traffic.
Speaking of which, engine displacement is roughly as fair as a measure in the automotive sector as voltage in silicon. It's the nature of the internal combustion engine that these engines are far from their peak efficiency at low to medium throttle, which is why having a lot of power you rarely use is no free lunch. If you accept that a typical 2 liter engine is more efficient than a typical 3 liter engine, why would voltage as a proxy for power be any different?
Regarding my previous post: that makes two markup failures in the same week. Time to shake the crusties out from underneath the punctuation plungers. I usually click "preview" when I'm insulting 300 million people. Yes, I suspect it's a good policy.
blockquote> Sure, but if you're above the people you are sacking, you look like you took corrective action to your own supervisor.
By the logic of this firing, firing the subordinate only after you become aware of the transgression is not soon enough. You my friend, are fired too. Darn, that escalates all the way to the top. Oops, loophole. The CEO recognizes the error of his ways right before it applies to him.
Michael Moore points out in Sicko that fear over loss of health care benefits (severely emphasized in an "at will" employment environment) greatly contributes to the docility of the working middle-income underclass. The big oaf compounds his irritable behaviour by sometimes being right. Baaa, America.
After reading the article, I'm surprised to read someone expressing highest respect for Mr Crockford, because my take from what I read was that he was just some fringe guy.
He introduces the idea of relaxing br and hr to accept the ambiguous empty form (which doesn't distinguish an empty tag from an unclosed tag error). Then he wraps up his discussion with the statement:
These changes significantly improve the reliability, security, and performance of HTML applications. I can see the merit of reducing the punctuation mark clutter on a couple of extremely common and "well known" (in the IANA sense) empty tags. But this constitutes an exception that introduces a new ambiguity to document parsing for the sake of removing a single punctuation mark whose insertion either by humans or in the application context is purely automatic. Where precisely is the simplicity in that trade-off?
I would not be opposed to introducing the concept of relaxed validation in which well-known empty tags are presumed to be self-closing. Any instance of a presumed self-closing tag used as an end-tag would fail relaxed validation (make up your mind, and stick to it). Relaxed validation could be allowed on the basis that translation to strong validation is purely mechanical.
My opinion is that anyone who makes the bare statement X simplifies Y is up to no good. The properly motivated form is: X simplifies Y for the sake of A at the expense of B. But then simplicity doesn't seem simple any more, and that defeats the purpose of vending the snake oil in the first place, to position oneself as the bearer of light and goodness, whose stock price is very high lately.
After writing this sans RSI entities, I discovered that ecode appears to function as a kind of nowiki with autoquote plus font change:
<e c o d e>&</e c o d e> renders &
Slash doesn't even provide a hyperlink to a description of the allowed HTML. Shamed by Wikipedia. How low can you go?
Slashdot Mirror
Math software shows up here fairly regularly. I keep taking notes, but never get around to using any. The R statistical package also gets frequent good mention, which I understand is accessible from within Sage.
Where does Octave stand relative to Sage?
And what's more, error handling needs to be part of the specification, because bad error handling creates road-blocks towards progressive adoption. How many good intentions toward standards have been thwarted by inscrutable error messages received on the bumpy road to enlightenment? As I recall it, the C standard explicitly makes no mention of error messages. This was not such a big deal as most syntactic errors were easily resolved. Then the C++ standard continued this tradition of not specifying error diagnostics while adding the template mechanism to the language. In my opinion this was a grossly negligent self-inflicted wound on the C++ community, and provided unnecessarily ammunition to the denigrationists who view C++ as a pox on humanity.
So yes, when a standard involves parsing down through multiple layers of syntax, it does matter to specify sane error responses. Unless you want your language/protocol to become known as the fat kid with pimples and a contagious itch.
Actually, one of the main challenges with natural language is that we humans perform so badly to begin with. Half the time we neither say what we mean, nor mean what we say. But it hardly matters: far more than half the time, the person (or people) listening hear either what they expected to hear, or what they wanted to hear, or they already knew they would disagree with whatever you were about to say before you even opened your mouth.
Sometimes it does matter. However, by the time you design a linguistic study to isolate the human gift for parsing grammar, the experimental task is about as "natural" as writing a law exam.
I think the contribution of grammar to early human language is way overstated. You don't need much grammar to handle everyday events, such as determining how to dress for dinner when the report from the field comes back "mammoth tusk hunter" or "hunter spear mammoth": in the former case (x3!) you'll be polishing your nose bone.
Where word order begins to matter is parsing the daily scuttlebutt. Did Adam tell Carol about Bob and Eve, or was it Eve telling Adam what she overheard between Bob and Carol? It's not easy keeping the cheaters distinct from the cheated upon. Plus Adam has to remember when to look surprised when Eve tells him something he learned from Carol just the other day. Not keeping your past/present/future and your cheatee/cheaters straight was a certain recipe for not sleeping on the warm side of the fire pit.
Later on, the grammar we acquired to parse who's zooming who became useful for digesting the BBQ assembly manual, but of course, that remains an evolutionary work in progress.
Maybe when children of the current MySpace generation reach the age to pop the big question ("What's an iPod?") and we've given up the fight to prevent our every indiscretion and peccadillo from being publicly archived for all posterity, we'll actually need a natural language interface to really drill down into the zettaflood of who said what to whom and who first posted it online and whether revenge was sweet.
For some definition of 90%. I think it will break down according to the 80-20 rule: the 20% of applications that produce 80% of the value will remain on the desktop, while the 80% of applications that produce 20% of the value will migrate to the network. You outsource a large slice of your IT hassle, but only lose 20% of your activity for the duration of a network outage, which are fairly rare events if you have a good ISP. If Google apps takes off, it will probably drive demand for a more reliable last mile, and even small companies will not have much reluctance to pay for this if their IT costs are significantly reduced.
It really doesn't make a lot of sense to have your in-house IT people supporting the 80% that produces 20% of your value.
No doubt the NSA is quite pleased to see people yammering on about encryption as if that mattered. On a practical level, the most important data stream is traffic analysis: who is talking to whom. Traffic analysis forms a graph. What's a graph worth? Tough question. Perhaps ask Google.
... traffic analysis ... unless the node is almost completely private. A private steganographic Tor node is your best cover, but it will severely compromise your bandwidth and latency. We've all got one of those in our back pockets. Neither would I wish to be apprehended in possession of that particular coke spoon: it has Gitmo express written all over it.
On this graph, what does the NSA know about the various nodes and links? They know who uses crypto, and when. They know who you talk to and in what order, the duration, and the interval between communications. They also know if you are talking to a Tor node, or any other node suspected of laundering one or more end-points.
Think you have a secret Tor node. Well, these nodes are fairly easy to detect by
For any node or link where the probers that be have the slightest suspicion, they can also determine in the majority of all cases what protocols you are running. Even if you tunnel your anonymity through SSL, packet timing profiles is likely to leak significant information about the protocols employed. Even if you leak no timing information, you distinguish your SSL segment from every other protocol which does.
With commercial software, it is almost impossible to know if the NSA hasn't found some clever way to leak key bits through the "random" number generator. What to do? Obtain hard core anon steg crypto from an open source project? Don't be seen doing it. That will flag your packets upward for years to come. Maybe through your functioning personal anon steg Tor server? That poses two problems: first of all, you don't have one yet. And second, even if your recently purchased four-digit Slashdot ID from the Dread Pirate Roberts included a secret anon steg Tor treasure map, your anon steg Tor server is severely bandwidth constrained (supposing you wish to continue flying under the radar much longer).
Even supposing, how well do you actually choose your password? It's a virtual certainty the NSA maintains a list of the billion most popular passwords in every language of the world, plus the one million most popular mnemonic devices (including all forms of keyboard mambo), and all the most popular transformations of the former against the later. Think your mnemonic device is unique? Guess again. Not unless it is almost as hard to remember as the password it replaces. Have you ever used any portion of that password in a context less strongly encrypted (such as any Microsoft or Apple program ever written?) The elephant never forgets. Remembering is cheap, cracking costs money.
Should you actually prevail and manage to conduct your electronic communications immune to any form of analysis yet mentioned, congratulations, you have now achieved exclusive membership among the hardened targets of the world, where brute force and white vans finally make good economic sense.
Those little chocolate mints that showed up on your pillowcase the other day are the sign of a job well done.
The mass of a vehicle makes very little difference to highway mileage in non-mountainous terrain. It does matter for start/stop traffic, such as quick trips back to the 200,000 sq ft hardware store in the not-so-local mall because you bought the wrong bolt.
The reason heavy vehicles have much poorer MPG ratings is that the engine is usually sized to achieve inside-lane passing speed while pulling a 10% grade with several passengers and as much again in cargo. Excess displacement kills at low load. If you can't drop the hammer and pass two Winnebagos in 500m while cresting into Boulder the studly man-mobile is worthless to 90% of blue collar America. That's a lot of displacement for fetching the bolt you should have bought in the first place.
First, if this new scheme works, it will likely lead to Google ranking their own knols up and Wikipedia pages down, which could severely reduce Wikipedia's reach. I think there is a conflict of interest brewing here, unless the Google knols are displayed among the ad results.
As for article quality on Wikipedia, I think Fermi estimation applies: you get 80% quality by throwing many bits of text into the air to see what sticks, but rarely much above that. If you think about it, the original Nupedia approach generated quality at an appalling pace. It's never been obvious to me why quality suddenly becomes easier after you have erected 2 million shacks in various stages of dilapidation. Most the shacks have something useful inside, few are pretty to look at.
It not surprising, either, that many of the people who thrived at Wikipedia in the barn raising stage are leaving the project as it meanders sideways when confronted with a much harder problem. One could argue that Wikipedia is the most successful knock-up ever.
I've never ceased to be amused by the number of people who stumble upon Wikipedia in relation to pop culture, then learn about the vastness and rapid growth and conclude "well, if everyone is doing this, it must be good" only to examine the actual articles and find obvious flaws abounding. Anyone who uses MySpace as their basis step (if everyone is doing it, it must be good) was due for a rude eye opening to begin with.
When Bill Cameron, a wonderful CBC news personality, was dying of cancer he wrote, in an award-winning article for the Walrus, "Hair loss. If you want to wound a television performer, take his hair away." Here's a man who devoted his life to the service of truth and poetry, who in his dying days referenced his hair as among his chief instruments.
I'm never sure when people comment on quality whether the underlying question is not actually "does it have a good hair?" The woman pictured in the picture beside the Google knol has a thick head of hair. What could be better than that: a Wikipedia with good hair?
I think Wikipedia begins to stumble the moment it triggers the human grooming reflex among its contributors. At given point in time, 99% of the articles on Wikipedia are having a bad hair day. So far, it can be anything but consistently well groomed.
Is anyone here factoring into all this the "cycle cost" of the batteries themselves? Most bottled electricity technologies (excepting supercaps) have a finite number of charge cycles before an expensive replacement/refurbish is required. It can depend on strange parameters such as the depth of the discharge, or relative frequency of deep discharge cycles.
How will it be handled when the power company cycles someone's fancy new battery pack into the recycle bin in under a month because a controller glitch caused them to change the charge/discharge direction twice an hour in one obscure subdivision?
Hard to believe he would self-publish that story. Did that really happen, or is it the self-promotional fantasy of a man who knows he'll never get laid?
http://drrobertepstein.com/downloads/FROM_RUSSIA_WITH_LOVE-Epstein-Sci_Am_Mind-Oct-Nov2007.pdf
I'll believe in the chatbot when it dives under my desk for existential reasons.
http://dir.salon.com/story/books/review/2004/05/21/slater/
Slashdot degrades itself when it runs stories by Dvorak. It can't look good from any perspective when half the regular membership is tagging a story submission as troll. I'll show you how it works. Watch me degrade myself by making references to Woody Allen. In fact, I did watch Sleeper the other night. (An interesting calculation: when Allen wakes up in the year 2173, Soon-Yi will be 203 years old. Gaaa! Given enough time, he'll prove us wrong yet.) Dvorak is the Howard Cosell of the IT industry, and that's probably paying him a complement he doesn't deserve.
I agree with another poster. It is not obvious for any software project that given an approach that is sometimes the right solution, that it benefits the indented audience to inflict support for that approach upon the feature set.
Typically when an application has narrow requirements that are so absolutely right you can't live without them, the application does not turn out to be suitable for RAD tools of any kind.
OTOH, if you can demonstrate that adding composite keys will *simplify* the design and implementation and use and documentation and performance and backward compatibility of Rails, then it should of course be added.
You are promoting reproducibility far beyond its practical reality. If an analysis program is shown to have severe defects, any paper whose analysis is tied to that package will come into disrepute, until either A) the analysis of the data presented is repeated with a different program (assuming the raw data remains available); or B) the whole experiment is repeated (possibly also bigger and better) and the conclusion is corroborated by independent research.
It would be nice to think that every research paper out there archives it's original data for all eternity, so that the analysis can be repeated by anyone who cares to do so at any point in the future, but this is not the world we presently live in.
In many cases, the raw data has privacy implications, and is quite a hassle to reliably and safely archive.
Science, in any case, does not depend upon reproducible analysis; it depends upon reproducible experimental protocols.
I am a huge believer that open source should set the precedent, where the data allows this, that the original data *is* archived, along with the tools of its analysis, so that past mistakes *can* be investigated at any point down the road. This doesn't come so much from the culture of science as you claim. It comes from the culture of openness and long-term accountability which tools like Sage promote.
Besides, in most cases it's a fiction that a calculation can be reproduced (by practical effort) at some distant point in the future. Platforms change, code experiences bit rot, essential libraries mysteriously vanish. TeX, in particular, is notable for bucking this trend, and not without a great deal of insularity in its design, which is the polar opposite of what Sage has set out to accomplish.
You lost me at could be shown. The whole point of the nature/nurture debate is that it can't be shown.
At the risk of firing an arrow so far over your head it passes into an alternate universe, this can be quite readily demonstrated from the vantage point of cryptographic theory. This page describes a hash function which fails to achieve uniform cryptographic avalanche.
http://home.comcast.net/~bretm/hash/6.html
What this means is that correlations remain between which input bit is modified and the distribution of the output bits. Consider that some of these input bits are nature, other input bits are nurture, and if you like, set aside some of the bits as hidden state variables for the subject concerned.
There are correlates for some restricted input subsets, but few general correlates over the full range. The human organism is complex enough that nature and nurture interleave into partial avalanche. Too many sub-correlations are exposed to make it much use as a cryptographic function, but all the same far too intricate to think it could ever be shown that genes and outcomes exist in any fixed relationship.
This conceptual fallacy was also debunked in the context of Laffer curves. Check out the Neo-Laffer curve. It also serves as a good illustration of the typical complexity of any gene determining any definable outcome.
http://en.wikipedia.org/wiki/Laffer_curve#The_Neo-Laffer_Curve
My cryptographic example merely shows that it is possible to draw a Laffer curve over multiple inputs and outputs with no correlation anywhere. Anyone here surprised that genetics can't be reduced to the smooth hump of a Reagan-era tax slogan?
Now, at some point some genes will be shown to have strong effects under broad conditions, but even then, not without a twenty page appendix of fine print covering assumptions and conditions. Conversely, many of our genes will prove to have such a baffling array of possible downstream effects, that soon the refrain from the peanut gallery will become "why do we bother, if nothing is ever proved?" Get used to it.
It can't be shown.
Absolutely. When I program in C++ on top of a sane foundation, I have no expectation of leaking memory any more so than I would in any other language. What I might say, though, is that C++ is a little less tolerant of divided attention. I rarely program in C++ while monitoring IM sessions or RSS feeds. Maybe the proper definition of a scripting language is a language where you can program approximately between the lines while moaning about your weekend hangover to your buddies list.
http://www.rotor.com/portals/12/Safety2002-2006.pdf
Accident rate for helicopters in America is 1 per 100,000 flying hours (11.4 S.R.Hadden-years) and trending downward. This includes the significantly more dangerous reciprocating designs. Few helicopter pilots chat on IM while flying, and only a vanishing minority shows up to work within 48 hours of non-subsistence substance ingestion.
I think C++ partially gets the bad reputation because so many dabblers out there have applied the worse-is-better philosophy to rotor proliferation. It's easy to architect a system in C++ with seven rotors. Not recommended; however, there is nothing to stop you but your own good judgment. Which seems to be the core problem with C++.
Larry is right on the money with his distinction about what a language forces you to say. C++ can become terrifically cumbersome in this department, to the point where C++ is a bad language choice by default unless you know in advance what abstraction C++ brings to the table (with some work) that simplifies your problem domain.
For example, as a C++ programmer I know exactly what Larry was talking about in his dispatch-by-committee passage. You can easily achieve this in C++ (on a compile-time basis for the idiom I have in mind): create a template function that does nothing but dispatch according to any set of rules concerning the types supplied that you can devise (which can be supplied to the template as policy templates, if you wish to do so). These dispatch functions can become ugly creations, but manageable if you are careful and not chatting on IM at the same time.
Personally, I wouldn't code in any multi-dispatch language with a lingering hangover from the previous morning.
One more parting shot about "leakage". It's like Gere's "The gun, the gun, the gun" ditty in Chicago: the purpose of his obsession with the gun is to dim the mind of the jury on related concerns such as motive and opportunity. I don't think about leakage when I'm writing C++ code. I think about correctness, establishing my post-conditions, and not violating my preconditions. Obviously, this can't be achieved if you are programming on top of a hideous library where a complete statement of such is too complex (or too mysterious) to write down for any given routine you need to call. But then you can't write correct code by any metric on top of such a library, so it fails my comprehension that any serious programmer would brag about not leaking memory faced with such an abomination. Many a sailor has drowned going to sea in a ship that never leaked.
Actually, Slashdot banned 99.9% of the IP range from contributing to the front page, and left the rest of us with no ability other than entering into endless discussions on the talk page, none of which ever appear in Google's top ten search results for common search terms.
Yet one more data point in the universal requirement of introducing learned helplessness into any large-scale human interaction. Promise too much, you get a riot. Promise too little, the circumscribed become defenders of righteousness.
Even with the quick fingered (and occasionally abusive blocks) on Wikipedia, it remains 100 times more open than Slashdot. But since we never labored under the delusion that this was more than "news for nerds" social unrest on Slashdot has no ability to cure a slow news day on the Inquirer/Register (with their own sordid feuds to better help them condemn the feuds of others).
Lately the Inq/Reg are suffering from the hiatus in the Intel/AMD Nvidia/ATI sand squabbles. Eight cores. Yawn. Opteron errata. Yawn. The Everywhere Girl played out her string. Yawn. Fiorina and Cappellas are both gone. Now all they have left to scratch eyeballs is the Wackypedia. If AMD finally delivers a compelling product from their ATI ingestation, they might have eight weeks worth of IT news to report. And then what? I hate to think.
The fundamental problem at Wikipedia right now is that sock-puppets are conceptually impossible to patrol. In the old model, the problem was a fringe lunatic who needed to adjust his/her meds. In the old days, spam was likewise largely the result of isolated cowboys. The first round of defensive measures made it difficult to succeed at spam on a disorganized basis. Systems theory response? Shift to organized spam. The same will happen with sock puppets at Wikipedia, if interest groups perceive enough incentive to do so.
It would be a bit of work to pull off. You'd need several hundred identities in good standing, and enough "real" people to keep these identities in good standing with a diversity of edit histories. Maybe you can seed the pool by purchasing on the sly a long established identity from someone who has since fallen on hard times. You would probably need some orchestration to ensure that your puppets don't cross paths and congregate on the same issues too often.
I'm not saying I see the economic justification for anyone to go to this trouble, but if someone did, the existing sock-puppet defenses would not prove adequate to the task. I believe that this uncomfortable, naked fact that defense against orchestrated sock-puppets is presently unworkable contributes to the paranoia level of the beleaguered admins who teeter on the precipice of failure. I can recall a point in time when some of self-appointed spam-fighters got themselves into the same emotional state (and the same fights over abuse of blacklists erupting).
The crucial difference is that Wikipedia is centralized whereas spam was decentralized (although some of the largest players such as AOL, Yahoo, MSN, and Google could sample a large fraction of the effluent flow). Wikipedia needs to further exploit that centralization to find other mechanisms to accomplish the same end (preventing vested interests from banding together to shout down opposing perspectives) without having to resort to their shadowy order of paranoid puppet police.
The word "valid" is not a synonym for "prevail". Valid means worthy of respectful consideration by the other involved parties. An opinion such as "all opinions are equal (and deserve to prevail)" is unlikely to prevail because it doesn't lead to a workable outcome.
Intelligent design and evolution are less incompatible than most people think. What evolution teaches is that we have made a very large number or precise observations (across a broad swath of distinct disciplines ranging from astronomy to geology to biochemistry) that fit coherently within a 13-billion year story, with a handful of memorable moments in the theatrical trailer: 2.5Ga (toxigen), 500Ma (bilateria), 250Ma (first CGI extravaganza), 65Ma (second CGI extravaganza), 1.5Ma (habilis/erectus overlap), and 30ka (good riddance, hairy brute). Science digs up a lot of stuff that fits coherently within this narrative.
I.D. asserts that this could not have unfolded as described without guidance at the hand of an unexplained power, uh, the almighty basis step. Quite possible. It's quite possible the laws of probability were bent by a guiding force. Science can't readily dispute this. That whole period between 5Ma and 1.5Ma needs a little cleaning up before we tackle statistical plausibility of the narrative as a whole relative to any chosen cosmology.
It's also quite possible that the universe winked into existence at 8ka, with all the fossils embedded in granite and sandstone in their present configuration, and starlight of the released in space as-if it had been in transit for billions of years. Perhaps the almighty basis step likes to cover his/her/its creative seams.
What we need to teach in school is the coherence of the many observations we have gathered relative to the evolutionary narrative they fit within (thus far). There is no need whatsoever to assert that any of this actually happened. That's not a matter of science. Science describes, it does not explain. Evolution is a descriptive narrative, not a causative narrative.
I.D. is a causative narrative with little concern for unifying what sciences observes. It surprises me how little the I.D. movement cares about the handiwork of the almighty basis step. He/she/it apparently went to a lot of trouble to start things rolling with an entertaining yet exquisitely implausible back story.
We tend to forget that collusion, like any other business model, is subject to cost-benefit analysis. It's likely the cost of this spectrum will be high enough that the incentives will be aligned toward exploiting the resource for value, rather than choking it off to obtain collusive side-benefits.
I don't see the point of libertarian extrapolation of "free" markets. When you model the interactions of independent agents in game theory, the concept of "freedom" is nowhere to be found in the bare equations. Everyone makes their own choices and suffers the benefits and consequences accordingly.
What you can do is set up some rules of the game (cue sound of freedom pissing down the drain) so that the game is less likely to degenerate into fixed collusive alliances. Collusion is a fact of life in any multiparty competitive system. In much the same way that we attempt to prevent petty crime from amalgamating into organized crime, we also attempt to prevent petty (implicit) collusion for amalgamating into organized, institutional collusion. Certain small freedoms are sacrificed in the transmutation. The main freedom one gains in the exchange is the freedom to disassociate (aka "bugger off"). If two parties become entangled in an exploitative relationship which greatly benefits one party at the expense of the other party, the abused party can elect to disassociate, and enter into fresh relations that function on a more equitable footing.
Historically, there have been many social impediments to disassociation: slavery, marriage, citizenship, conscription, and debt. An emotional list. No accident there.
Debt it one of the smoother plays. A tin pot dictator accepts $100 million in loans from the World Bank, squirrels half of it away in Swiss bank accounts, then the (captive) citizenry toils for decades to repay this loan in support of the high 1st world standard of living. Sweet. When the abused citizenry moves to disassociate themselves from this corruption (which was of no original benefit to them in the first place) we label it an insurrection and bring out the tanks (the other half of the $100 million was well spent).
I don't see how an ideological extrapolation of freedom has much to contribute to this debate.
This has been understood in the industry for decades: in a given silicon process, power consumption fits roughly within the envelop V^2 * F, where F represents frequency. Given a process shrink, this relation might or might not hold true. For a long time it was a good rule of thumb, but then in the era of excessively high leakage current it did not hold true, more recently with better control over leakage, the relationship is again a good rule of thumb. The upshot is that, over two decades, almost every reduction in voltage for a given class of part corresponds to a significant increase in power efficiency.
What the article failed to explain is this long history of voltage serving as a proxy for power efficiency.
The other relationship is that a given part will usually demonstrate a relationship where lower frequencies are stable at lower voltages. If increasing the voltage by 20% allows you to overclock a processor from 2GHz to 3GHz, you can estimate your increased power draw as 1.2^2 * 1.5, about double where you started.
It's almost pointless to convert this measure into watts, as so many other variables change in tandem. The new part has different bandwidth, different latency, different leakage, different dynamic consumption. There's no simple number that gets you apples vs apples. Most of the time, however, voltage is fair proxy. Peak consumption figures are mostly worthless from an efficiency perspective, except for sizing your power and cooling requirements.
On a side note, I'm wondering when we hit the floor on practical CMOS voltage levels. Surely the band-gap will come into play in the near future, and then what? Does the efficiency graph suddenly develop a crimp and stagger forward on a much reduced slope? This happened with hard drives, where there was a period of accelerated capacity increase (PRML/GMR/pixie-dust era) only to return to the more sedate curve once again later on. It wasn't long ago that F hit thin air (due to thermal issues) and now F is increasing at half the rate it sustained for a least a decade prior.
Long ago apparently respectable sources used to proclaim "silicon will hit the brick wall at 0.1um". In turns out S-curves hardly ever play out that way. The curve begins to taper downward when the easy gains are exhausted. The phrase "peak oil" is another one of those conceptual nightmares, much like the chimeric brick-wall on photo lithography. It's not going to be a peak, is it? It's going to be a wavy plateau. On any particular graph, you can point to a "peak" (though none of the graphs will agree), it's just that there won't be a momentous Alderan-disturbance that ripples though planet earth as the precocious metaphor suggests. Much like the silicon people had to finally confess, driving F higher and higher as your primary performance metric (at the cost of absolute efficiency) makes about as much sense in the long run as a single-occupancy air-conditioned Hummer in rush hour traffic.
Speaking of which, engine displacement is roughly as fair as a measure in the automotive sector as voltage in silicon. It's the nature of the internal combustion engine that these engines are far from their peak efficiency at low to medium throttle, which is why having a lot of power you rarely use is no free lunch. If you accept that a typical 2 liter engine is more efficient than a typical 3 liter engine, why would voltage as a proxy for power be any different?
Regarding my previous post: that makes two markup failures in the same week. Time to shake the crusties out from underneath the punctuation plungers. I usually click "preview" when I'm insulting 300 million people. Yes, I suspect it's a good policy.
blockquote>
Sure, but if you're above the people you are sacking, you look like you took corrective action to your own supervisor.
By the logic of this firing, firing the subordinate only after you become aware of the transgression is not soon enough. You my friend, are fired too. Darn, that escalates all the way to the top. Oops, loophole. The CEO recognizes the error of his ways right before it applies to him.
Michael Moore points out in Sicko that fear over loss of health care benefits (severely emphasized in an "at will" employment environment) greatly contributes to the docility of the working middle-income underclass. The big oaf compounds his irritable behaviour by sometimes being right. Baaa, America.
He introduces the idea of relaxing br and hr to accept the ambiguous empty form (which doesn't distinguish an empty tag from an unclosed tag error). Then he wraps up his discussion with the statement: These changes significantly improve the reliability, security, and performance of HTML applications. I can see the merit of reducing the punctuation mark clutter on a couple of extremely common and "well known" (in the IANA sense) empty tags. But this constitutes an exception that introduces a new ambiguity to document parsing for the sake of removing a single punctuation mark whose insertion either by humans or in the application context is purely automatic. Where precisely is the simplicity in that trade-off?
I would not be opposed to introducing the concept of relaxed validation in which well-known empty tags are presumed to be self-closing. Any instance of a presumed self-closing tag used as an end-tag would fail relaxed validation (make up your mind, and stick to it). Relaxed validation could be allowed on the basis that translation to strong validation is purely mechanical.
My opinion is that anyone who makes the bare statement X simplifies Y is up to no good. The properly motivated form is: X simplifies Y for the sake of A at the expense of B. But then simplicity doesn't seem simple any more, and that defeats the purpose of vending the snake oil in the first place, to position oneself as the bearer of light and goodness, whose stock price is very high lately.
After writing this sans RSI entities, I discovered that ecode appears to function as a kind of nowiki with autoquote plus font change: Slash doesn't even provide a hyperlink to a description of the allowed HTML. Shamed by Wikipedia. How low can you go?