On a serious note, I greatly welcome the 4 billion transaction limit being raised. Sure, it's 126 per second over the period of a year -- but it's one step closer to being the database for a major system.
Jeez. I had no idea. If our MySQL server had this limit:
Uptime: 10277498 Threads: 7 Questions: 3678405287 Slow queries: 1584 Opens: 2299222 Flush tables:
1 Open tables: 256 Queries per second avg: 357.909
I'd be receiving a phone call in about a month from my client wondering why the site stopped working.;)
As others have pointed out, there's a difference
between "Queries" and "Transactions". Now, people
have often facetiously accused MySQL people from
not knowing what a Transaction is, but here we
have some solid proof.:-)
Seriously, the previous PostgreSQL limitation
probably did have some serious ramifications for
bigger installations, so it's a good thing it's
history now. When I bring PostgreSQL up in polite
public these days, Oracle jockeys now feel the
need to critique it more seriously than the "don't make me laugh" attitude of even a couple of years
ago. Not because PostgreSQL is that likely to
sink a real high-end RDBMS, but it will be an increasingly plausible contender for users who
don't really need or use all of the cool (but expensive) features of Oracle or DB2.
Sure, we can all laugh about the idea of people emailing half a gig of video to each other, or downloading them onto their PDA, or say "wow, how cool would having digital archives of my favourite tv programs be", but the real issue here is - how do media artists make a living when their product can be copied an infinite number of times for virtually zero cost?
Gosh, if only all the questions I get would be
so easy. If the problem here is only that the
media artists don't get paid because the means
of copying (blank CDs, DVDs, and network bandwidth) have zero cost, then the obvious answer is to raise the cost of copying and use
the funds raised to pay people. We already have a (broken) system for doing that on
the CD-R(W) front. There is no reason why realistic copying fees couldn't be added to blank
media, or to the cost of network bandwidth. (For
that matter, we could and should charge for networkk bandwidth down to the personal level so
that free-riding on bandwidth becomes less of an issue.)
That part is easy. The fair distribution of revenue is a bit trickier. Basically, artists should be paid proportionally to the volume of
traffic/copying of their works on the net. The
way to do this is to add in some kind of wartermarking or other unique signature to the
content being shipped around, and then regularly
check random samples of traffic to see what people really are watching or ripping. (If this sounds silly, I'll point out that this is how virtually any media rating service works, and how over-the-air residuals and royalties are computed.) So, in February 2002, it might turn
out that U2 had 0.01% of all media traffic on the
web, while Gene Autry had 0.000005%. Total media
and bandwidth charges might have been US$2 billion. So U2 gets $2 million, while the Gene
Autry foundation gets...$100. That would be a fair if not deserving outcome.:-)
The beauty of the system here is that people get
paid, but people also get freedom of access. It
also allows rights holders to take the moral and
artistic high ground against (what would have to be illegal and heavily punished) subversion of the
system. If you strip out the identifying information, you really are depriving an artist
of income. If you butcher a soundtrack or a video
by down-sampling or screwing around with the original, you have made a derivative and inferior
work, and the artist has every right to legal
action on those grounds.
One mild disadvantage of the system I suggest is
that it would be hard initially for sellers to set differential prices for different performances/performers. (Differential pricing has not had as high an impact as it should have so
far in many of these industries, but that's a different story.) I think the way to make that
work is for publishers to make potentially available high/medium/low quality versions of the
work to be distributed, but then make sure that the stuff they want to charge more for is (say)
not available in a low quality (and low-bit-rate) stream. Some artists might be deeply sensitive
to the quality issue and insist that their stuff
only be available at higher bit rates, while the
starving and waiting to be discovered might be
thrilled to make their work available more cheaply. I think this could really work.
The problem, as always, is that there are definitely vested interests who would not want
it to work, or are concerned about the obviously
immense changes in their business models. That
is how movie studios reacted to the home VCR,
of course. The thing to like about it is that
it decriminalizes copying, yet generates a revenue stream.
who cares about market share. The real question is, how do Apple's profit earnings compare to Microsoft and to Dell (need to compare both since Apple does OS and the box).
Thanks for pointing out that you need to compare
Apple with some appropriately weighted combination
of MS and and PC makers, to get the relevant "platform" numbers. Dell is probably the best single proxy, but even here there are some tricks. In particular, most PC makers have been getting creamed this year. The total
market has not grown much. Microsoft has done
somewhat better since they have been able to profit from the upgrade treadmill in addition to
the new sales treadmill.
But the big news here is that the PC industry is now mature. I don't see how
we can return to the go-go double digit growth
years of the 90s for the overall industry. From
here on out, it really is all about market share
and margins.
Apple's margins are pretty impressive if you compare them to other PC makers, but they look
pretty weak compared to MS. The last time I checked, they were ahead of the MS+box makers combo. Now, I understand why you'd like to focus on the real bottom line numbers; things like earnings and dividends. The problem is that we don't have dividend numbers for any of these companies, and the earnings numbers, especially for MS, are basically "whatever we say they are". MS makes
a ton of money, don't get that wrong, but what
keeps the stock high and the inevitability
factor high also, is the belief that the business will continue to grow. The brutal fact
of the matter is that the boxes part of the industry will not be growing as fast, so what happens next should be very interesting.
The cheeriest possible Apple scenario I see is
that the one segment of the market that *is*
growing the best is the one that they are probably strongest in competitively: notebook computers. If current trends continue, it would not be that surprising to see Apple's market share go up; they
hold their share of the home/SOHO desktop market,
don't lose too much of their share for corporate
desktops, but win big with notebooks across the
board. How good could that get? I don't believe
they can ever top 20% of the total platforms market, but they could end up being the dominant
player for notebooks.
OK, so there is one more fantasy Apple scenario
that should be mentioned. In a world of little
real growth in the PC market, a focus on home
users and the "digital convergence" strategy
suddenly makes them look a *lot* like a consumer
electronics firm, and, specifically, a lot like
Sony. MS makes the XBox, and wants to invade Sony
turf on other fronts as well. No, I don't think
this is very likely, but I think it is thinkable,
and that means something.
Now, has Microsoft, for example, used untoward means to maintain its monopoly? The courts have unequivocally said yes. However, bear in mind that, in general, it achieved it's OS monopoly fair and square.
This is actually not true. In the 80s and early
90s, MS "negotiated" many kinds of bundling deals
with manufacturers of computers that had the effect of requiring licensing fees to be paid for
each computer sold, not each copy of the OS shipped. This practice, which is illegal in the US, was the subject of the original DOJ anti-trust
complaint against MS that was (*stupidly*) settled
for basically nothing but a wrist slap of a consent decree right after Clinton took office.
This turned out to be unfortunate, since the original case against MS was overwhelmingly strong (not that the second one wasn't strong enough), and a meaningful resolution of that one would have
probably done something useful in at least some market sectors.
If Apple hadn't stumbled after Jobs left in the 80s, well, we might be bellyaching about the Beast from Cupertino instead of the Beast from Redmond.
Funny you should mention Apple, since they have
had their own history of anti-competitive actions (e.g., when they used to forcibly control prices
on their hardware back in the day). But the
only way that Apple could have gotten a monopoly
on PCs would have been to displace the IBM PC completely, and that was really not going to happen, at least on my planet.:-)
We have three native Polish speakers in my office. I asked one of them to translate the professor's reply. She said the gist of it is that he was upset they released his name, he didn't authorize any information release, etc.
Wow; that's what it felt like to me. I feel my
"random people who returned email queries" now
has some support from native speakers.
Now this has to be the beauty of Usenet; working
from isolated keywords and the power of google (tm), you could follow what appears to be a scam
from press announcement to debunking in a couple of hours, despite the fact that the smoking gun
was in a Polish math specialty news group and
had to be translated by a third party...
Someday, this kind of thing will save people some
real money.:-)
Since he added margin and retail margin afterwards, I think the $300 was supposed to be apple's OEM cost. But the other prices seemed kind of high for an OEM cost so I think it was just a bunch of numbers pulled out of an uncomfortable place (no, not the backseat of a volkswagon).
No, these numbers weren't completely of rectal
origin. Just the incorrect ones.:-)
Seriously, the *point* is that the margin on
the new $1800 iMac need not be very slim. To
make that point, you only need to use comfortably
padded cost numbers and show that there's still room for margin. The only tricky things are new technology, how fast R&D is amortized, and making sure you include all costs.
Now, apparently you agree with me that most of these are as cushy as a plump backside (and I don't mean cache:-)). Apple knows they're going to make a couple million of these at least, and I
budgeted something like 200 man-years for development costs, so I *think* I'm safe there. The only really tricky number is the superdrive. Retail costs for these are not low ($500 is the number somebody used), but most people don't order 500,000 of them at a crack. And, indeed, one reason why retail single unit costs are high is probably that one or more very large companies have diverted a large proportion of the total manufacturing capacity for these things to themselves...
Okay, the mysterious Dr. Wlodzimierz Holtzinski doesn't get a single hit on Google.
Well, that's because they mis-spelled his name.
Seriously, I bet they are really trying to refer
to Wlodzimierz Holsztynski, who posts to Polish
newsgroups from the address "sennajawa@yahoo.com".
His last contribution to the one Usenet thread that mentions "zeosync" and his name
uses the word "nonsens" a lot, also the phrase "nie autoryzowalem", and the sentence "Bylem ich konsultantem, moze znowu bede, a moze nie,
z nimi nie wiadom." Somebody who really knows Polish could probably have a field day with this
and other posts...
I'm getting the idea that some people on the scientific team might be better termed "random people we sent email to who actually responded once or twice".
If they don't have these in place by then--with a nise Ghz+ clock speed to go with them--then I think you're right; they'll lose a lot of profit margin from creaming their high end sales.
Hmm...Mr. Back Of Envelope isn't so sure
about that. Apple is using some pretty cheap components *very cleverly* in this package. For the $1800 unit, you're realistically looking at something like this:
60 gig HD => $100
LCD screen => $200
mobo w/ proc => $250
RAM, packaging and nicknacks => $100
Superdrive => $300
R&D amortization => $150
TOTAL COST: $1100
If we add 30% to this for Apple's margin,
then the retailer gets 20%, and we've got
our $1800 price tag. Now, it could be even
better than this for Apple, since they'll sell
a lot of these directly over the Internet or
in their new Apple retail outlets.
I think the interesting story for Apple margins
on the previous versions of the iMac was that
components like RAM and CD/RW burners only got really, really cheap after their sales of the
things had basically run their course, and they'd
had to blow out at least 3 rounds of excess inventory as they did updates over the last 3 years. Other than the superdrive, there really
isn't any new/expensive technology in these things.
So if you play a DVD of This Is Spinal Tap on it, will the Stonehenge monument come out to be the right size?
There are three possible answers here, but they are all "no".:-)
Screen reality: on my ~14" screen, I estimate the 18" tall Stonehenge from the movie comes out to be no more than 2" tall, so on a 14' screen, you're only looking at 24" (2') tall.
Psychological reality: tragically in this case, human perceptual abilities will still be able to tell the Spinal Tap Stonehenge is only 18" tall, and that those are just dwarfs dancing around it.
Physical reality: actually, the 18" Stonehenge might even be over 2 feet tall on the 14' iMac
screen, so it would *not* be according to spec!
The spec value written on the napkin in the movie was very clearly that the edifice was to be 18" tall, and 24" tall is way out of spec.
Disclaimer: the target article was either horribly written or horribly edited; it's tough to tell which. My comments aren't on the article itself
but about what the responses to the article say
about what we might really want.
I think the main problem with this article is that the authors have forgotten what the desktop metaphor represents. It represents a desktop (surprise!). On a real desktop, if you run out of space you start filling stuff away into folders.
Not true. On a real and productive desktop, everything on the desktop is currently being used.
If it wasn't being used, then it might be in a folder but that folder would be in a filing cabinet, not on your desk. Direct
manipulation is a brilliant idea, but only for
objects that need to be (or should be) manipulated
directly.
Now think about the file cabinet itself. The cabinet has (say) 4 drawers, each drawer has a couple of dozen folders, and each folder quite possibly contains everything I have on one particular project. Now check that out: only
two levels of hierarchy (drawer and folder).
If I have multiple file cabinets, I can get 3 levels of hierarchy. With the use of some amount of pendaflexiness, I can get up to 4 levels of hierarchy. But that's about the limit to physical file cabinets. Moreover, each
of these levels is distinct from the others in a
fairly obvious way, and you don't really mix up
levels of the hierarchy: you really would never
have the equivalent of a file drawer and a file
folder on your desktop at the same time. Maybe
you'd have two or three folders on your desk, or
maybe you'd be riffling through your file cabinet
(but you wouldn't put it on your desktop first).
I think it is very revealing that the physical file cabinet metaphor stops when the depth gets
to level 3 or 4. Indeed, I'd argue that people
allocate their attention such that they are "really" only dealing with at most 2 levels: containers and things contained.
Hierarchical file systems, however, are arranged with the notion that the physical model can become "virtualized" to structures of arbitrary depth.
People really don't deal with things this
way, however. Deeply nested (more than 2 levels) hierarchical menus or window/directories
really don't work very well.
You DON'T buy a second desk and constantly switch between them.
Oh, but you do. Many (or most?) people who do desk work and computer work, when given a choice, will choose to have separate physical desktops for the computer and for the paper. And some of those
people will also have a "work table" for non-paper
projects. People have and use many different physical desk-top-like surfaces. Really, they can't get enough of them, as you could tell if you've seen any recent kitchen designs. Indeed,
people really don't like to do two different physical activities in the same space; nobody really likes to eat at his or her desk, for example.
You certainly don't end up with dozens of desks.
Actually, I'd argue you do end up with dozens of
different work surfaces that are specialized for
what it is that you are doing at the moment (see above). This is expensive (and costly in time) to pull off in a single physical office; it *should*
be easier to pull off in the virtual space of a computer, even if each desktop is only 2-d.
I have over 150,000 files. How many desktops would I need?
This, it seems to me, is the real issue. In the
OS sense, I have no doubt that you have 150,000
files in the filing system. But I *know* you probably don't poke into each and every one of them at the level of direct manipulation very often. Even if each interaction lasted only one
second, that's 41 hours of poking around to make
the rounds. A more realistic 5 minute time to
play with something means it takes you almost a
year and a half to make the rounds. NOBODY does
that. Everybody realizes that the 150,000 files
on your hard drive differ drastically in the amount of time you ever spend with them, yet the
file system user interface (a tree structure) is
essentially optimized for the case where you need to get at each and every one of them equally often...assuming you can navigate the deep hierarchies involved. It's nuts.
Now, how many files on your disk do you have to
directly use? That's an interesting question.
It may well be thousands, but even here there is
some useful structure that isn't well captured
by most hierarchical file systems. So, I've got hundreds of mp3 files at home, but that's not
the way it seems to me: what I have is a scrolling
menu of choices in iTunes that I can customize
in many ways (including playlists). Now I also have hundreds of pdf files corresponding (mostly) to scientific articles. No such luck on the organization front here. What I *want* is a relational view of the content, not one big directory or pdf files strewn throughout the file system corresponding to different projects. Indeed, some papers I want to see or use in two
different places, without futzing around with stuff to get it that way. I'd be thrilled if the
lastest Nature Genetics article (say) would arrive in my system and magically appear in every project
where it was relevant. Now, this could really actually start happening, but if it does, we will want to (and need to) learn to think of hierarchical file systems as (at best) some implementation level detail that really does not need to be relevant.
How many people that buy consumer-level computers upgrade them? Very few.
That certainly used to be the case. Interestingly, however, Apple has gone out of their way to make the iMac and the iBook almost completely trivial to upgrade, at least if you're doing the most common upgrades (RAM and an Airport card). In other words, now that hard drives are
getting to the point of being "big enough" (many
fewer people are getting to the point of being squeezed for hard disk space even in the age of
mp3s) and video cards are "fast enough", if you build everything else in, you really can make all reasonable upgrades possible for a person armed only with simple instructions and a US quarter.
In essence, people are empowered but not hassled,
which could basically be Apple's new slogan.
I don't know about you, but I wouldn't want my email address to end with@cox.com, sounds a little too pornoish to me;-)
It could be worse. The first electronic contact I had with
my current employer (the University of Missouri)
came from somebody whose email address was, I believe, michelle@showme.missouri.edu, which got cut off to "michelle@showme" in the mbox view of
my email agent. Now, what's the chance that *that* would be legit? So I nearly ended up junking the
travel itinerary for my job talk, until I realized that Missouri is, after all, the showme state...:-)
I followed the link to Toshiba site. They will sell me the 5 GB little hard drive for $399 retail. Apple will sell me a complete iPod for $399.
[snip]
Guess Apple's price for the iPod isn't really a rip off.
Must...control...fist...of death.
Oh, certainly; every OEM in the world buys these
drives by going to the Toshiba website and putting
in an order for 10,000 of them @ $399 each...
Use your heads, people. I have
seen so many smug comments about how Apple is not
making any money on this produt, with the premise
being that the drive alone costs them $399.
It doesn't, of course. Now, the drive probably isn't cheap, but you can be *sure* they have got
as sweet a deal as is humanly possible on these
drives (which are now no doubt getting great press). The high price at the website probably partly reflects the fact that a good, big,
chunk of Toshiba's entire production is going to a certain hardware vendor in Cupertino, with an option for them to buy even more drives, so they
don't now have much incentive to lower prices for anybody else until they've got their production
process ramped up a bit more.
If Apple is making less than a 20% margin on the
iPod, I'd be completely stunned. Not that this
is a rip-off (many people will pay it), but let's
not get all silly here...
As reasons for still using Slackware, Spud Zeppelin writes:
It's rock-stable when it's released. I'm writing this on Slackware 8.0 now, in fact. It actually fulfills the promise of being useful both on servers and workstations with a single distro.
Historically, however, I'd like to point out that
this has not always been the case. A lot of people back in the day
switched away from Slackware to the new upstart
Red Hat when the maintainer of Slackware put out releases where it was absolutely clear that nobody had ever tested *anything* much. Really, it was
off-scale that way. This was interesting given
that Slackware had started out as rebadged, bug-fixed version of SLS, which itself had problems in that the maintainer would make big changes to things like the kernel in for-real releases that ended up breaking things.
Note that this is not a flame against either
McDonald or Volkerding; Linux might have gone
*nowhere* without their contributions. But it
is pretty much a historical fact that distributions like these have tapered off as Linux has taken off. This is not surprising: the amount of effort it takes to deal with the complexity of packaging a modern Linux system has gone beyond
what a single person can do. Corporations can
do it, as can the very impressive distributed organization that is Debian, but not any lone
hacker.
It does make you a bit nostalgic.
According to the lore, Ken Thompson used to send
little "with love" notes along with the magtapes
of what we now call Unix to remote sites. I can
tell my grandkids one day about downloading SLS
onto 40 or more 5.25" floppy disks and secretly
installing a rebel Linux partition on the PC
in my office at UCSD. It has to start that way
if it's any good. It has to change at least a
little bit if it's going to stay any good, or be
good for more than a small circle of friends.
What I think the Apple iPod will do is raise the stakes for everybody else.
Well, technically speaking, it will not raise the stakes, but it will raise the bar or the standards. Do choose your words carefully.
Actually, I really meant "raise the steaks", which is synonymous with
"raise the bar", itself a less baroque abbreviated form of the now obsolete phrase "raise the barbeque".
Hmm...so you're not buying that one? Me neither.
OK, so I'm really embarrassed. I mean, if I write that crappily then people will never believe I was once a Yale English major, now will they?
It's not a 2.5 inch notebook drive. The iPod itself is too slim and too narrow for a notebook drive. A plain FireWire drive based on this particular unique 5GB super-slim hard disk is also $399. The additional music playing features are free. The storage in iPod is totally cutting-edge.
OK, so I think this argument is way too strong.
Yes, Toshiba will sell you one of these directly
for $399...but that price probably reflects a nice profit margin for them, plus the
fact that a substantial chunk of their production
has probably already been spoken for by somebody who just got into the high-end mp3 player business. My guess is that Apple's margin on these is their favored 30% or so; I'm guessing the
drive costs them about $250, the rest of the
hardware about $40, and the development costs about $40 when spread over the 500,000 units they'd like to sell in the next year. In six
months, $220 will buy them a 10 GB drive, $30 will
get them the rest of the hardware, and the price
will be down to $350. In another six months, it's
$250 for an iPod with a 20 GB drive. So it goes.
The Nomad has a 6Gb and a 20Gb version. The 20Gb version is the same price as the 5Gb Apple iPod.
Please, please, please, learn how not to be bought by apple's marketing hoardes.
The 20 gig Nomad is the same price, but it *is*
substantially larger and *isn't* firewire, with
all of the advantages we've already heard about.
Now, would *I* pay $400 for either one? Probably
not, since marketing hordes or no marketing hordes, I don't need it that badly. What I think
the Apple iPod will do is raise the stakes for
everybody else. Nobody will be satisfied with
an mp3 player that weighs almost a pound, or that
can't be used for both songs and software. Firewire is also clearly a better interface for these things than USB, and by enough of a margin
that I suspect that this might be what drives the
next wave of Firewire peripheral growth.
Apple never provides the cheapest solution, and
often doesn't even provide the "best" solution, but it tends to produce the solution that most
resembles the cheapest and best solution within
a year or three.
For instance, how would you search for the source of the quotation "Now is the time for all good men to come to the aid of their party"? That'll get cut down to "now time good men come aid party", and while I can think of a lot of web sites that might appear as a result of such a query, about 99% of them will probably not be what I'm looking
for.
OK, so it has been pointed out that you can get
what you really want with judicious use of "+"
terms in the original search. But you know what?
There's no reason why Google couldn't do that
for you if it knew you were really searching for
a quote or an exact phrase. So I therefore predict that that Google will develop a special
"quotes" section like they already have for "groups" and "images". Or maybe just a search directive like "quote:" (cf some of the special forms used in google groups).
I mean, people whine about this a lot now, so
it something that should get fixed. And Google
does tend to fix what's broken.
While we're on the subject... Has anyone else noticed that Google now includes PDF files in its searching? It indexes the content of the files and even lets you view them as plaintext. That's the best thing since bread came sliced, IMO.
I have to say that I almost cried with joy when
the indexing of PDF became a feature. The view
of them as plaintext...needs some work. But you
know what? I have some confidence the work will
get done, because one of these corporations that
Google indexes for pay will fork over some money
to make this happen better, and then we're really
going to town.
One thing about Google that people haven't mentioned is that they really do work constantly
to improve the service. When they first took over
deja, the results were...disappointing. But they've kept at it, and now it's pretty usable.
We'll know that's really arrived when everybody
starts pointing links at pages in that archive on
the scale they used to when dejanews was at its
peak.
Yahoo has gone from the king of all search engines to a portal for sex chats, and a messaging client quickly losing its own little war.
I can't quite agree with this, since Yahoo showed
true cluefulness by adopting Google for
their own search engine purposes. And
I think you are really missing the big huge honkingly important things about Yahoo:
easy access to news, and (drumroll, please)
Yahoo Finance. Screw sex; show me the money!
Google is the king of all search engines. It is clean and pure, without the convoluted portal structure that has wrecked the others. Bow before Google, beg it to bestow upon you its collection of wisdom, and love it for being so great.
Well, yeah, that goes without saying. Yet, strangely, it bears repetition.:-)
Apple stole Xerox's OS interface metaphor and released the Lisa/Macintosh.
Total horse droppings. They paid for the right
to take a look at what Parc was doing and to hire the people who did it. They paid money, and they actually shipped products that people bought.
I mean, would you have rather had Xerox sit on
this stuff for another five years? I saw the
Xerox Star system, back in the day, and there has
never been a company that had less of a clue as to
what they really had than Xerox.
It was unquestionably a wonderful product. However, Jobs et al managed to squander a 10 year technology lead by failing to develop a good licensing strategy, pricing badly, marketing terribly, and failing to go after big business
customers until it was too late.
Wow. First off, as for "Jobs et al", you should
really be aware that it was almost all "et al";
Jobs was out of Apple in 1985, one year after
the Mac was introduced. One can argue whether earlier licensing would have been a good idea or not, but note that outside licensing would probably have slowed down technological development for the simple reason that the more backwards compatibility you have to provide, the more backwards you become.
More interestingly, it was their big joint (cross-licensing) venture to develop "Pink" with IBM that really screwed them over with respect
to advancing the software part of the platform.
The bright side of that was gaining the PowerPC
platform.
As far as pricing badly goes, there was certainly some of that; Apple had also become notorious for
completely mispredicting demand for their products.
The chasing big business question is an interesting one, since the history of the Mac
is trying to get into that market (first through
"innovative" products like Lotus Jazz, then via
the agreement with MS that gets Excel out for the
Mac before the PC...it went on and on, but it really didn't get too far. You could blame Apple,
or you could just realize that being #2 to start
was a nearly insurmountable barrier in the business world, especially when #1 is MS itself.
Despite an overwhelming technical superiority and an early window in which it might have successfully competed against wintel, Macintosh stagnated, and became an expensive 5%-of-the-market niche machine, a status which it retains to this day.
For starters, I think this just goes to show *yet again* that overwhelming technical superiority does not necessarily mean very much. I think the
early window you're talking about was very short
indeed if it was in the post-1984 era. The biggest relative gap between the two platforms probably occurred in late 1985, when the Mac
was the Mac featuring built-in networking with
Appletalk and the world's first PostScript printer. Microsoft Windows 1.0 was released in November 1985. It may have
sucked, but it looked more like a Mac, and that
was enough to guarantee the Mac permanent minority
status.
Apple has always been at war with its engineering talent. Its R&D and development elan has gradually eroded ever since its storied beginnings (Woz etc)
Oh, I suppose you could think that. But note
that Woz is out of Apple by 1985, and it was *since* then that the adoption of really good
outside ideas like SCSI, PostScript, built-in
networking, etc. happened. And then there are all of the inside Apple inventions: Quickdraw, TrueType, Firewire, ADB (followed up by USB),
the continuous reinvention of the notebook...yeah, I suppose you could say their engineering sucks.
[snip]
especially on the software end. For over a decade, stewardship of the OS and platform came down to maintenance, incremental improvements, and hardware and functionality kluges. Microsoft, IBM and Intel, meanwhile, slowly but steadily closed the UI gap.
Intel closed the UI gap?? Nevermind. Yes, MS did clone the Mac. But IBM and Apple were supposed to be co-developing the ultimate OS (Pink) for an embarassingly long time; that stupendously failed effort was what really got them behind the curve.
In other words, it wasn't a failure to try, but a
(spectacularly) failed attempt. For that, Apple
and IBM deserve some blame, to be sure. But I
don't see how this would have allowed them to seize the majority market share away from Wintel.
Meanwhile Jobs, who is by all accounts an arrogant, fairly ignorant and antagonistic manager, took off/was booted out and started NeXT, another company with a variety of similar problems.
Note that Jobs was out of Apple by 1985. Also
note that the first thing he did was buy...Pixar.
Yes, the people who did Toy Story ten years later.
That was a move that undoubtedly demonstrated his
ignorance.
NeXT was to sell a new kind of computer. It had some interesting hardware/software ideas and a completely stupid/unrealistic idea about how to
package/sell/market them. NeXT packages cost deep in the 5 figures, eliminating any chance at a mass market.
Yes, I would have to say that Mach + NextStep + Display Postscript + removable writable media TWELVE YEARS AGO would qualify as "interesting".
As for "deep in the five figures", you're full
of it. The Cube cost $10,000 for non-educational
markets in 1989, and went way down from there.
Moreover, it isn't surprising that NeXT never hit
the mass market because...it wasn't aimed at the
mass market AT ALL. Now, I think you're right
that Jobs grossly overestimated the market he
did aim at, but that sure wasn't the same market
as the Mac.
Jobs actually _discouraged_ porting of 3rd party applications to his system, saying "the NeXT developer community will do things its own way, and it will be better than anything else out there." If you can believe that.
That was arrogant as hell, to be sure, but you
don't actually argue with the fact that it could
have been...true. Where NeXT competed the best
was in vertical financial markets (rapid application development with a pretty face), and
in that (admittedly limited) environment, I have
never heard that it didn't completely rule.
[snip]
and then was weeks away from receivership when Jobs magically conned Apple into buying it from him (for a ridiculously inflated price)
Yes, and we all know how horribly Apple has been
doing since they brought Jobs back. Come on.
You might have issues with OS X (I sure do), but
*the sucker works*. Which is way more than we can
say about the decade of failure (Pink, Copland,
others) in OS development that preceded it.
The ex-NCR exec who had (fairly successfully) cleaned house at Apple after its disastrous slide in the early nineties (during which it had been so close to bankruptcy it had to take investment from Microsoft!) saw Jobs coming, was unable to prevent his return, and split.
First of all, Amelio was from National Semiconductor, and, gosh, that's an outfit that
has just whipped Intel, right? You're right that
a lot of costs got cut under his regime, but some
of the more painful things actually happened under
Jobs, including the MS investment, the canning of
the Newton, the final turfing of Hypercard, and
yet more layoffs.
[snip]
Depending on how you interpret Pixar, Jobs has a consistent track record for failure.
How *I* interpret Pixar is pretty simple: they
made Toy Story, Toy Story 2, and Bug's Life.
They make great stuff, have a market cap of close
to $2 billion, and have outperformed the
S&P 500 over the last 5 years. I guess
that defines "failure" upwards.
Regardless of what you think of Jobs, Apple's glory days were over before Regan left office.
I find that hard to believe. For starters, I believe they sold more Macs last year than they sold during the entire Reagan administration.
Look, I'm not arguing that Apple is a perfect
company, since it isn't, but if you want to
lay into them, do get your facts straight, and
do learn to think critically. Apple has been
a relatively poor stock to own over the last 15
years, and that's a serious condemnation. But
to argue that they can't do anything right...is
just silly.
Even the best search engine can only answer questions, it cannot teach how to ask new questions.
I actually disagree to some extent; the search engine alone can't teach this, but a decent human can, if he or she sees what the student is trying to do. I've actually sat down with undergraduates before when they
have problems getting anything useful out of search engines. In many cases, the real problem (as you note) is that they don't know what question they are really asking. That itself can either happen because they lack some kinds of critical thinking skills, or because they don't know what *any* of the terms they're trying to use
really mean, because they don't spend the time to
get even that far. In both cases, it's easy to point out the fatal flaw, and only a bit harder to get them to improve to some degree. And, in most
cases, you can get people to the point where they
themselves can realize what's going wrong.
I mean, it sure beats trying to learn how to ask
questions when no answers at all ever come back...
Yup, 8 millibits per second. This counts as the first implementation of IP over humpback whale song, with ones encoded as "AHOOOOOOOOHHHhhh..." and zeroes encoded as "EEEEEEeeeeEEEEEeeEEEEE..."
Not only do you get phenomenal 8mbps download speeds, but also this development brings e-commerce and pr0n to the cetacean community.
Yes, there are serious advantages to be had here (no need to lay undersea cable; your routers are
powered by krill and are usually protected by international treaties). But I'm surprised that nobody has yet mentioned the economies of scale.
OK; that was a joke; get it? Economies of scale? Whales don't *have* scales. OK, how about this:
another downside is that your devices only work
in promiscuous mode.
No dice there, either. Right, so moving on along,
the one real effect that Sprint shutting down ION
will have is the likely tanking of the alternative club scene in Overland Park, Kansas.
Now, if you live in the KC area and don't find *that* one funny, you probably just
got down-sized. Just like the whales are going
to be down-sized.
Every time a story about a company doing something irresponsible or evil gets posted on Slashdot, somebody invariably makes this argument. "Don't blame them! They're just trying to make a profit!" Apparently the idea is that anyone who's trying to make money is exempt from any moral responsibility whatsoever.
Not quite. Individuals are (or should be) always responsible for their actions. It is the diffusion of responsibility within large organizations that makes it very difficult to even discuss moral issues when discussing a corporation. So let's not even try to do that.
Instead, I think it is far better for people to have accurate notions of what corporations are all about. Really, where we get into problems is when people start believing the self-serving blurbs that companies put into the media (itself a collection of corporate entities) about how great they are or how much they care.
If people really understood that each and every
public corporation (at least) has a legal imperative to maximize return on shareholder investment, then they would treat corporations much differently, and support measures designed
to make corporations as little like individual
people as possible.
So, for example, there is the psychotic weirdness
of corporate income taxes, corporate lobbying, and
corporate contributions to political organizations. I think the only real way to deal
with this mess is to (and please read the whole thing before you flame) eliminate
all corporate income taxes. If corporations did
not pay taxes, people would see no real reason for them to engage in any of the rest of this nonsense (including so-called corporate charitable contributions, which is essentially a kind of bribery). Basically, it would make the
whole system much more transparent. As far as
contributions to political activity would go,
that would only be allowed to proceed from private
individuals, and similarly with lobbying. Once it became clear just how truly concentrated interests in changing regulations or enacting legislation really were, politicians would face a serious dilemma when accepting large contributions
from individuals who were known to benefit directly from particular courses of action. Always being able to follow the political money trail back to specific individuals would be an
incredibly powerful tool in understanding how
the system really worked.
Obviously, the rich and powerful would find themselves really hating this scenario, precisely
because it would require them to be judged for
their actions as individuals and not as some small part of a larger and (almost by definition) unresponsible and opaque organization that just
happens to serve as their proxy. Politicians would not be very happy with this either. A lot of very valuable information would be more directly communciated by looking at a pattern of acceptance or denial of campaign contributions.
People could really see what it was they were voting for or against.
I really want this. Of course, my kids want a
pony. Alas, the likelihoods involved are comparable in the two cases.
I've never heard a good explanation for why this is supposed to be true. In fact I've never heard any explanation for it; some people
just assume that the profit motive is enough to justify any misdeed, as long it stops short of breaking an actual law.
Slashdot Mirror
As others have pointed out, there's a difference between "Queries" and "Transactions". Now, people have often facetiously accused MySQL people from not knowing what a Transaction is, but here we have some solid proof. :-)
Seriously, the previous PostgreSQL limitation probably did have some serious ramifications for bigger installations, so it's a good thing it's history now. When I bring PostgreSQL up in polite public these days, Oracle jockeys now feel the need to critique it more seriously than the "don't make me laugh" attitude of even a couple of years ago. Not because PostgreSQL is that likely to sink a real high-end RDBMS, but it will be an increasingly plausible contender for users who don't really need or use all of the cool (but expensive) features of Oracle or DB2.
Gosh, if only all the questions I get would be so easy. If the problem here is only that the media artists don't get paid because the means of copying (blank CDs, DVDs, and network bandwidth) have zero cost, then the obvious answer is to raise the cost of copying and use the funds raised to pay people. We already have a (broken) system for doing that on the CD-R(W) front. There is no reason why realistic copying fees couldn't be added to blank media, or to the cost of network bandwidth. (For that matter, we could and should charge for networkk bandwidth down to the personal level so that free-riding on bandwidth becomes less of an issue.)
That part is easy. The fair distribution of revenue is a bit trickier. Basically, artists should be paid proportionally to the volume of traffic/copying of their works on the net. The way to do this is to add in some kind of wartermarking or other unique signature to the content being shipped around, and then regularly check random samples of traffic to see what people really are watching or ripping. (If this sounds silly, I'll point out that this is how virtually any media rating service works, and how over-the-air residuals and royalties are computed.) So, in February 2002, it might turn out that U2 had 0.01% of all media traffic on the web, while Gene Autry had 0.000005%. Total media and bandwidth charges might have been US$2 billion. So U2 gets $2 million, while the Gene Autry foundation gets...$100. That would be a fair if not deserving outcome. :-)
The beauty of the system here is that people get paid, but people also get freedom of access. It also allows rights holders to take the moral and artistic high ground against (what would have to be illegal and heavily punished) subversion of the system. If you strip out the identifying information, you really are depriving an artist of income. If you butcher a soundtrack or a video by down-sampling or screwing around with the original, you have made a derivative and inferior work, and the artist has every right to legal action on those grounds.
One mild disadvantage of the system I suggest is that it would be hard initially for sellers to set differential prices for different performances/performers. (Differential pricing has not had as high an impact as it should have so far in many of these industries, but that's a different story.) I think the way to make that work is for publishers to make potentially available high/medium/low quality versions of the work to be distributed, but then make sure that the stuff they want to charge more for is (say) not available in a low quality (and low-bit-rate) stream. Some artists might be deeply sensitive to the quality issue and insist that their stuff only be available at higher bit rates, while the starving and waiting to be discovered might be thrilled to make their work available more cheaply. I think this could really work.
The problem, as always, is that there are definitely vested interests who would not want it to work, or are concerned about the obviously immense changes in their business models. That is how movie studios reacted to the home VCR, of course. The thing to like about it is that it decriminalizes copying, yet generates a revenue stream.
Thanks for pointing out that you need to compare Apple with some appropriately weighted combination of MS and and PC makers, to get the relevant "platform" numbers. Dell is probably the best single proxy, but even here there are some tricks. In particular, most PC makers have been getting creamed this year. The total market has not grown much. Microsoft has done somewhat better since they have been able to profit from the upgrade treadmill in addition to the new sales treadmill.
But the big news here is that the PC industry is now mature. I don't see how we can return to the go-go double digit growth years of the 90s for the overall industry. From here on out, it really is all about market share and margins.
Apple's margins are pretty impressive if you compare them to other PC makers, but they look pretty weak compared to MS. The last time I checked, they were ahead of the MS+box makers combo. Now, I understand why you'd like to focus on the real bottom line numbers; things like earnings and dividends. The problem is that we don't have dividend numbers for any of these companies, and the earnings numbers, especially for MS, are basically "whatever we say they are". MS makes a ton of money, don't get that wrong, but what keeps the stock high and the inevitability factor high also, is the belief that the business will continue to grow. The brutal fact of the matter is that the boxes part of the industry will not be growing as fast, so what happens next should be very interesting.
The cheeriest possible Apple scenario I see is that the one segment of the market that *is* growing the best is the one that they are probably strongest in competitively: notebook computers. If current trends continue, it would not be that surprising to see Apple's market share go up; they hold their share of the home/SOHO desktop market, don't lose too much of their share for corporate desktops, but win big with notebooks across the board. How good could that get? I don't believe they can ever top 20% of the total platforms market, but they could end up being the dominant player for notebooks.
OK, so there is one more fantasy Apple scenario that should be mentioned. In a world of little real growth in the PC market, a focus on home users and the "digital convergence" strategy suddenly makes them look a *lot* like a consumer electronics firm, and, specifically, a lot like Sony. MS makes the XBox, and wants to invade Sony turf on other fronts as well. No, I don't think this is very likely, but I think it is thinkable, and that means something.
This is actually not true. In the 80s and early 90s, MS "negotiated" many kinds of bundling deals with manufacturers of computers that had the effect of requiring licensing fees to be paid for each computer sold, not each copy of the OS shipped. This practice, which is illegal in the US, was the subject of the original DOJ anti-trust complaint against MS that was (*stupidly*) settled for basically nothing but a wrist slap of a consent decree right after Clinton took office. This turned out to be unfortunate, since the original case against MS was overwhelmingly strong (not that the second one wasn't strong enough), and a meaningful resolution of that one would have probably done something useful in at least some market sectors.
Funny you should mention Apple, since they have had their own history of anti-competitive actions (e.g., when they used to forcibly control prices on their hardware back in the day). But the only way that Apple could have gotten a monopoly on PCs would have been to displace the IBM PC completely, and that was really not going to happen, at least on my planet. :-)
Wow; that's what it felt like to me. I feel my "random people who returned email queries" now has some support from native speakers.
Now this has to be the beauty of Usenet; working from isolated keywords and the power of google (tm), you could follow what appears to be a scam from press announcement to debunking in a couple of hours, despite the fact that the smoking gun was in a Polish math specialty news group and had to be translated by a third party...
Someday, this kind of thing will save people some real money. :-)
No, these numbers weren't completely of rectal origin. Just the incorrect ones.:-)
Seriously, the *point* is that the margin on the new $1800 iMac need not be very slim. To make that point, you only need to use comfortably padded cost numbers and show that there's still room for margin. The only tricky things are new technology, how fast R&D is amortized, and making sure you include all costs.
Now, apparently you agree with me that most of these are as cushy as a plump backside (and I don't mean cache :-)). Apple knows they're going to make a couple million of these at least, and I
budgeted something like 200 man-years for development costs, so I *think* I'm safe there. The only really tricky number is the superdrive. Retail costs for these are not low ($500 is the number somebody used), but most people don't order 500,000 of them at a crack. And, indeed, one reason why retail single unit costs are high is probably that one or more very large companies have diverted a large proportion of the total manufacturing capacity for these things to themselves...
Now here's the interesting part: they used to spell his name right in a previous version of their official bios section. This could just be sloppiness, of course.
Well, that's because they mis-spelled his name. Seriously, I bet they are really trying to refer to Wlodzimierz Holsztynski, who posts to Polish newsgroups from the address "sennajawa@yahoo.com". His last contribution to the one Usenet thread that mentions "zeosync" and his name uses the word "nonsens" a lot, also the phrase "nie autoryzowalem", and the sentence "Bylem ich konsultantem, moze znowu bede, a moze nie, z nimi nie wiadom." Somebody who really knows Polish could probably have a field day with this and other posts...
I'm getting the idea that some people on the scientific team might be better termed "random people we sent email to who actually responded once or twice".
Hmm...Mr. Back Of Envelope isn't so sure about that. Apple is using some pretty cheap components *very cleverly* in this package. For the $1800 unit, you're realistically looking at something like this:
If we add 30% to this for Apple's margin, then the retailer gets 20%, and we've got our $1800 price tag. Now, it could be even better than this for Apple, since they'll sell a lot of these directly over the Internet or in their new Apple retail outlets.
I think the interesting story for Apple margins on the previous versions of the iMac was that components like RAM and CD/RW burners only got really, really cheap after their sales of the things had basically run their course, and they'd had to blow out at least 3 rounds of excess inventory as they did updates over the last 3 years. Other than the superdrive, there really isn't any new/expensive technology in these things.
There are three possible answers here, but they are all "no". :-)
Ask a silly question, get a silly answer...
Not true. On a real and productive desktop, everything on the desktop is currently being used. If it wasn't being used, then it might be in a folder but that folder would be in a filing cabinet, not on your desk. Direct manipulation is a brilliant idea, but only for objects that need to be (or should be) manipulated directly.
Now think about the file cabinet itself. The cabinet has (say) 4 drawers, each drawer has a couple of dozen folders, and each folder quite possibly contains everything I have on one particular project. Now check that out: only two levels of hierarchy (drawer and folder). If I have multiple file cabinets, I can get 3 levels of hierarchy. With the use of some amount of pendaflexiness, I can get up to 4 levels of hierarchy. But that's about the limit to physical file cabinets. Moreover, each of these levels is distinct from the others in a fairly obvious way, and you don't really mix up levels of the hierarchy: you really would never have the equivalent of a file drawer and a file folder on your desktop at the same time. Maybe you'd have two or three folders on your desk, or maybe you'd be riffling through your file cabinet (but you wouldn't put it on your desktop first).
I think it is very revealing that the physical file cabinet metaphor stops when the depth gets to level 3 or 4. Indeed, I'd argue that people allocate their attention such that they are "really" only dealing with at most 2 levels: containers and things contained. Hierarchical file systems, however, are arranged with the notion that the physical model can become "virtualized" to structures of arbitrary depth. People really don't deal with things this way, however. Deeply nested (more than 2 levels) hierarchical menus or window/directories really don't work very well.
Oh, but you do. Many (or most?) people who do desk work and computer work, when given a choice, will choose to have separate physical desktops for the computer and for the paper. And some of those people will also have a "work table" for non-paper projects. People have and use many different physical desk-top-like surfaces. Really, they can't get enough of them, as you could tell if you've seen any recent kitchen designs. Indeed, people really don't like to do two different physical activities in the same space; nobody really likes to eat at his or her desk, for example.
Actually, I'd argue you do end up with dozens of different work surfaces that are specialized for what it is that you are doing at the moment (see above). This is expensive (and costly in time) to pull off in a single physical office; it *should* be easier to pull off in the virtual space of a computer, even if each desktop is only 2-d.
This, it seems to me, is the real issue. In the OS sense, I have no doubt that you have 150,000 files in the filing system. But I *know* you probably don't poke into each and every one of them at the level of direct manipulation very often. Even if each interaction lasted only one second, that's 41 hours of poking around to make the rounds. A more realistic 5 minute time to play with something means it takes you almost a year and a half to make the rounds. NOBODY does that. Everybody realizes that the 150,000 files on your hard drive differ drastically in the amount of time you ever spend with them, yet the file system user interface (a tree structure) is essentially optimized for the case where you need to get at each and every one of them equally often...assuming you can navigate the deep hierarchies involved. It's nuts.
Now, how many files on your disk do you have to directly use? That's an interesting question. It may well be thousands, but even here there is some useful structure that isn't well captured by most hierarchical file systems. So, I've got hundreds of mp3 files at home, but that's not the way it seems to me: what I have is a scrolling menu of choices in iTunes that I can customize in many ways (including playlists). Now I also have hundreds of pdf files corresponding (mostly) to scientific articles. No such luck on the organization front here. What I *want* is a relational view of the content, not one big directory or pdf files strewn throughout the file system corresponding to different projects. Indeed, some papers I want to see or use in two different places, without futzing around with stuff to get it that way. I'd be thrilled if the lastest Nature Genetics article (say) would arrive in my system and magically appear in every project where it was relevant. Now, this could really actually start happening, but if it does, we will want to (and need to) learn to think of hierarchical file systems as (at best) some implementation level detail that really does not need to be relevant.
Needless to say, we're not there yet.
That certainly used to be the case. Interestingly, however, Apple has gone out of their way to make the iMac and the iBook almost completely trivial to upgrade, at least if you're doing the most common upgrades (RAM and an Airport card). In other words, now that hard drives are getting to the point of being "big enough" (many fewer people are getting to the point of being squeezed for hard disk space even in the age of mp3s) and video cards are "fast enough", if you build everything else in, you really can make all reasonable upgrades possible for a person armed only with simple instructions and a US quarter.
In essence, people are empowered but not hassled, which could basically be Apple's new slogan.
It could be worse. The first electronic contact I had with my current employer (the University of Missouri) came from somebody whose email address was, I believe, michelle@showme.missouri.edu, which got cut off to "michelle@showme" in the mbox view of my email agent. Now, what's the chance that *that* would be legit? So I nearly ended up junking the travel itinerary for my job talk, until I realized that Missouri is, after all, the showme state... :-)
Must...control...fist...of death.
Oh, certainly; every OEM in the world buys these drives by going to the Toshiba website and putting in an order for 10,000 of them @ $399 each...
Use your heads, people. I have seen so many smug comments about how Apple is not making any money on this produt, with the premise being that the drive alone costs them $399.
It doesn't, of course. Now, the drive probably isn't cheap, but you can be *sure* they have got as sweet a deal as is humanly possible on these drives (which are now no doubt getting great press). The high price at the website probably partly reflects the fact that a good, big, chunk of Toshiba's entire production is going to a certain hardware vendor in Cupertino, with an option for them to buy even more drives, so they don't now have much incentive to lower prices for anybody else until they've got their production process ramped up a bit more.
If Apple is making less than a 20% margin on the iPod, I'd be completely stunned. Not that this is a rip-off (many people will pay it), but let's not get all silly here...
As reasons for still using Slackware, Spud Zeppelin writes:
Historically, however, I'd like to point out that this has not always been the case. A lot of people back in the day switched away from Slackware to the new upstart Red Hat when the maintainer of Slackware put out releases where it was absolutely clear that nobody had ever tested *anything* much. Really, it was off-scale that way. This was interesting given that Slackware had started out as rebadged, bug-fixed version of SLS, which itself had problems in that the maintainer would make big changes to things like the kernel in for-real releases that ended up breaking things.
Note that this is not a flame against either McDonald or Volkerding; Linux might have gone *nowhere* without their contributions. But it is pretty much a historical fact that distributions like these have tapered off as Linux has taken off. This is not surprising: the amount of effort it takes to deal with the complexity of packaging a modern Linux system has gone beyond what a single person can do. Corporations can do it, as can the very impressive distributed organization that is Debian, but not any lone hacker.
It does make you a bit nostalgic. According to the lore, Ken Thompson used to send little "with love" notes along with the magtapes of what we now call Unix to remote sites. I can tell my grandkids one day about downloading SLS onto 40 or more 5.25" floppy disks and secretly installing a rebel Linux partition on the PC in my office at UCSD. It has to start that way if it's any good. It has to change at least a little bit if it's going to stay any good, or be good for more than a small circle of friends.
Actually, I really meant "raise the steaks", which is synonymous with "raise the bar", itself a less baroque abbreviated form of the now obsolete phrase "raise the barbeque".
Hmm...so you're not buying that one? Me neither.
OK, so I'm really embarrassed. I mean, if I write that crappily then people will never believe I was once a Yale English major, now will they?
OK, so I think this argument is way too strong. Yes, Toshiba will sell you one of these directly for $399...but that price probably reflects a nice profit margin for them, plus the fact that a substantial chunk of their production has probably already been spoken for by somebody who just got into the high-end mp3 player business. My guess is that Apple's margin on these is their favored 30% or so; I'm guessing the drive costs them about $250, the rest of the hardware about $40, and the development costs about $40 when spread over the 500,000 units they'd like to sell in the next year. In six months, $220 will buy them a 10 GB drive, $30 will get them the rest of the hardware, and the price will be down to $350. In another six months, it's $250 for an iPod with a 20 GB drive. So it goes.
The 20 gig Nomad is the same price, but it *is* substantially larger and *isn't* firewire, with all of the advantages we've already heard about.
Now, would *I* pay $400 for either one? Probably not, since marketing hordes or no marketing hordes, I don't need it that badly. What I think the Apple iPod will do is raise the stakes for everybody else. Nobody will be satisfied with an mp3 player that weighs almost a pound, or that can't be used for both songs and software. Firewire is also clearly a better interface for these things than USB, and by enough of a margin that I suspect that this might be what drives the next wave of Firewire peripheral growth.
Apple never provides the cheapest solution, and often doesn't even provide the "best" solution, but it tends to produce the solution that most resembles the cheapest and best solution within a year or three.
OK, so it has been pointed out that you can get what you really want with judicious use of "+" terms in the original search. But you know what? There's no reason why Google couldn't do that for you if it knew you were really searching for a quote or an exact phrase. So I therefore predict that that Google will develop a special "quotes" section like they already have for "groups" and "images". Or maybe just a search directive like "quote:" (cf some of the special forms used in google groups).
I mean, people whine about this a lot now, so it something that should get fixed. And Google does tend to fix what's broken.
I have to say that I almost cried with joy when the indexing of PDF became a feature. The view of them as plaintext...needs some work. But you know what? I have some confidence the work will get done, because one of these corporations that Google indexes for pay will fork over some money to make this happen better, and then we're really going to town.
One thing about Google that people haven't mentioned is that they really do work constantly to improve the service. When they first took over deja, the results were...disappointing. But they've kept at it, and now it's pretty usable. We'll know that's really arrived when everybody starts pointing links at pages in that archive on the scale they used to when dejanews was at its peak.
I can't quite agree with this, since Yahoo showed true cluefulness by adopting Google for their own search engine purposes. And I think you are really missing the big huge honkingly important things about Yahoo: easy access to news, and (drumroll, please) Yahoo Finance. Screw sex; show me the money!
Well, yeah, that goes without saying. Yet, strangely, it bears repetition. :-)
OK, but let's not be too revisionist, okay?
Total horse droppings. They paid for the right to take a look at what Parc was doing and to hire the people who did it. They paid money, and they actually shipped products that people bought. I mean, would you have rather had Xerox sit on this stuff for another five years? I saw the Xerox Star system, back in the day, and there has never been a company that had less of a clue as to what they really had than Xerox.
Wow. First off, as for "Jobs et al", you should really be aware that it was almost all "et al"; Jobs was out of Apple in 1985, one year after the Mac was introduced. One can argue whether earlier licensing would have been a good idea or not, but note that outside licensing would probably have slowed down technological development for the simple reason that the more backwards compatibility you have to provide, the more backwards you become.
More interestingly, it was their big joint (cross-licensing) venture to develop "Pink" with IBM that really screwed them over with respect to advancing the software part of the platform. The bright side of that was gaining the PowerPC platform.
As far as pricing badly goes, there was certainly some of that; Apple had also become notorious for completely mispredicting demand for their products.
The chasing big business question is an interesting one, since the history of the Mac is trying to get into that market (first through "innovative" products like Lotus Jazz, then via the agreement with MS that gets Excel out for the Mac before the PC...it went on and on, but it really didn't get too far. You could blame Apple, or you could just realize that being #2 to start was a nearly insurmountable barrier in the business world, especially when #1 is MS itself.
For starters, I think this just goes to show *yet again* that overwhelming technical superiority does not necessarily mean very much. I think the early window you're talking about was very short indeed if it was in the post-1984 era. The biggest relative gap between the two platforms probably occurred in late 1985, when the Mac was the Mac featuring built-in networking with Appletalk and the world's first PostScript printer. Microsoft Windows 1.0 was released in November 1985. It may have sucked, but it looked more like a Mac, and that was enough to guarantee the Mac permanent minority status.
Oh, I suppose you could think that. But note that Woz is out of Apple by 1985, and it was *since* then that the adoption of really good outside ideas like SCSI, PostScript, built-in networking, etc. happened. And then there are all of the inside Apple inventions: Quickdraw, TrueType, Firewire, ADB (followed up by USB), the continuous reinvention of the notebook...yeah, I suppose you could say their engineering sucks.
Intel closed the UI gap?? Nevermind. Yes, MS did clone the Mac. But IBM and Apple were supposed to be co-developing the ultimate OS (Pink) for an embarassingly long time; that stupendously failed effort was what really got them behind the curve. In other words, it wasn't a failure to try, but a (spectacularly) failed attempt. For that, Apple and IBM deserve some blame, to be sure. But I don't see how this would have allowed them to seize the majority market share away from Wintel.
Note that Jobs was out of Apple by 1985. Also note that the first thing he did was buy...Pixar. Yes, the people who did Toy Story ten years later. That was a move that undoubtedly demonstrated his ignorance.
Yes, I would have to say that Mach + NextStep + Display Postscript + removable writable media TWELVE YEARS AGO would qualify as "interesting". As for "deep in the five figures", you're full of it. The Cube cost $10,000 for non-educational markets in 1989, and went way down from there. Moreover, it isn't surprising that NeXT never hit the mass market because...it wasn't aimed at the mass market AT ALL. Now, I think you're right that Jobs grossly overestimated the market he did aim at, but that sure wasn't the same market as the Mac.
That was arrogant as hell, to be sure, but you don't actually argue with the fact that it could have been...true. Where NeXT competed the best was in vertical financial markets (rapid application development with a pretty face), and in that (admittedly limited) environment, I have never heard that it didn't completely rule.
Yes, and we all know how horribly Apple has been doing since they brought Jobs back. Come on. You might have issues with OS X (I sure do), but *the sucker works*. Which is way more than we can say about the decade of failure (Pink, Copland, others) in OS development that preceded it.
First of all, Amelio was from National Semiconductor, and, gosh, that's an outfit that has just whipped Intel, right? You're right that a lot of costs got cut under his regime, but some of the more painful things actually happened under Jobs, including the MS investment, the canning of the Newton, the final turfing of Hypercard, and yet more layoffs.
How *I* interpret Pixar is pretty simple: they made Toy Story, Toy Story 2, and Bug's Life. They make great stuff, have a market cap of close to $2 billion, and have outperformed the S&P 500 over the last 5 years. I guess that defines "failure" upwards.
I find that hard to believe. For starters, I believe they sold more Macs last year than they sold during the entire Reagan administration.
Look, I'm not arguing that Apple is a perfect company, since it isn't, but if you want to lay into them, do get your facts straight, and do learn to think critically. Apple has been a relatively poor stock to own over the last 15 years, and that's a serious condemnation. But to argue that they can't do anything right...is just silly.
I actually disagree to some extent; the search engine alone can't teach this, but a decent human can, if he or she sees what the student is trying to do. I've actually sat down with undergraduates before when they have problems getting anything useful out of search engines. In many cases, the real problem (as you note) is that they don't know what question they are really asking. That itself can either happen because they lack some kinds of critical thinking skills, or because they don't know what *any* of the terms they're trying to use really mean, because they don't spend the time to get even that far. In both cases, it's easy to point out the fatal flaw, and only a bit harder to get them to improve to some degree. And, in most cases, you can get people to the point where they themselves can realize what's going wrong.
I mean, it sure beats trying to learn how to ask questions when no answers at all ever come back...
Yes, there are serious advantages to be had here (no need to lay undersea cable; your routers are powered by krill and are usually protected by international treaties). But I'm surprised that nobody has yet mentioned the economies of scale.
OK; that was a joke; get it? Economies of scale? Whales don't *have* scales. OK, how about this: another downside is that your devices only work in promiscuous mode.
No dice there, either. Right, so moving on along, the one real effect that Sprint shutting down ION will have is the likely tanking of the alternative club scene in Overland Park, Kansas.
Now, if you live in the KC area and don't find *that* one funny, you probably just got down-sized. Just like the whales are going to be down-sized.
I guess it's just not working for me today...
Not quite. Individuals are (or should be) always responsible for their actions. It is the diffusion of responsibility within large organizations that makes it very difficult to even discuss moral issues when discussing a corporation. So let's not even try to do that. Instead, I think it is far better for people to have accurate notions of what corporations are all about. Really, where we get into problems is when people start believing the self-serving blurbs that companies put into the media (itself a collection of corporate entities) about how great they are or how much they care.
If people really understood that each and every public corporation (at least) has a legal imperative to maximize return on shareholder investment, then they would treat corporations much differently, and support measures designed to make corporations as little like individual people as possible.
So, for example, there is the psychotic weirdness of corporate income taxes, corporate lobbying, and corporate contributions to political organizations. I think the only real way to deal with this mess is to (and please read the whole thing before you flame) eliminate all corporate income taxes. If corporations did not pay taxes, people would see no real reason for them to engage in any of the rest of this nonsense (including so-called corporate charitable contributions, which is essentially a kind of bribery). Basically, it would make the whole system much more transparent. As far as contributions to political activity would go, that would only be allowed to proceed from private individuals, and similarly with lobbying. Once it became clear just how truly concentrated interests in changing regulations or enacting legislation really were, politicians would face a serious dilemma when accepting large contributions from individuals who were known to benefit directly from particular courses of action. Always being able to follow the political money trail back to specific individuals would be an incredibly powerful tool in understanding how the system really worked.
Obviously, the rich and powerful would find themselves really hating this scenario, precisely because it would require them to be judged for their actions as individuals and not as some small part of a larger and (almost by definition) unresponsible and opaque organization that just happens to serve as their proxy. Politicians would not be very happy with this either. A lot of very valuable information would be more directly communciated by looking at a pattern of acceptance or denial of campaign contributions. People could really see what it was they were voting for or against.
I really want this. Of course, my kids want a pony. Alas, the likelihoods involved are comparable in the two cases. I've never heard a good explanation for why this is supposed to be true. In fact I've never heard any explanation for it; some people just assume that the profit motive is enough to justify any misdeed, as long it stops short of breaking an actual law.