For years, my word processor of choice was WordStar, which fit comfortably on a 160K floppy with enough space left over for a semester's worth of papers and notes. (In fact, for a while I kept my copy of dBase II, WordStar, and Turbo Pascal on the same 360K floppy. It was years before I had the cash for a hard disk...). I also used a version of MS Word that ran off a single 720K floppy -- its interface would be immediately recognizable to any use of Office today. The only thing most people would notice is the absence of Clippy and a dearth of fonts.
The tools today have more features, but they're also 100 times larger (if not 1000) and run on systems with 1000 times more processor power, memory, and disk. Are you 100 times more productive? Or even twice as productive? Unless you're doing something that you simply couldn't do without a feature that didn't exist twenty years ago, I'll bet the answer is no.
However, it has not escaped from this phenomenon. Ten years ago,
QNX had a downloadable demo image that would fit on a floppy and
included a networking stack, windowing system, web browser, and
some other nice stuff. How big is the distro today, now that it
bundles things like Java and WebSphere?
L4 is a microkernel's microkernel, and another wonderful piece
of engineering. But it's small because it's
defined to be small. It's not a system, it's a component.
(strlen hasn't gotten any larger over the years, but libc sure has...)
It's the modules on top of the microkernel that have gotten larger.
True, the fact that windows is bloated does not imply that all other
operating systems are bloated. But the fact remains that they are.
I've seen OSs and apps like word processors and databases grow from things
that a handful of people could put together in a few months
into things that require 1,000s of engineers years to create millions of lines
of code, and each new
feature or bug fix seems to require an exponential number of new engineers to add.
Nobody
can comprehend the whole system any more, except at a very high level. Eventually
some sort of event horizon is passed and it's impossible to add anything new because
every new engineer gets sucked into fixing bugs...
The isn't a new phenomenon (remember "The Mythical Man Month"?) but the change
is that it seems to have become ubiquitous -- more and more software projects
are growing past the manageable size. Hopefully there's another Fred Brooks out
there, who will tell us how to deal with all this...
I have a theory; call it "Pedantic Bore's Law": The number of lines of code in a
typical release doubles every two years.
But you're right, the sexist spin of the article is both disgusting and archaic -- but easily explained. "Nth space tourist" doesn't get a headline; "First female space tourist" does. Given that these companies are trying to promote space tourism, they are likely to issue press releases with whatever hook will help get them published. "First space tourist with green eyes", etc. might not play in Peoria, but I bet we'll see "First grandma space tourist", "first cancer survivor space tourist", etc sooner or later.
The Information Dispersal Algorithm is due to Michael Rabin. Shamir's secret-sharing algorithm uses a similar idea (it's essentially the same as Rabin's algorithm, except that the data is padded with random gibberish).
Small beans. The difference works out to about 300Km/h. Compare that with
typical orbital velocities of 10-20,000Km/h for low orbit
(of course, it depends on altitude).
Please remind me why this is exciting...
Slackware is cool to run. It's not particularly
interesting to look at. (the OSX folks have that wrapped up)
Just tell me when it's ready to download.
I'm all for garbage collected languages -- they save me a lot of work -- but
it's essential to remember that the algorithms they use to detect garbage use
strict algorithms and cannot read the mind of the programmer!
As long as you've got a live reference to something, it won't go away. If you
don't clean up your references, the referants are going to clutter up your
world. It's nice that one foo = NULL can replace free'ing everything
foo points to, item by item, and then foo itself... but
it doesn't remove the need for the programmer to make sure that no dangling
pointers to foo don't leak out.
So, you'd strip search Mother Theresa and Gandhi and let Dan Cooper walk right onto a plane.
Sorry, you're not making sense. There are lunatics of every race. If you don't think so,
please consider who has actually used
of weapons of mass distruction.
I'd try to trick them in to rewriting some crucial piece of the
security infrastructure at the last possible minute. That way,
I'd never run out of new holes to fine.
Perhaps I'd do this by smiling and saying that the OS was so
secure that I couldn't find anything wrong with it
and recommending, no, begging that they ship it
in exactly its current form.
I can't help but feel that there's something morally wrong
with spending that much money on something that lasts
a few minutes, can't be shared, etc., while back in the
real world people are
starving, school budgets are shrinking...
Like XML, the notation is just a beginning. It's nice if
everyone agrees to use the same syntax to express information
(even if it's somewhat gnarly, like XML) but that just saves
everyone the effort of writing a bunch of boilerplate code.
As someone who has been using IDLs and markup languages for
decades, XML and/or RDF doesn't excite me much. It's those
other problems -- the ones beyond their scope --
that remain unaddressed.
Writing the URIs is where all the pain is. I don't see any
difference here this and hashing out any other protocol spec.
Some SemWeb advocates tell me that they've almost got this
solved and pretty soon we won't have to do this any more --
they'll all be written automagically. This is the claim --
that this new notation, if we all used it correctly,
will suddenly make the problem easy --
that smells like bullshit to me.
This is a little bit like saying "Computer science is easy. It's all just one's and zero's."
The representation isn't the problem. The problem is agreeing what the
the relationships mean. What does "#friend" mean? Does it mean the same thing
to program X as it does to program Y? How can you tell? What do you do when there's
a conflict -- who gets to decide what #friend means, and whether this is a global or
local definition? These are questions that I've never heard answered in any
believable manner.
I'd be a lot happier (and so would my co-workers)
standing while working if I wasn't in a cubicle,
looming over my neighbors...
I do a lot of work standing up, but I wish my cube
walls were a bit higher (i.e., like all the way to
the ceiling...).
That's assuming that mac users live 20,000 years. Otherwise, Apple will
have to pick up some new users -- they don't necessarily have to switch
from windows, but they do need to be born.
OMG! That's 0.0004% of their installed user base! In a single week!
Nerds are a small demographic, but they can also be the canary in the coal mine with stuff like this. Or not. Jeepers. Someone out to FUD Apple this week, or something?
Very unlikely, with these numbers. Unless you mean perhaps "a distant fourth."
IBM had server sales of more than five billion dollars last year (or three billion, if you don't count mainframes).
Even lowly Sun beats out Dell,
which comes in at almost $1B.
Keep in mind that this is just for one year. Pick your favorite guess for how large
Googles server farm is and divide by the average age of those machines. Do you still
think they're assembling more than a billion dollars of hardware per year?
This "tube" metaphor doesn't seem bad at all, especially
given his audience. As the parent post pointed out,
if he'd used "pipes" instead of "tubes" it wouldn't be
a slashdot story...
Seriously -- do you expect him to hand out copies of a few
dozen RFCs and a map of the backbone sites and say
"here, read this, and everything will be
crystal clear." Politicians have better things to do than
try to understand BGP.
Slashdot Mirror
For years, my word processor of choice was WordStar, which fit comfortably on a 160K floppy with enough space left over for a semester's worth of papers and notes. (In fact, for a while I kept my copy of dBase II, WordStar, and Turbo Pascal on the same 360K floppy. It was years before I had the cash for a hard disk...). I also used a version of MS Word that ran off a single 720K floppy -- its interface would be immediately recognizable to any use of Office today. The only thing most people would notice is the absence of Clippy and a dearth of fonts.
The tools today have more features, but they're also 100 times larger (if not 1000) and run on systems with 1000 times more processor power, memory, and disk. Are you 100 times more productive? Or even twice as productive? Unless you're doing something that you simply couldn't do without a feature that didn't exist twenty years ago, I'll bet the answer is no.
However, it has not escaped from this phenomenon. Ten years ago, QNX had a downloadable demo image that would fit on a floppy and included a networking stack, windowing system, web browser, and some other nice stuff. How big is the distro today, now that it bundles things like Java and WebSphere?
L4 is a microkernel's microkernel, and another wonderful piece of engineering. But it's small because it's defined to be small. It's not a system, it's a component. (strlen hasn't gotten any larger over the years, but libc sure has...) It's the modules on top of the microkernel that have gotten larger.
I've seen OSs and apps like word processors and databases grow from things that a handful of people could put together in a few months into things that require 1,000s of engineers years to create millions of lines of code, and each new feature or bug fix seems to require an exponential number of new engineers to add. Nobody can comprehend the whole system any more, except at a very high level. Eventually some sort of event horizon is passed and it's impossible to add anything new because every new engineer gets sucked into fixing bugs ...
The isn't a new phenomenon (remember "The Mythical Man Month"?) but the change is that it seems to have become ubiquitous -- more and more software projects are growing past the manageable size. Hopefully there's another Fred Brooks out there, who will tell us how to deal with all this...
I have a theory; call it "Pedantic Bore's Law": The number of lines of code in a typical release doubles every two years.
But you're right, the sexist spin of the article is both disgusting and archaic -- but easily explained. "Nth space tourist" doesn't get a headline; "First female space tourist" does. Given that these companies are trying to promote space tourism, they are likely to issue press releases with whatever hook will help get them published. "First space tourist with green eyes", etc. might not play in Peoria, but I bet we'll see "First grandma space tourist", "first cancer survivor space tourist", etc sooner or later.
The Information Dispersal Algorithm is due to Michael Rabin.
Shamir's secret-sharing algorithm uses a similar idea (it's
essentially the same as Rabin's algorithm, except that the
data is padded with random gibberish).
Every little bit helps, but not that much.
Or is there some deeper hidden difference here?
Please remind me why this is exciting... Slackware is cool to run. It's not particularly interesting to look at. (the OSX folks have that wrapped up) Just tell me when it's ready to download.
As long as you've got a live reference to something, it won't go away. If you don't clean up your references, the referants are going to clutter up your world. It's nice that one foo = NULL can replace free'ing everything foo points to, item by item, and then foo itself... but it doesn't remove the need for the programmer to make sure that no dangling pointers to foo don't leak out.
So, you'd strip search Mother Theresa and Gandhi and let Dan Cooper walk right onto a plane. Sorry, you're not making sense. There are lunatics of every race. If you don't think so, please consider who has actually used of weapons of mass distruction.
Perhaps I'd do this by smiling and saying that the OS was so secure that I couldn't find anything wrong with it and recommending, no, begging that they ship it in exactly its current form.
People said the same thing about Microsoft.
Like XML, the notation is just a beginning. It's nice if everyone agrees to use the same syntax to express information (even if it's somewhat gnarly, like XML) but that just saves everyone the effort of writing a bunch of boilerplate code. As someone who has been using IDLs and markup languages for decades, XML and/or RDF doesn't excite me much. It's those other problems -- the ones beyond their scope -- that remain unaddressed.
Writing the URIs is where all the pain is. I don't see any difference here this and hashing out any other protocol spec. Some SemWeb advocates tell me that they've almost got this solved and pretty soon we won't have to do this any more -- they'll all be written automagically. This is the claim -- that this new notation, if we all used it correctly, will suddenly make the problem easy -- that smells like bullshit to me.
The representation isn't the problem. The problem is agreeing what the the relationships mean. What does "#friend" mean? Does it mean the same thing to program X as it does to program Y? How can you tell? What do you do when there's a conflict -- who gets to decide what #friend means, and whether this is a global or local definition? These are questions that I've never heard answered in any believable manner.
It's always good to decouple the specs and the implementation.
Especially if the rules appear to be an incomprehensible ad-hoc mix of principles taken from a dozen not-quite-fully-baked AI dissertations.
I still don't think I truly understand how RDF is supposed to work...
I don't think anyone does.
I'm not saying that the semantic web is bullshit, but it does trigger my bullshit detector. At least one of them must be broken.
I'd be a lot happier (and so would my co-workers) standing while working if I wasn't in a cubicle, looming over my neighbors... I do a lot of work standing up, but I wish my cube walls were a bit higher (i.e., like all the way to the ceiling...).
It's more accurate to say that they're not afraid to buy large-scale projects.
Nerds are a small demographic, but they can also be the canary in the coal mine with stuff like this. Or not. Jeepers. Someone out to FUD Apple this week, or something?
IBM had server sales of more than five billion dollars last year (or three billion, if you don't count mainframes). Even lowly Sun beats out Dell, which comes in at almost $1B.
Keep in mind that this is just for one year. Pick your favorite guess for how large Googles server farm is and divide by the average age of those machines. Do you still think they're assembling more than a billion dollars of hardware per year?
Seriously -- do you expect him to hand out copies of a few dozen RFCs and a map of the backbone sites and say "here, read this, and everything will be crystal clear." Politicians have better things to do than try to understand BGP.