Take a look at Unskilled and Unaware of It: How Difficulties in Recognizing One's Own Incompetence Lead to Inflated Self-Assessments for an interesting look at why results like that are to be expected.
Ah, thanks for beating me to this recommendation.:-) A cute point about this particular paper is that it actually won an Ig Nobel Prize a few years back. Now, it sure ain't a flawless piece of work, but it is a result that you ignore at your peril. For those who won't bother to click through or read the linked paper, the punch line is exactly what the title says: not only do *most* people from a given population think they're at about the 60th percentile in ability for X, for almost values of X, but they do not correct their inflated self-assessments even when confronted with data that should clue them in. So, you might think that somebody who was in the bottom 10% but who thought they were better than the average student at, say, "proper" English grammar could recognize that this might not be true if you confronted him or her with their own written work and a representative sample of student work. But they don't; if anything, they now think they are even better than they did before.
Now, I suppose the Ig Nobel was awarded to them because in some sense this is a "duh" result. But the real point is that it really does completely crush what might seem to be an obvious and humane teaching strategy: provide students with models of superior work and have them strive to meet that ideal. I hope some of you just had your blood run cold when you just realized why this won't work.
Now it gets even better once you realize that
this same effect can help explain why education
about science and technology is especially hard
to design. A big strong argument in favor of Real Science in comparison with PseudoScience is that the Real kind eventually leads to very tangible yet nearly miraculous things. So Real Science gives rise to miraculous stuff like rewritable CD players and genetic engineering, while astrology
and ESP only seem to lead to bad TV specials. Now, you think that this difference would be clear, and that you would listen to the people who brought you the Magical Machines when they point out that astrology is complete crap. But they
don't.
Imitation is the sincerest form of flattery, right? I don't think ripoff is the right word, either. I posted the link to the Eckel's comment because I hoped to see some reaction from Ruby fans. I admire Eckel, but everyone is wrong sometimes, and Ruby folks should be the first to
have valid arguments with Eckel's points.
Indeed. I think a lot of Ruby users don't find themselves very interested in this issue any more. For a wide sampling of opinion on this and related issues, back in the day, you might be interested in this 138-post thread
Or maybe not.:-) More seriously, the best response in this thread is, I think, the second
one, by the author of the book involved.
OTOH, I'm still not totally convinced that importable modules can actually replace multiple inheritence. So far I haven't run into any show stoppers, but I keep having the nagging feeling that one is lurking somewhere in the future... and that by the time I hit it I may have several thousand lines of code to re-write.
Given the number of object-oriented languages (including Java) that have been doing Just Fine
without MI, I suspect your nagging feeling is
probably just needless worry. I have no doubt
that your code could have a different conceptual
feel without multiple inheritance, but, hey,
single (implementation) inheritance itself is probably over-used. Stepanov, god of the STL would argue (I think) that what people really want in most cases is generic programming on well-structured data. Interestingly, Ruby can
be used in this way, too, despite the pretensions
to being "pure OO".
"Why did the language designers bother doing this? So far I keep coming to the conclusion that Ruby is just a bad ripoff of Python"...
http://www.mindview.net/Etc/FAQ.html#Ruby
Yes, people in what you could call the Ruby Community did in fact wonder what Bruce Eckel
was thinking when he wrote that. This quote is particularly weird given that he goes on to point out that the syntax is more Perlish. I think what is really going on here is that Ruby did fairly self-consciously tried to implement some features of particular languages that the author thought would be especially useful. But a lot of the similarities with various scripting languages are almost certainly less "rip-off" than the inevitable consequence of working in a constrained design space. It is actually quite reminiscent of a situation you see in plagiarism lawsuits against screenwriters who are accused of "clearly copying"
some part of their work from that of some amateur
they may or may not have barely seen once upon a time. But it turns out that the courts recognize that once you have the nucleus of a specific idea,
a lot of plot and character similarities basically come along for the ride.
The real reason Ruby got written was that the author thought it was a better language for his uses than the alternatives available to him. Obviously, you won't get universal agreement on something like this, but to assume that something
like Ruby would be an ignorant rip-off of your favorite language strikes me as a bit weird.
Even if it is more popular in Japan, who the heck cares? Python is more widely used, supported, etc. everywhere else.
The problem of support is, of course, a real one in many applications. The fallacy of the "who cares?" argument here is that the same argument could probably have been used against every programming language since the first
assembler was written. Not every new language "makes it" by a long shot, but not every
well-supported language lives forever, either. Python was barely on people's radar for the very
longest time; it has succeeded because enough people were convinced it had advantages that outweighed a temporary (or possibly even permanent) disadvantage in publicity, support, or
whatever reason-to-fail you might give.
I choose Python over Ruby because Ruby is actually *proud* of its Perl heritage. Gimme a break!
Well, you can be proud of your heritage without
repeating the sins of your parents.:-) As far
as I can tell (I've programmed in Perl since 1991
and in Ruby for a few months), Ruby retains much
of Perl's flexibility while hugely reducing
the amount of (non-regexp) "line noise" feel that
even I think Perl sometimes has, and dramatically reducing the number of oddball special cases. I
know you could say the same thing about Python,
but I never really "got into" Python, for whatever
reason. Like it or not, I think we all realize
that tastes vary somewhat in these things.
The Ruby Programming Language, Matsumoto Yukihiro, ISBN:020171096X
I would love to see this book get published, but the last I heard, as suggested by this
message from the author himself, the book has been cancelled. Does anybody know if they
changed their minds on this or found a new publisher (Addison Wesley did the original contract)?
[1] Insert obligatory "51st state" or "not a real country" or "Royalist traitor" joke here.
:)
First off, congratulations on holding off (longer
than I did) on the grading of the papers that I predict are currently stacked on the "visitor" chair in your office.:-)
Secondly, what part of Canada would you really
want to have in the US anyway? OK, so B.C. might
be worth something, but anything else? Alberta would end up being just as much a PITA as Montana and Wyoming currently are, and we already have a full complement of boring flat states, so Manitoba and Saskatchewan are out. The Maritime provinces are basically Appalachia with a (pretty, I'll admit) coastline, Quebec is over-run with people who speak French, and that leaves
you with Ontario. Do we have any takers for a
bankrupt, boring, grid-locked province filled with
former Nortel employees? Didn't think so.
So let's *not* joke about making Canada our 51st
state; if we really wanted it that badly, we surely would have tried harder back in 1812.:-)
Horror Story 1: Delivered an HP something-or-other with 15" monitor to some 40 year old yuppie woman. She wants the thing setup on her roll-top desk. Only problem is that with the monitor setup, the roll-top wouldn't close. Rather than just leaving the roll-top open, though, or moving the computer to some other desk (of which she had plenty), we had to find her a monitor that would fit under that roll-top. We end up putting a floor model
15" that had a slightly different form, but was over a year older. All because the roll-top had to close.
OK, sport, so what's wrong with wanting your roll-top desk to close? Seriously, this is an
article thread about the death of ugly beige boxes which was brought on in part
by the fact that people are using computers more
in the home, but don't want to have something as
hideous looking as your average PC there. Or, let's put it another way: geeks are quite well-known for their refusal to compromise on principle, and often get insults or abuse as their reward for this. Now, you may be unfamiliar with
the idea or unsympathetic to the notion that most
traditional computer monitors make really crummy additions to virtually every decor. But, of course, the first thing you do is insult and try to make fun of somebody who has a different idea.
Forgive me if I'm just not very impressed.
I really like Mac OS X, and I would use it for
pretty much everything in life...except
there is no native version of Matlab for it.
To be brutally honest about it, the lack of Matlab is probably costing Apple tens of thousands of lost
sales per year (really! I seriously do mean it!). If Apple is serious about selling the new Macs to technical users, they really, truly are going to have to do something about the complete lack of Matlab. Heck, it might even be worth it for them to jump into Octave development with both feet
and bring *that* up to speed. R is wonderful,
Mathematica is beautiful, but when you have to
use Matlab, there really is no substitute.:-(
Tax advantages aside, you can't sell what you do not have. Any complaints to the effect that Amazon or anyone else are ripping off authors by not promoting new copies have to address the fact that so much material is OOP.
I guess I was unclear on this. My point was that
one reason (among others) for so much being OOP
was that neither publishers nor wholesalers could
afford (with the new depreciation rules) to carry
as much backstock as they used to. In print really only meant that you could order copies from the publisher, not that the smell of hot ink was still in the air.:-) The surprise then was how little it turned out to matter in terms of industry profitability.
The difference between selling a paperback novel at $25 and selling it at $10 is one of seeing it sit gathering dust and seeing it sell. I agree the economics means that you'll want to price some books at more than others, but there's such a thing as excessive.
I will admit that I don't read many novels anymore, but I *think* the most expensive one I remember seeing recently was more like $20 in
paperback. In other words, waaay more than mass market, to be sure, but not $25 per. The great
thing about the economics of this is that we have
a good definition of "excessive price" we can use:
the price where your overall profit would be higher if you lowered the the price. The trick,
of course, is knowing what that price is for any
given book... And the tragedy is that most books
don't make any money, so you'd be just as well off
selling copies for the production cost and then "updating" the price stickers for the few titles
that you could essentially give away...
Oh, and a note to the authors and publishers: If you really don't want us to buy second hand books:
Keep your books in-print. That's, believe it or not, the #1 rule. People can't buy new copies if you refuse to sell them.
Don't try selling paperback fiction for $25. I don't care how good the book is, that's excessive.
Keep all types of your book (hardback, paperback, etc) in print, rather than just the MMP.
This is an interesting list, but I'm not sure these are moves that would benefit publishers, and it
is publishers who would have to make these moves.
As far as keeping books in print goes, the tricky
part there is that the inventory you build up that
way becomes a hot potato under the current
tax and accounting regulations, as far as I can understand the situation. Back in the 1980s, I believe, rules for the depreciation of certain kinds of assets (like books piled up in warehouses) changed, and after that point, neither publishers nor sellers really wanted to warehouse a lot of stuff. Moreover, it became clearer and
clearer that the chance that you would publish a
"sleeper" that would suddenly more than justify
your having a large initial print run two years
down the line really were not very large. Your
profits are greater if you can use marketing to
move a more limited number of titles from "proven" authors. All the back catalog stuff is there
basically to find the next possible Hot Author that
you can market...for their next book. Whose rights you are likely to own, given the standard
contract.
As far as the price of books goes, I'm afraid I
don't see very much evidence that lower prices
would increase profits much, either. The books
that sell best are the books that sell well, and
publishers don't have to discount these much (although booksellers might). If you've got a title that does not take off at $25 a copy (in a
relative way), then you don't really expect it to
magically take off at $15 or $10, and even if it
did, you'd have to do a huge volume to make up
for the discount. (Plus, authors royalties would
certainly not go up in this scenario.)
There are two real problem here. One is that individual books are not commodities in the same way that many other goods are. If everybody is
reading Grisham and you want to talk with those
people, reading a discounted book by Bette R. Author won't necessarily achieve that goal. Another problem is that there is not very much risk in the business for publishers, since they
have a diversified "portfolio" of offerings. Individual authors only have an interest
in their specific books. So authors bear a lot of the risk, but have no say in many of the
decisions that are made to buy and sell and price
and market. I can see how this could be very frustrating, but I don't see how trying to quash
the sales of used books is going to help much.
As another poster on this thread pointed out, used
books were previously new books that were bought,
read (or not) and then resold. If the "used" share of a newly published book is relatively large, that probably does not mean good things for
the sale of the new copies. Amazon might make
the sale of used books "prominent" (or something),
they didn't have much to do with the fact that the used books existed in the first place.
As an overall comment, I would have to point out
that I think companies like Amazon have done far
more to help the authors of lightly purchased (and stocked) books by making them visible *at all*
than any harm they have done by offering up used
copies with the new copies.
Disclaimer: my last royalty check amounted to a
whopping $12 and change, so it's not as if I don't
share some of the disappointment or bemusement of
writers represented by the guild...
I actually don't think we're as far apart as you think, but I'm not sure that many libertarians really do take the historical record into account as much as they think they do.
Utilities backed by government monopolies enjoyed higher profits far too long because of their monopoly status. Hence, "deregulation" being all the rage.
I don't like to accuse people of selective quotation, but electrical power *access*
was one several examples I gave. And I chose
those words for a reason: it's the delivery part
of the monopoly whose justification is clearest,
and whose implementation could be most fair.
Now that there really is getting to be a very large integrated grid, electrical power *generation* should be (in theory) a decent area for competition in markets large enough for the idea to make sense. Heck, wouldn't it be fun to
send 80% of your check to the cheapest provider
or the greenest provider, at your choice?
As a side issue, I'm not sure I quite believe
the point about utility profits being too large
for too long. The question there is: compared to what? Utilities were never growth stocks, and
the whole thing basically traded like one big
pile of bonds for a long time. Now, many utilities had no incentive to be especially efficient, but I'm not sure this really resulted
in tremendous profits.
Canada, and Quebec, in particular (admitedly, not part of the U.S.)
What? We haven't conquered them yet?:-)
are notorious for building roads, at great expense, to nowhere, simply to bolster employment of road-building workers. Not quite the same as a monopoly, but such disasters do not inspire confidence in the ability to plan and regulate anything.
Seriously, this is a weak example. Your original
argument is that government can't jump start infratrucure. I point out that the US interstate
highway system might just be a teeny weeny exception to this, and you counter with the fact
that some Canadian politicians have made some
Quebec highway building projects into political
pork and/or make-work projects.
I don't see this as a compelling argument. For one, I suspect that the Canadians involved had at least some hint that the roads being built weren't really that important, but found the program vaguely palatable as workfare. For another, there
is no doubt that a ton of pork went into the US
interstate system, too (check out Interstate 99,
or even Interstate 72). Weirdly enough, though,
this didn't prevent almost all of the necessary
stuff from being built, and even some of the fairly useless roads have surprised people by suddenly making some non-places into someplaces.
Not a great argument for building them, of course,
but you can't wish away the whole network as some
mirage, I don't think.
The problem with your analysis is that it empowers the state to overcome "activation energy" at the taxpayer's expense so that a desirable steady-state is achieved sooner, rather than later.
OK, so I'd written a much larger reply to this
that got taken out by a Netscape crash. Serves
me right. All I will say here is that I really
don't see any connection between what you say
here and what I said above. I was discussing
situations where no good is provided unless something gets started, and that one way for things to get started is via a regulated monopoly.
Also, that the problems creep in when the purpose
for which the monopoly was desirable is lost and
replaced (often through technological innovation)
by some other purpose which should not enjoy such
a protection. Cable TV is a pretty natural monopoly to grant when all you're doing is giving
people the chance to see local stations with less
interference and something like HBO; then, it is just a content conveyer. The monopoly is no longer desirable when it becomes associated with the identity of *the* big-time content provider to most homes, and the supplier of most internet bandwidth besides. How do we get out of this? It's not easy, but I severely doubt that saying "let's let the market sort it all out" will work
because a monopoly is now involved. And (wait for it...) monopolies and markets don't mix well.
There are many industries where economies of scale are enormous. The PC industry is one: it costs an enormous amount of money to make the "first" new-fangled CPU. After that, they're cheap as dirt, almost literally. No government intervention was required for this industry to take off. And, while I would have liked to see cheaper
PCs sooner, it would be wrong to tax my fellows to achieve this.
OK, so I might have been a bit unclear in my original post. The obvious reasons why governmental intervention is a bad idea in the PC industry are simply:
No limited public good (like spectrum, land, right of way, right to pollute, etc.) is directly involved.
There are no "positive network effects" that
need to be encouraged here. (It doesn't matter
whether you and I use brand X or brand Y, or whether we are the only computer users on earth.)
So while you could argue cases for electrical power delivery, the telephone, the interstate
highways system, and old-tyme cable TV, you wouldn't argue for PCs.
The record on government intervention to "jump-start" infrastructure is generally poor, the odd success notwithstanding extended scrutiny of the track record.
This looks like a statement unsupported by facts.
I would claim that the government was intimately
involved with successful projects like:
universal telepohone access
universal electrical power access
clean and drinkable water
high quality highways (interstates)
research that sparked innovation in a variety of fields including darpanet
Now, this doesn't mean that whatever government touches turns to gold; involvement in agricultural
subsidies, Amtrak, military protection afforded
to oil companies, resource management in national
forests would all be poor examples. I would argue that the common thread through most of these are
that these are situations that do not involve common goods and/or positive network effects. (OK, so losing money in the timber industry is just blatant theft, but that's a different thread.)
I actually don't think we're as far apart as you think, but I'm not sure that many libertarians really do take the historical record into account as much as they think they do.
To all of you anti-government people, I say "get a clue!" The current system is not working and the free market is, by and large, not solving the problem.
The free market IS solving the problem (see the article that spawned this thread) when it is permited to do so.
The problem here stems from monopolies propped up by the government in the first place, leaving you with no legal alternatives. In fact, about the only thing that justifies government regulation to any extent (and not enough, in my book), is the existance of a government-enforced monopoly.
The most fascinating thing about some radical
libertarians is their consistency. The big, gaping piece left out of the above argument about
government regulation and monopolies is any
comment on how or why we might have some government-sanctioned monopolies in the first place. We have them because there was a time
for each monopoly industry when it was believed
that the best and cheapest way to provide some
capital-intensive service where economies of scale were required was through a private sector company (rather than, say, a municipal utility). In some
cases, the decision was probably a good one; in
others, possibly less so. In *all* cases, the stage was set for a time when changes in technology would lead to the situation where the
status quo (government-regulated monopoly) was
awkward. So, I think ATT did about as good and as
fast a job of wiring up huge pieces of the US in
its day, and provided a level of consistent service that served a genuine public good. But
the possibility of commodity long-distance pretty
much completely changed that reality, so you had the ATT break-up and now the really awkward situation situation where the Verizons and Qwests of the world have some shadowy status of quasi-monopolies via the effective use of bureaucracy.
Wiring up big metro areas for cable
was certainly expensive, costly, and not the kind
of thing that any company would jump to do unless they had a guaranteed revenue stream. This was
fine when all cable amounted to was higher quality
feeds of local channels and a few exotic notions
like HBO. Now the capacity and reach of the system has gone waaaay beyond that for which the
original monopoly made sense...but getting a system with universal access and anything like free market pricing and any kind of reasonable structure is tough.
I could go on. I think it's pretty clear that
in most cases, what has gone wrong is neither
some raw "failure of the market" or "failure of
governmental regulation" but changes in the whole technology landscape that have had completely unforseen consequences on companies, governments,
and citizens. It is going to be a mess. I think
a market can set prices when it is competitive,
but we don't really have many of those going now (except for long distance telephone, the big success story for de-regulation). I think regulated monopolies can spend the money it takes
to get infrastructure built, but they are much less useful when their reach extends to the content that is broadcast or the services that are provided.
For that matter, some of the near-future technologies that can help remove us from this
quagmire will also require some amount of regulation to do any good. I think wireless IP
will be the best thing since sliced bread within
a decade. With the appropriate set-up,
you *could* do voice, date, TV...you name it with
a much different kind of capital outlay (look ma,
no cable being laid in our street!). But we already know that unfettered 802.11b can have some, um, interesting consequences if there is no
planning or regulation of the use of the specrum
involved.
I have no idea what the best answer is, but I have
little faith that either hard-core governmental
regulation approaches or cowboy networking provide the best answer to all concerned.
If this bill passes, Microsoft would be able to pick and choose which token 'competitors' survive by deciding to whom they would and would not license the mandated technology... and it certainly wouldn't be to GNU/Linux.
That is certainly one possibility, but another
one is that Microsoft (a company that is certainly *not* run by dummies) really does know that the value of their current OS monopoly cannot possibly
grow at the rate that would continue to make them
the dominant player they surely wish to be.
Recent MS moves actually seem more focused on becoming the kind of company that gets 1% or 2%
of every transaction rather than one that gets 95% of a more limited pie. If that's the real strategy, they have no reason not to license their
stuff to anybody else. Indeed, if licensing DRM
technology substantially slows any effort in the
Linux community to work around it, re-implement it, or come up with a competing standard, they would have to be silly not to do so.
Or even give it away, if the client use was associated with a revenue stream at the server end. (Indeed, that's *classic* MS behavior.)
I personally also find a parable in the story
of MS and its manipulation of the.DOC format.
Once upon a time, you couldn't do the.DOC format
if you weren't MS, then the licensing became less
and less restrictive as the format became more and
more prevalent. Last I checked, you could do anything with it you liked except (I believe) use it as a default file format for your software. This isn't really giving anything away, though, since Word is what 98% of mankind will use to edit any.DOC files. Plus, they had to do something once it became clear that XML and stylesheets really would be a serious contender to anything
that MS put out...so guess who leads the w3.org
efforts on XSL?
I don't think anyone's saying it's wise to set up a high-tech startup in the middle of Kansas farmland... but I also think many people play down the suitability of larger midwestern cities.
Surprisingly, there *is* a decently viable tech/professional place that was Kansas farmland
not very long ago: Overland Park, Kansas. (OK,
so it is adjacent to Kansas City.:-)) But Sprint
just laid off thousands of people there, and I have to wonder what might be coming out of the
garages in the area in the next year or three. (For all I know, nothing.)
For example, look at Chicago. You've got a number of major players there (AKA. Motorola), and IBM certainly has a large presence. In a recent survey, Chicago was in the top 20 for tech-savvy cities. You can't tell me it's impossible to find tech-knowledgable people in the Chicagoland area!
No I can't tell you that, but I can tell you that
while there is *some* cost advantage to Chicago
over California, it might not be as obvious a relocation spot as you might think for a SV concern. So it was no surprise that Boeing decided to go there, but that's perhaps a different story.
There are many disadvantages to the Silicon Valley area. Questions about electrical power available should certainly be key, along with the high risk of earthquake damage, heavy taxes and govt. regulation, occasional water shortages, and an expectation of high wages so employees can afford the high cost of living in the area.
The funny thing about the electrical power shortage is that it disappeared the moment that the state signed high-price long-term contracts.
The potential for earthquake damage is worrisome,
but it would be interesting to compare the size of that risk with the bite that Chicago's winters take out of the infrastructure each and every year. California certainly has some bizarre governmental problems, but Chicago doesn't?
Anyway, I think the big key point in California's favor, and it's a huge one, is the quality of the
public and private universities there. The UC system is as elitist a public institution as you're likely to find, and that's *exactly* what
it should be. Throw in places like Stanford, and
you've got a recruiting dream boat.
As far as employment costs go, you could actually make an argument that (at least in non-bubble times) real costs are likely to be lower in California than some other places, because people like California and are willing to pay more rent and/or take less money to work there. (I won't
bore you with the details, but if you're puzzled by this, search for "hedonic regression" at google.
Now, I happen to agree with you that places like
Chicago, Pittsburgh, KC and St. Louis might be expected to develop larger tech sectors than they already have, but it hasn't yet happened that way.
Of these 4, perhaps the strangest omission is not
Chicago, but Pittsburgh: There's a big university town with many cultural offerings, a large and growing international population, and every reason to succeed except that it just doesn't quite do it.
At least in the last several years, it's because that's where you could find the best (or already experienced) technical talent. That's not to say that technically talented people don't exist in other parts of the world, but they certainly do concentrate in that area.
You are obviously right, however...:-)
The success of many manufacturing industries used to depend on their location in a very systematic
way. You had to be "close" to markets, to materials, and to an appropriate labor force. So all the steel mills ended up in places like Pittsburgh, cars were made in Detroit, textiles and shoes in New England. In today's economy,
manufacturing can be put almost anywhere, but there is still a big clustering effect in industries that need what have been (annoyingly) called "knowledge workers". And part of that
surely makes sense. For one thing:
Take, for example, technical support. Not a highly skilled position, but one requiring a certain skill level and generally some previous experience. In the Silicon Valley, you had NO problem finding qualified and experienced technicians. When my previous company moved it's technical support operations to a midwest location, recruiting became VERY difficult (even with the stellar wages, for the area anyway) - finding any quantity of technically skilled and experienced people was very difficult.
And the reason was obvious: anybody who thought
they were somebody wanted to be in {Silicon Valley, Dallas, Boston, wherever} because that's
where all of the jobs were... In the short run,
I think you are absolutely right that this is a
critical problem. In the longer run, I'm not
so sure. And the reason is, ironically enough:
So, why, you ask? You set up shop where you can find and retain people.
Now, I think the second part (retention) is likely to become a more important issue. During the internet bubble days, you could rely on a huge influx of very young talent into the SV area, since people were willing to blow off college,
regional (or even national) ties, and a lot of
other things to be involved in the Gold Rush. After the Gold Rush, though, some things can and
will revert to normal. Fortunately for the people who want to stay in Silicon Valley, it has way more attractions than many former gold rush towns.
Unfortunately for the people who expect SV to continue to be the be-all and end-all of high technology, the workers you may want and need in the coming years might be a bit different than
the starry-eyed youth who flew out to California
in the late 90s.
In particular, they are older and wiser. Many more of them are likely to have families, and
suddenly things that didn't matter before (really boring stuff like housing affordability and public schools) *will* become more important. At that
point, some locations away from the coasts suddenly seem more attractive. Yes, we'd all love to live next to an art museum at some point, but
the slightly older geek might want a back yard, too.
Now, I'm not saying that people will spontaneously move out of the Valley *just* to capture a lower
cost of living, but I'm willing to bet that a lot of people who turned up their noses at non-coastal
living could be persuaded to try it.
Disclaimer: I live in a growing college town in the Midwest, so I may be wishcasting here (I would love there to be more geeks around here). On the other hand, I know my house would have cost over
a million in San Diego (about seven times as much as what you'd pay here), and we have much better public schools. Also much less traffic. Hmm, on second thought, maybe you all should just stay where you are.:-)
As far as I can tell, at least this WOZ shit is part of the new tech bubble: wireless. Wireless is great and all, but the adoption rate of wireless usage, at least in the US, is incredibly low.
Up to a point, what you say is true. People expecting that wireless anything will suddenly
take off overnight probably burned off as many
billions as any other one single thing. Worse,
the success of wireless crucially depends on the
adoption rate, given that much of the value comes
from network effects.
That said, it's remarkably clear that wireless *is* generally going to be where the world is going. We don't know who are going to be the billionaires yet, but billionaires there will be.
For that matter, it's not necessarily true that
Silicon Valley will be where the Next Big Wireless Thing will come from, but this is a reasonable guess. So I suspect that dozens of smaller start-ups will do what start-ups do, and eventually hit on the combination of application and technology that will become what people must have.
People use cellphones, but only the super-geeks actively go beyond that to use a PDA, WAP, etc.
Absolutely. And ten years ago, cellphones could be something of a status symbol and nobody's grandma ever used email. Ten years ago, I'm not
sure you could get GPS in anything but the most
expensive or pretentious cars; five years from
now, I'd be shocked if it weren't stock equipment
on a Corolla. Yes, it takes a surprisingly long time for things to happen, and lots of bad ideas get funded in the mean time.
It's maybe a discouraging thing that past and current results do not predict future returns in the stock market, but it's a very encouraging thing that they don't help you much when it comes
to technological innovation.
Cut to the chase: human population is rapidly rising. Everything else is just a byproduct. Seriously, just look at population statistics for the root cause.
I think you are correct that, in the long run, human population levels and the rate of technological advancement are the most important things to look at. This is why I am cautiously optimistic about the
situation. Technological advancement should drastically improve energy efficiency over the
next century, but that improvement could conceivably be defeated by dramatically rising
populations.
The good news on the population front is that the
UN can now give population estimates under some
reasonable assumptions that show human populations
as leveling off substantially. (See http://esa.un.org/unpp/
for more information.) Not every fertility scenario leads to a levelling off, and some might
argue that 7 billion to 10 billion is still too many people for the planet, but I now believe that
the second derivative of population growth over
time is negative, and the first is declining as well.
This should not induce a lot of self-congratulation or allow us to favor the continuation of truly wasteful policies, but it should tell us that improvements in living standards can and most likely will lead to decreases in family size and a decline in the population growth rate.
Alas, there are still some pretty severe speed issues on things like the Preferences dialog box,
and something is definitely wrong with some of the fonts (or is this just because they aren't ani-aliased or something)?
The last Mozilla I tried on the Mac was 0.9.7, and that lasted all of 10 minutes. So far, this is much
better than that one, but I really did want to see
some MathML one of these days, too.
IAAT (I am a taxonomist). This is a major pipe dream (at least the do it in 25 years worth).
In insect taxonomy if you are a highly trained (world class) you can describe around 50 species PER YEAR (at least doing an adequate job). The (small) family I work in has over 2000 undescribed species. There are fewer than 8 experts in the world on this group, only 2-4 are actually producing names actively, and these at rate of much fewer than 50/year. This is a relatively small family of Insects, there are many many larger ones with many many more undescribed species. You do the math.
Wow; I am *so glad* you responded to this thread,
and I do also hope you respond to these questions.:-)
I would really like to do the math, but I am worried that most of the estimates tossed around
for the total number of species on the planet are, um, likely to be wrong. (There, I said it.)
So, for starters, and since you are a trained taxonomist, what do you think is the fair ballpark
number of species on the planet?
OK, so I know that question is unfair. Somewhat
more fair is: what is the current ratio of described to undescribed species in the family of
insects you specialize in? Has that ratio changed
in any predictable way? By that, I mean this:
does every new collecting expedition that goes out into a fresh area come up with just about as many
new possible species in your family as the old ones did? More? Fewer? What I'm looking for is
the ability to make a statement like "If the number of undescribed species follows the same discovery curve as [your insect family here], then
there are X species still to be discovered.
That's the kind of statement I can really deal with.
I had to tranfer an 80mb drive over appletalk . . .
then again, it beat the 100k transfer over cat net--transmission easy, but reception is tough--sometimes "Meow" and "WROWRR" are hard to distinguish . . .
This is why we've taught our cat to use Morse Code. Short "Merp" like sounds are dit while
anything that could be described as a caterwaul
is dash. Unfortunately, we found out that she
perseverates on the same seven messages:
Feed me.
I smell mice in that cupboard.
Let me tear into that cardinal, please?
Open the door you doofus.
It's freezing out here you yutz.
The box is even smellier than you are.
I left its miserable little corpse in the usual location. Dispose of it, plase.
hawk, still not correcting this stack of papers
Hah! I *knew* it. Once I saw how many slashdot
posts you'd made today, I could tell you were
trying to avoid correcting papers. Myself,
I'm putting off writing the review that was due
last week. The stack of papers isn't due until
Monday.:-)
What about getting laws that say that unsolicitated mail is illegal?
How much time do you expect a Chinese bureaucrat to spend prosecuting a fellow countryman because he made 1000 foreigners delete a bothersome message?
That depends. If sending SPAM happens to be a capital crime in China and the sender is Chinese, the odds are pretty good given the Chinese practice of harvesting organs for profit.
Now in the Internet Bubble there were a lot of good, sound businesses that were really more like Jim's and less like eBay. Duh. In fact, I think the internet is more suited for small profitable outfits -- it doesn't scale very well. With an office, some good coders, a few routers and you can reach the world.
Actually, it's even better than that. If you've
got the right business, you can ditch the coders and routers and just get an ISP to
set you up something ultra-basic. There's
a fabric store in our town that is now moving
into a higher rent district because they became part of a business called "bridalfabrics.com". As best as I can understand it, they make surprisingly good money by selling bulk fabric suitable for wedding gowns and things like bridesmaid's
dresses over the web. The reach of the 'net
meant that some dorky little fabric store in
Columbia, MO could now outsell Walmart
in a carefully selected niche market. Yow.
And then there are all the small independent
bookstores who owe their fortunes in no small degree to amazon.com. We have one
of *those* in town, too. It was an okay small
independent that is getting run out of the mall
by a Barnes and Noble. But not run out of
business, since their thousands of carefully
chosen, slightly odd-ball titles are turning over
at a much better rate given national exposure
that cost them basically only a database merge of some kind. Now they'll get a store
that's twice as big a couple of blocks away
from campus, and I suspect they will do just
fine. Basically, it looks as thought the internet
is best at selling things that end up in random
locations but can be pulled off of shelves and
shipped. Weird books on Missouri archeology? Somebody wants those. Import
CD-single from Europe? Yup, that too. Aunt
Esmerelda's Limoge china? That's a bidding
war on e-bay. Anything you can get from Target or Walmart? No, let's just drive down
to the store and pick it up, thanks.
There's no deep message here, but it's amazing that so many missed it so wildly.
Slashdot Mirror
Ah, thanks for beating me to this recommendation. :-) A cute point about this particular paper is that it actually won an Ig Nobel Prize a few years back. Now, it sure ain't a flawless piece of work, but it is a result that you ignore at your peril. For those who won't bother to click through or read the linked paper, the punch line is exactly what the title says: not only do *most* people from a given population think they're at about the 60th percentile in ability for X, for almost values of X, but they do not correct their inflated self-assessments even when confronted with data that should clue them in. So, you might think that somebody who was in the bottom 10% but who thought they were better than the average student at, say, "proper" English grammar could recognize that this might not be true if you confronted him or her with their own written work and a representative sample of student work. But they don't; if anything, they now think they are even better than they did before.
Now, I suppose the Ig Nobel was awarded to them because in some sense this is a "duh" result. But the real point is that it really does completely crush what might seem to be an obvious and humane teaching strategy: provide students with models of superior work and have them strive to meet that ideal. I hope some of you just had your blood run cold when you just realized why this won't work.
Now it gets even better once you realize that this same effect can help explain why education about science and technology is especially hard to design. A big strong argument in favor of Real Science in comparison with PseudoScience is that the Real kind eventually leads to very tangible yet nearly miraculous things. So Real Science gives rise to miraculous stuff like rewritable CD players and genetic engineering, while astrology and ESP only seem to lead to bad TV specials. Now, you think that this difference would be clear, and that you would listen to the people who brought you the Magical Machines when they point out that astrology is complete crap. But they don't.
Be afraid. Be very afraid.
Indeed. I think a lot of Ruby users don't find themselves very interested in this issue any more. For a wide sampling of opinion on this and related issues, back in the day, you might be interested in this 138-post thread
Or maybe not. :-) More seriously, the best response in this thread is, I think, the second
one, by the author of the book involved.
Given the number of object-oriented languages (including Java) that have been doing Just Fine without MI, I suspect your nagging feeling is probably just needless worry. I have no doubt that your code could have a different conceptual feel without multiple inheritance, but, hey, single (implementation) inheritance itself is probably over-used. Stepanov, god of the STL would argue (I think) that what people really want in most cases is generic programming on well-structured data. Interestingly, Ruby can be used in this way, too, despite the pretensions to being "pure OO".
Yes, people in what you could call the Ruby Community did in fact wonder what Bruce Eckel was thinking when he wrote that. This quote is particularly weird given that he goes on to point out that the syntax is more Perlish. I think what is really going on here is that Ruby did fairly self-consciously tried to implement some features of particular languages that the author thought would be especially useful. But a lot of the similarities with various scripting languages are almost certainly less "rip-off" than the inevitable consequence of working in a constrained design space. It is actually quite reminiscent of a situation you see in plagiarism lawsuits against screenwriters who are accused of "clearly copying" some part of their work from that of some amateur they may or may not have barely seen once upon a time. But it turns out that the courts recognize that once you have the nucleus of a specific idea, a lot of plot and character similarities basically come along for the ride.
The real reason Ruby got written was that the author thought it was a better language for his uses than the alternatives available to him. Obviously, you won't get universal agreement on something like this, but to assume that something like Ruby would be an ignorant rip-off of your favorite language strikes me as a bit weird.
The problem of support is, of course, a real one in many applications. The fallacy of the "who cares?" argument here is that the same argument could probably have been used against every programming language since the first assembler was written. Not every new language "makes it" by a long shot, but not every well-supported language lives forever, either. Python was barely on people's radar for the very longest time; it has succeeded because enough people were convinced it had advantages that outweighed a temporary (or possibly even permanent) disadvantage in publicity, support, or whatever reason-to-fail you might give.
Well, you can be proud of your heritage without repeating the sins of your parents. :-) As far
as I can tell (I've programmed in Perl since 1991
and in Ruby for a few months), Ruby retains much
of Perl's flexibility while hugely reducing
the amount of (non-regexp) "line noise" feel that
even I think Perl sometimes has, and dramatically reducing the number of oddball special cases. I
know you could say the same thing about Python,
but I never really "got into" Python, for whatever
reason. Like it or not, I think we all realize
that tastes vary somewhat in these things.
I would love to see this book get published, but the last I heard, as suggested by this message from the author himself, the book has been cancelled. Does anybody know if they changed their minds on this or found a new publisher (Addison Wesley did the original contract)?
First off, congratulations on holding off (longer than I did) on the grading of the papers that I predict are currently stacked on the "visitor" chair in your office. :-)
Secondly, what part of Canada would you really want to have in the US anyway? OK, so B.C. might be worth something, but anything else? Alberta would end up being just as much a PITA as Montana and Wyoming currently are, and we already have a full complement of boring flat states, so Manitoba and Saskatchewan are out. The Maritime provinces are basically Appalachia with a (pretty, I'll admit) coastline, Quebec is over-run with people who speak French, and that leaves you with Ontario. Do we have any takers for a bankrupt, boring, grid-locked province filled with former Nortel employees? Didn't think so.
So let's *not* joke about making Canada our 51st state; if we really wanted it that badly, we surely would have tried harder back in 1812. :-)
OK, sport, so what's wrong with wanting your roll-top desk to close? Seriously, this is an article thread about the death of ugly beige boxes which was brought on in part by the fact that people are using computers more in the home, but don't want to have something as hideous looking as your average PC there. Or, let's put it another way: geeks are quite well-known for their refusal to compromise on principle, and often get insults or abuse as their reward for this. Now, you may be unfamiliar with the idea or unsympathetic to the notion that most traditional computer monitors make really crummy additions to virtually every decor. But, of course, the first thing you do is insult and try to make fun of somebody who has a different idea. Forgive me if I'm just not very impressed.
To be brutally honest about it, the lack of Matlab is probably costing Apple tens of thousands of lost sales per year (really! I seriously do mean it!). If Apple is serious about selling the new Macs to technical users, they really, truly are going to have to do something about the complete lack of Matlab. Heck, it might even be worth it for them to jump into Octave development with both feet and bring *that* up to speed. R is wonderful, Mathematica is beautiful, but when you have to use Matlab, there really is no substitute. :-(
I guess I was unclear on this. My point was that one reason (among others) for so much being OOP was that neither publishers nor wholesalers could afford (with the new depreciation rules) to carry as much backstock as they used to. In print really only meant that you could order copies from the publisher, not that the smell of hot ink was still in the air. :-) The surprise then was how little it turned out to matter in terms of industry profitability.
I will admit that I don't read many novels anymore, but I *think* the most expensive one I remember seeing recently was more like $20 in paperback. In other words, waaay more than mass market, to be sure, but not $25 per. The great thing about the economics of this is that we have a good definition of "excessive price" we can use: the price where your overall profit would be higher if you lowered the the price. The trick, of course, is knowing what that price is for any given book... And the tragedy is that most books don't make any money, so you'd be just as well off selling copies for the production cost and then "updating" the price stickers for the few titles that you could essentially give away...
Hmm, I should probably patent that idea. :-)
This is an interesting list, but I'm not sure these are moves that would benefit publishers, and it is publishers who would have to make these moves.
As far as keeping books in print goes, the tricky part there is that the inventory you build up that way becomes a hot potato under the current tax and accounting regulations, as far as I can understand the situation. Back in the 1980s, I believe, rules for the depreciation of certain kinds of assets (like books piled up in warehouses) changed, and after that point, neither publishers nor sellers really wanted to warehouse a lot of stuff. Moreover, it became clearer and clearer that the chance that you would publish a "sleeper" that would suddenly more than justify your having a large initial print run two years down the line really were not very large. Your profits are greater if you can use marketing to move a more limited number of titles from "proven" authors. All the back catalog stuff is there basically to find the next possible Hot Author that you can market...for their next book. Whose rights you are likely to own, given the standard contract.
As far as the price of books goes, I'm afraid I don't see very much evidence that lower prices would increase profits much, either. The books that sell best are the books that sell well, and publishers don't have to discount these much (although booksellers might). If you've got a title that does not take off at $25 a copy (in a relative way), then you don't really expect it to magically take off at $15 or $10, and even if it did, you'd have to do a huge volume to make up for the discount. (Plus, authors royalties would certainly not go up in this scenario.)
There are two real problem here. One is that individual books are not commodities in the same way that many other goods are. If everybody is reading Grisham and you want to talk with those people, reading a discounted book by Bette R. Author won't necessarily achieve that goal. Another problem is that there is not very much risk in the business for publishers, since they have a diversified "portfolio" of offerings. Individual authors only have an interest in their specific books. So authors bear a lot of the risk, but have no say in many of the decisions that are made to buy and sell and price and market. I can see how this could be very frustrating, but I don't see how trying to quash the sales of used books is going to help much. As another poster on this thread pointed out, used books were previously new books that were bought, read (or not) and then resold. If the "used" share of a newly published book is relatively large, that probably does not mean good things for the sale of the new copies. Amazon might make the sale of used books "prominent" (or something), they didn't have much to do with the fact that the used books existed in the first place.
As an overall comment, I would have to point out that I think companies like Amazon have done far more to help the authors of lightly purchased (and stocked) books by making them visible *at all* than any harm they have done by offering up used copies with the new copies.
Disclaimer: my last royalty check amounted to a whopping $12 and change, so it's not as if I don't share some of the disappointment or bemusement of writers represented by the guild...
I don't like to accuse people of selective quotation, but electrical power *access* was one several examples I gave. And I chose those words for a reason: it's the delivery part of the monopoly whose justification is clearest, and whose implementation could be most fair. Now that there really is getting to be a very large integrated grid, electrical power *generation* should be (in theory) a decent area for competition in markets large enough for the idea to make sense. Heck, wouldn't it be fun to send 80% of your check to the cheapest provider or the greenest provider, at your choice?
As a side issue, I'm not sure I quite believe the point about utility profits being too large for too long. The question there is: compared to what? Utilities were never growth stocks, and the whole thing basically traded like one big pile of bonds for a long time. Now, many utilities had no incentive to be especially efficient, but I'm not sure this really resulted in tremendous profits.
What? We haven't conquered them yet? :-)
Seriously, this is a weak example. Your original argument is that government can't jump start infratrucure. I point out that the US interstate highway system might just be a teeny weeny exception to this, and you counter with the fact that some Canadian politicians have made some Quebec highway building projects into political pork and/or make-work projects.
I don't see this as a compelling argument. For one, I suspect that the Canadians involved had at least some hint that the roads being built weren't really that important, but found the program vaguely palatable as workfare. For another, there is no doubt that a ton of pork went into the US interstate system, too (check out Interstate 99, or even Interstate 72). Weirdly enough, though, this didn't prevent almost all of the necessary stuff from being built, and even some of the fairly useless roads have surprised people by suddenly making some non-places into someplaces. Not a great argument for building them, of course, but you can't wish away the whole network as some mirage, I don't think.
OK, so I'd written a much larger reply to this that got taken out by a Netscape crash. Serves me right. All I will say here is that I really don't see any connection between what you say here and what I said above. I was discussing situations where no good is provided unless something gets started, and that one way for things to get started is via a regulated monopoly. Also, that the problems creep in when the purpose for which the monopoly was desirable is lost and replaced (often through technological innovation) by some other purpose which should not enjoy such a protection. Cable TV is a pretty natural monopoly to grant when all you're doing is giving people the chance to see local stations with less interference and something like HBO; then, it is just a content conveyer. The monopoly is no longer desirable when it becomes associated with the identity of *the* big-time content provider to most homes, and the supplier of most internet bandwidth besides. How do we get out of this? It's not easy, but I severely doubt that saying "let's let the market sort it all out" will work because a monopoly is now involved. And (wait for it...) monopolies and markets don't mix well.
OK, so I might have been a bit unclear in my original post. The obvious reasons why governmental intervention is a bad idea in the PC industry are simply:
So while you could argue cases for electrical power delivery, the telephone, the interstate highways system, and old-tyme cable TV, you wouldn't argue for PCs.
This looks like a statement unsupported by facts. I would claim that the government was intimately involved with successful projects like:
Now, this doesn't mean that whatever government touches turns to gold; involvement in agricultural subsidies, Amtrak, military protection afforded to oil companies, resource management in national forests would all be poor examples. I would argue that the common thread through most of these are that these are situations that do not involve common goods and/or positive network effects. (OK, so losing money in the timber industry is just blatant theft, but that's a different thread.)
I actually don't think we're as far apart as you think, but I'm not sure that many libertarians really do take the historical record into account as much as they think they do.
The most fascinating thing about some radical libertarians is their consistency. The big, gaping piece left out of the above argument about government regulation and monopolies is any comment on how or why we might have some government-sanctioned monopolies in the first place. We have them because there was a time for each monopoly industry when it was believed that the best and cheapest way to provide some capital-intensive service where economies of scale were required was through a private sector company (rather than, say, a municipal utility). In some cases, the decision was probably a good one; in others, possibly less so. In *all* cases, the stage was set for a time when changes in technology would lead to the situation where the status quo (government-regulated monopoly) was awkward. So, I think ATT did about as good and as fast a job of wiring up huge pieces of the US in its day, and provided a level of consistent service that served a genuine public good. But the possibility of commodity long-distance pretty much completely changed that reality, so you had the ATT break-up and now the really awkward situation situation where the Verizons and Qwests of the world have some shadowy status of quasi-monopolies via the effective use of bureaucracy.
Wiring up big metro areas for cable was certainly expensive, costly, and not the kind of thing that any company would jump to do unless they had a guaranteed revenue stream. This was fine when all cable amounted to was higher quality feeds of local channels and a few exotic notions like HBO. Now the capacity and reach of the system has gone waaaay beyond that for which the original monopoly made sense...but getting a system with universal access and anything like free market pricing and any kind of reasonable structure is tough.
I could go on. I think it's pretty clear that in most cases, what has gone wrong is neither some raw "failure of the market" or "failure of governmental regulation" but changes in the whole technology landscape that have had completely unforseen consequences on companies, governments, and citizens. It is going to be a mess. I think a market can set prices when it is competitive, but we don't really have many of those going now (except for long distance telephone, the big success story for de-regulation). I think regulated monopolies can spend the money it takes to get infrastructure built, but they are much less useful when their reach extends to the content that is broadcast or the services that are provided.
For that matter, some of the near-future technologies that can help remove us from this quagmire will also require some amount of regulation to do any good. I think wireless IP will be the best thing since sliced bread within a decade. With the appropriate set-up, you *could* do voice, date, TV...you name it with a much different kind of capital outlay (look ma, no cable being laid in our street!). But we already know that unfettered 802.11b can have some, um, interesting consequences if there is no planning or regulation of the use of the specrum involved.
I have no idea what the best answer is, but I have little faith that either hard-core governmental regulation approaches or cowboy networking provide the best answer to all concerned.
That is certainly one possibility, but another one is that Microsoft (a company that is certainly *not* run by dummies) really does know that the value of their current OS monopoly cannot possibly grow at the rate that would continue to make them the dominant player they surely wish to be.
Recent MS moves actually seem more focused on becoming the kind of company that gets 1% or 2% of every transaction rather than one that gets 95% of a more limited pie. If that's the real strategy, they have no reason not to license their stuff to anybody else. Indeed, if licensing DRM technology substantially slows any effort in the Linux community to work around it, re-implement it, or come up with a competing standard, they would have to be silly not to do so.
Or even give it away, if the client use was associated with a revenue stream at the server end. (Indeed, that's *classic* MS behavior.) I personally also find a parable in the story of MS and its manipulation of the .DOC format.
Once upon a time, you couldn't do the .DOC format
if you weren't MS, then the licensing became less
and less restrictive as the format became more and
more prevalent. Last I checked, you could do anything with it you liked except (I believe) use it as a default file format for your software. This isn't really giving anything away, though, since Word is what 98% of mankind will use to edit any .DOC files. Plus, they had to do something once it became clear that XML and stylesheets really would be a serious contender to anything
that MS put out...so guess who leads the w3.org
efforts on XSL?
Surprisingly, there *is* a decently viable tech/professional place that was Kansas farmland not very long ago: Overland Park, Kansas. (OK, so it is adjacent to Kansas City. :-)) But Sprint
just laid off thousands of people there, and I have to wonder what might be coming out of the
garages in the area in the next year or three. (For all I know, nothing.)
No I can't tell you that, but I can tell you that while there is *some* cost advantage to Chicago over California, it might not be as obvious a relocation spot as you might think for a SV concern. So it was no surprise that Boeing decided to go there, but that's perhaps a different story.
The funny thing about the electrical power shortage is that it disappeared the moment that the state signed high-price long-term contracts. The potential for earthquake damage is worrisome, but it would be interesting to compare the size of that risk with the bite that Chicago's winters take out of the infrastructure each and every year. California certainly has some bizarre governmental problems, but Chicago doesn't?
Anyway, I think the big key point in California's favor, and it's a huge one, is the quality of the public and private universities there. The UC system is as elitist a public institution as you're likely to find, and that's *exactly* what it should be. Throw in places like Stanford, and you've got a recruiting dream boat.
As far as employment costs go, you could actually make an argument that (at least in non-bubble times) real costs are likely to be lower in California than some other places, because people like California and are willing to pay more rent and/or take less money to work there. (I won't bore you with the details, but if you're puzzled by this, search for "hedonic regression" at google.
Now, I happen to agree with you that places like Chicago, Pittsburgh, KC and St. Louis might be expected to develop larger tech sectors than they already have, but it hasn't yet happened that way. Of these 4, perhaps the strangest omission is not Chicago, but Pittsburgh: There's a big university town with many cultural offerings, a large and growing international population, and every reason to succeed except that it just doesn't quite do it.
You are obviously right, however... :-)
The success of many manufacturing industries used to depend on their location in a very systematic way. You had to be "close" to markets, to materials, and to an appropriate labor force. So all the steel mills ended up in places like Pittsburgh, cars were made in Detroit, textiles and shoes in New England. In today's economy, manufacturing can be put almost anywhere, but there is still a big clustering effect in industries that need what have been (annoyingly) called "knowledge workers". And part of that surely makes sense. For one thing:
And the reason was obvious: anybody who thought they were somebody wanted to be in {Silicon Valley, Dallas, Boston, wherever} because that's where all of the jobs were... In the short run, I think you are absolutely right that this is a critical problem. In the longer run, I'm not so sure. And the reason is, ironically enough:
Now, I think the second part (retention) is likely to become a more important issue. During the internet bubble days, you could rely on a huge influx of very young talent into the SV area, since people were willing to blow off college, regional (or even national) ties, and a lot of other things to be involved in the Gold Rush. After the Gold Rush, though, some things can and will revert to normal. Fortunately for the people who want to stay in Silicon Valley, it has way more attractions than many former gold rush towns. Unfortunately for the people who expect SV to continue to be the be-all and end-all of high technology, the workers you may want and need in the coming years might be a bit different than the starry-eyed youth who flew out to California in the late 90s.
In particular, they are older and wiser. Many more of them are likely to have families, and suddenly things that didn't matter before (really boring stuff like housing affordability and public schools) *will* become more important. At that point, some locations away from the coasts suddenly seem more attractive. Yes, we'd all love to live next to an art museum at some point, but the slightly older geek might want a back yard, too.
Now, I'm not saying that people will spontaneously move out of the Valley *just* to capture a lower cost of living, but I'm willing to bet that a lot of people who turned up their noses at non-coastal living could be persuaded to try it.
Disclaimer: I live in a growing college town in the Midwest, so I may be wishcasting here (I would love there to be more geeks around here). On the other hand, I know my house would have cost over a million in San Diego (about seven times as much as what you'd pay here), and we have much better public schools. Also much less traffic. Hmm, on second thought, maybe you all should just stay where you are. :-)
Up to a point, what you say is true. People expecting that wireless anything will suddenly take off overnight probably burned off as many billions as any other one single thing. Worse, the success of wireless crucially depends on the adoption rate, given that much of the value comes from network effects.
That said, it's remarkably clear that wireless *is* generally going to be where the world is going. We don't know who are going to be the billionaires yet, but billionaires there will be. For that matter, it's not necessarily true that Silicon Valley will be where the Next Big Wireless Thing will come from, but this is a reasonable guess. So I suspect that dozens of smaller start-ups will do what start-ups do, and eventually hit on the combination of application and technology that will become what people must have.
Absolutely. And ten years ago, cellphones could be something of a status symbol and nobody's grandma ever used email. Ten years ago, I'm not sure you could get GPS in anything but the most expensive or pretentious cars; five years from now, I'd be shocked if it weren't stock equipment on a Corolla. Yes, it takes a surprisingly long time for things to happen, and lots of bad ideas get funded in the mean time.
It's maybe a discouraging thing that past and current results do not predict future returns in the stock market, but it's a very encouraging thing that they don't help you much when it comes to technological innovation.
I think you are correct that, in the long run, human population levels and the rate of technological advancement are the most important things to look at. This is why I am cautiously optimistic about the situation. Technological advancement should drastically improve energy efficiency over the next century, but that improvement could conceivably be defeated by dramatically rising populations.
The good news on the population front is that the UN can now give population estimates under some reasonable assumptions that show human populations as leveling off substantially. (See http://esa.un.org/unpp/ for more information.) Not every fertility scenario leads to a levelling off, and some might argue that 7 billion to 10 billion is still too many people for the planet, but I now believe that the second derivative of population growth over time is negative, and the first is declining as well.
This should not induce a lot of self-congratulation or allow us to favor the continuation of truly wasteful policies, but it should tell us that improvements in living standards can and most likely will lead to decreases in family size and a decline in the population growth rate.
So tell me: what am I missing here? What do they think their users are going to do? Use the mouse?
The last Mozilla I tried on the Mac was 0.9.7, and that lasted all of 10 minutes. So far, this is much better than that one, but I really did want to see some MathML one of these days, too.
Wow; I am *so glad* you responded to this thread, and I do also hope you respond to these questions. :-)
I would really like to do the math, but I am worried that most of the estimates tossed around for the total number of species on the planet are, um, likely to be wrong. (There, I said it.)
So, for starters, and since you are a trained taxonomist, what do you think is the fair ballpark number of species on the planet?
OK, so I know that question is unfair. Somewhat more fair is: what is the current ratio of described to undescribed species in the family of insects you specialize in? Has that ratio changed in any predictable way? By that, I mean this: does every new collecting expedition that goes out into a fresh area come up with just about as many new possible species in your family as the old ones did? More? Fewer? What I'm looking for is the ability to make a statement like "If the number of undescribed species follows the same discovery curve as [your insect family here], then there are X species still to be discovered.
That's the kind of statement I can really deal with.
This is why we've taught our cat to use Morse Code. Short "Merp" like sounds are dit while anything that could be described as a caterwaul is dash. Unfortunately, we found out that she perseverates on the same seven messages:
Hah! I *knew* it. Once I saw how many slashdot posts you'd made today, I could tell you were trying to avoid correcting papers. Myself, I'm putting off writing the review that was due last week. The stack of papers isn't due until Monday. :-)
That depends. If sending SPAM happens to be a capital crime in China and the sender is Chinese, the odds are pretty good given the Chinese practice of harvesting organs for profit.
Actually, it's even better than that. If you've got the right business, you can ditch the coders and routers and just get an ISP to set you up something ultra-basic. There's a fabric store in our town that is now moving into a higher rent district because they became part of a business called "bridalfabrics.com". As best as I can understand it, they make surprisingly good money by selling bulk fabric suitable for wedding gowns and things like bridesmaid's dresses over the web. The reach of the 'net meant that some dorky little fabric store in Columbia, MO could now outsell Walmart in a carefully selected niche market. Yow.
And then there are all the small independent bookstores who owe their fortunes in no small degree to amazon.com. We have one of *those* in town, too. It was an okay small independent that is getting run out of the mall by a Barnes and Noble. But not run out of business, since their thousands of carefully chosen, slightly odd-ball titles are turning over at a much better rate given national exposure that cost them basically only a database merge of some kind. Now they'll get a store that's twice as big a couple of blocks away from campus, and I suspect they will do just fine. Basically, it looks as thought the internet is best at selling things that end up in random locations but can be pulled off of shelves and shipped. Weird books on Missouri archeology? Somebody wants those. Import CD-single from Europe? Yup, that too. Aunt Esmerelda's Limoge china? That's a bidding war on e-bay. Anything you can get from Target or Walmart? No, let's just drive down to the store and pick it up, thanks.
There's no deep message here, but it's amazing that so many missed it so wildly.