...I'd like to see anyone on the Tour match his feat.
So why doesn't this Mike Curiak enter the TdF?
The fact is that the Tour de France and Great Divide Race are completely different events, and being an expert at one doesn't say anything about how you'd do at the other.
There's a certain price level where you get a quality bike, something that's not Wal-Mart crap. I don't know where it is now, probably around $400 or $500. But you know what, if spending $1000 instead of $500 makes you more likely to get out and ride, then do it. Buy the bike that will make you happy and make you ride more. Don't waste time sneering at the other guy because his bike is too cheap or too expensive
I do in fact have one of those fancy expensive bikes, and I also have a regular beater bike. I do all my own maintenance so the beater only looks like shit, it's well maintained so it's fine to ride. The biggest difference between that bike that cost me $300 in '91 and the fancy bike with Columbus SLX tubing and the Campagnolo components? Yes, the expensive one is much nicer to ride -- the weight difference is perceptible. But most important: it's easy to work on or adjust the expensive bike, and it stays adjusted. The cheaper one is a bit of a pain to work on and it's a little finicky.
The most fun bike is a track bike. Get rid of that heavy derailleur, freewheel and multiple chainrings and cogs. You can toss the rear brake too. That's the light bike that's fun to ride in the city!
... in the UK at least, weed is still sold in ounces.
I think he was mistaken -- weed is in ounces in the US too. It's crack, cocaine, amphetamines etc. that are in grams (or kilos if you're in the importation business).
Speak harshly to your little VAX,
And boot it when it crashes;
It only does it to annoy,
Because the paging thrashes.
(User of 4.2BSD on VAXen in '86... the 11/750, 11/780, and the monster 8650. I remember when NFS first came out -- '87 or so? Mt. Xinu 4.3BSD+NFS if I remember right. Ahh, the good ol' days!)
Memory controllers have been on the motherboard side for a long time, at least on x86.... Having the memory controller on the RAM stick would make RAM sticks very much more expensive to produce, and the margins are already razor-thin.
So ECC RAM is not just a question of "$3 more", then? I've never seen a $100 motherboard with ECC support -- only the "server" boards, which probably have a fatter margin as well.
(I would love to have ECC. In fact when I first heard x86 systems didn't, I was shocked, shocked, I tell you!)
What I remember of my computer hardware classes (a long, long time ago), ECC memory is self-contained: the memory sub-system implements it, it's transparent from outside. But I see that for some reason, x86 motherboards are either designed for ECC or non-ECC RAM. WTF? Why should the motherboard care that the memory is very reliable?
Or is the x86 motherboard design so crappy that the memory controller is actually on the motherboard, and the RAM sticks we buy are just the chips, no control? If that's the case, it's not just a question of ECC being "just a couple of dollars more"; I never see consumer motherboards advertising ECC support.
I don't know why I'm replacing to such obvious flame-bait, but no one said I was smart.
X11 is not like your little bitmapped Micros**t applications: colour depths, fonts, and of course extensions are implemented by the server, not by the client. They will vary between servers. If you want applications to move their connections from one server to another, they will have to re-map their fonts and colours and figure out what to do about missing extensions. Needless to say this will not be transparent to the application.
Everytime we've looked for a boss I've interviewed the candidates. In one case the VP of Engineering as well. Why is that strange? If the company hires an idiot, or someone engineers/developers won't be able to work with, they'll have unhappy employees who might quit. Far better to keep looking than to lose good people. Even in this dire economy, you don't want to fire someone who's competent and already knows the ropes, the code, the design. Middle-managers are much easier to replace than good developers. (I used to be a manager but I went back to being a hacker. I like it more, and it has more job security.)
I use the same thinking interviewing bosses as well as co-workers and employees: will I be able to work with this person? No assholes and no stupid people. Additionally with a manager I want to know if the person will be a good advocate for the team. A manager's primary responsibility is to stop any shit coming down from above.
...it existed at a time so far before many languages that pathetically failed to implement its features, so I'm a little confused at the way the computing world has ignored it, instead of trying to work its principles into modern languages.
Have you looked into Icon, or its extension Unicon? You can make custom control structures (using what are called co-expressions). It also has goal-directed evaluation (backtracking, think continuations of the LISP world) built into its expression evaluation. Of course it's dynamically typed and has all sorts of data structures like lists/trees, associative arrays, sets etc. built-in, plus X11 and Win32 graphics, automatic storage management (garbage collection) and lots of cool features for text manipulation (it's a development of the ideas in SNOBOL4). To all this Unicon adds object-orientedness, POSIX system calls and networking, message-passing, etc. All this in a easy to read block structured (C-like) syntax.
Has a nice optical zoom. How many X makes a nice optical zoom? I suppose that's up to the individual, but I think 10X or more. More is always better when it comes to optical zoom.
From a usage standpoint, yes, more zoom is better. However the laws of physics being what they are, making the zoom range wider means more compromises, more elements, more aberration, more flare, less contrast. A 10x zoom is pretty bad. It looks like around 3x is the best range, which means you need two lenses to cover that 10x range.
And in fact, in the Canon line, there's the 24-70 f/2.8L and the 70-200 f/2.8L -- they're excellent lenses, and f/2.8 means they're on the "large and expensive" side. (Well, they are "L" lenses!) Those lenses just over $1000 each, which for a "serious" amateur is doable.
Digital SLRs are nowhere near as cheap as film ones. The EOS Elan (film) and the EOS 10D (digital) are roughly comparable bodies in features, construction etc. But the Elan is $400, and the 10D is $1500. Plus the 10D has a smaller sensor than the 35mm frame, so that wide 28mm lens gives you the view angle of a 45mm -- just slightly wider than the "normal" 50mm. And it means "bokeh" -- the out-of-focus background -- isn't as nice and soft. (Conversely, you get better apparent depth of field with the smaller sensor.)
You're implying that, everything else being equal, the probabilities change if you pause between the coin flips?
You might be thinking about the conditional probabilities -- the probability of X givem Y has already happened. Or you might be thinking of Type I and Type II errors of predictors.
Wow! Assuming that it doesn't actually burn you, but only causes an intense burning sensation -- think of the possibilities! Make a smaller hand-held version and no more bothering with pesky wires attached to genitals while making sure the illegal combatant doesn't fall off the box.
This, coming from the "teh softwarez must be free-as-in-um-actually-i'm-just-cheap" crowd (which unfortunately makes up the majority of the people who use open source) is absolutely hilarious.
And you know that "the majority of the people who use open source" are cheap bastards, how?
My experience has been that people who use Unix tend to be technically oriented adults who are more aware of ethics, copyrights and patents than the general population. It is self-evident that the Microsoft-using population is where the demand for "cracked" Photoshop and Windows XP registration keys etc. comes from -- those programs are just not available/applicable on Unix.
Writing an OS is even harder than writing a compiler by an order of magnitude, and getting that done within a year may very well be too much for your average undergrad.
No. I used to be a CS professor. I'd expect my students to write a basic OS kernel in a semester, just as I did, and the others in school with me did. We also had to write a basic compiler in one semester. These things are not that hard once you get the basic concepts. The talented programmers had code that looked (was designed) better, was easier to understand, and had fewer bugs.
The time-consuming thing is getting all the little details right so it's actually usable for real work. (No one here is claiming that Linux 0.01 was usable, I hope.)
they can just babble, just as Einstein did about socialism and pacifism, and Bill Joy is doing about science.
What, then, does someone have to do to gain your permission to talk about socialism or pacifism? Where does this wonderful "education" come from that allows you to be infallible in subjects like socialism and pacifism? Why should journalists only talk to these infallible experts?
I'd rather hear Einstein's (and to a lesser degree Joy's) "babblings" than J. Random Blowhard on Slashdot.
With dimensions of 3x4x9, it might be wise to check for the ACME name before attempting to cram it in your pocket.
Nonsense. I've seen one of these things (working) -- it might be awkward if you just wear t-shirt and jeans, but it's no problem with those "cargo" pants that you kids wear these days, or jackets. (That's 0.9 in those numbers you quote, not 9 -- 3 x 4 x 0.9.) I'd find some way to justify the $1500.
Hmmm... I didn't think my experience was atypical. I was copying about 20GB of files across NFS. The machine (running RedHat 8) crashed five times. In fact I wrote a little script that echoed the directories copied, so when it crashed I knew where to start it after the machine came back up. A little while after that I happened to read the Beowulf network card notes (and/or Don Becker's opinions), and they matched me Realtek experiences, and suggested that the Intel EEPro 1000 was the one that worked. (Dang, I couldn't find that page after a quick googling.) I decided that although Intel is evil, I decided I'd switch all the machines over to eepros. After that I tried the same test, of transferring that 20GB via NFS. No crashes. Although I'm sure the temperature, the phase of the moon etc. probably had more to do with it, it was good enough for me.
The Realtek chips might do OK for light loads like web browsing, but try transferring large amounts of data with them. What's the point of a 100 Mbps full-duplex connection when the chip craps out (often crashing the machine)? Go read the Beowulf guys' recommendations. Dirt cheap is right -- you get what you pay for.
you will gain notoriety as the guy who opened maps to the world.
Work done by the US Government is in the public domain. This means USGS maps are free in every sense of the word. Scanned maps (DRGs) are available from a bunch of places; the 1:24,000 7.5' quads are probably what you want. Vector data are available from the Census' TIGER maps.
Yes indeed!
on
Open Maps?
·
· Score: 2, Informative
Free maps from the TIGER data, as well as the (free) software that draws them. Here's Gregg Townsend's package in Icon. (Icon is a free VHLL -- very high level language -- of which Unicon is the current development extension.
The plural of schema is schemata. Greek roots vs. Latin roots, etc.
Or, if you don't want to be called pedantic, just use schemas. If we borrow a word from another language there isn't really a good reason to follow its rules.
(And virii is never correct. It would be the plural of virius, not virus. cf. radius, radii. Just say viruses.)
Slashdot Mirror
I think dydndns sucks but that's just my opinion. Use something like ZoneEdit -- they allow you to add TXT records.
The fact is that the Tour de France and Great Divide Race are completely different events, and being an expert at one doesn't say anything about how you'd do at the other.
I do in fact have one of those fancy expensive bikes, and I also have a regular beater bike. I do all my own maintenance so the beater only looks like shit, it's well maintained so it's fine to ride. The biggest difference between that bike that cost me $300 in '91 and the fancy bike with Columbus SLX tubing and the Campagnolo components? Yes, the expensive one is much nicer to ride -- the weight difference is perceptible. But most important: it's easy to work on or adjust the expensive bike, and it stays adjusted. The cheaper one is a bit of a pain to work on and it's a little finicky.
The most fun bike is a track bike. Get rid of that heavy derailleur, freewheel and multiple chainrings and cogs. You can toss the rear brake too. That's the light bike that's fun to ride in the city!
And boot it when it crashes;
It only does it to annoy,
Because the paging thrashes.
(User of 4.2BSD on VAXen in '86... the 11/750, 11/780, and the monster 8650. I remember when NFS first came out -- '87 or so? Mt. Xinu 4.3BSD+NFS if I remember right. Ahh, the good ol' days!)
Now the Netflix founder is taking the credit? It's been variously attributed to Andy Tanenbaum and Dennis Ritchie. Maybe Karl Kleinpaste?
So ECC RAM is not just a question of "$3 more", then? I've never seen a $100 motherboard with ECC support -- only the "server" boards, which probably have a fatter margin as well.
(I would love to have ECC. In fact when I first heard x86 systems didn't, I was shocked, shocked, I tell you!)
Or is the x86 motherboard design so crappy that the memory controller is actually on the motherboard, and the RAM sticks we buy are just the chips, no control? If that's the case, it's not just a question of ECC being "just a couple of dollars more"; I never see consumer motherboards advertising ECC support.
X11 is not like your little bitmapped Micros**t applications: colour depths, fonts, and of course extensions are implemented by the server, not by the client. They will vary between servers. If you want applications to move their connections from one server to another, they will have to re-map their fonts and colours and figure out what to do about missing extensions. Needless to say this will not be transparent to the application.
I use the same thinking interviewing bosses as well as co-workers and employees: will I be able to work with this person? No assholes and no stupid people. Additionally with a manager I want to know if the person will be a good advocate for the team. A manager's primary responsibility is to stop any shit coming down from above.
And in fact, in the Canon line, there's the 24-70 f/2.8L and the 70-200 f/2.8L -- they're excellent lenses, and f/2.8 means they're on the "large and expensive" side. (Well, they are "L" lenses!) Those lenses just over $1000 each, which for a "serious" amateur is doable.
Digital SLRs are nowhere near as cheap as film ones. The EOS Elan (film) and the EOS 10D (digital) are roughly comparable bodies in features, construction etc. But the Elan is $400, and the 10D is $1500. Plus the 10D has a smaller sensor than the 35mm frame, so that wide 28mm lens gives you the view angle of a 45mm -- just slightly wider than the "normal" 50mm. And it means "bokeh" -- the out-of-focus background -- isn't as nice and soft. (Conversely, you get better apparent depth of field with the smaller sensor.)
You might be thinking about the conditional probabilities -- the probability of X givem Y has already happened. Or you might be thinking of Type I and Type II errors of predictors.
Wow! Assuming that it doesn't actually burn you, but only causes an intense burning sensation -- think of the possibilities! Make a smaller hand-held version and no more bothering with pesky wires attached to genitals while making sure the illegal combatant doesn't fall off the box.
My experience has been that people who use Unix tend to be technically oriented adults who are more aware of ethics, copyrights and patents than the general population. It is self-evident that the Microsoft-using population is where the demand for "cracked" Photoshop and Windows XP registration keys etc. comes from -- those programs are just not available/applicable on Unix.
The time-consuming thing is getting all the little details right so it's actually usable for real work. (No one here is claiming that Linux 0.01 was usable, I hope.)
Yeah, why should teachers be paid more than plumbers? They only take care of our children, not toilets, which everyone knows are more important.
I'd rather hear Einstein's (and to a lesser degree Joy's) "babblings" than J. Random Blowhard on Slashdot.
Hmmm... I didn't think my experience was atypical. I was copying about 20GB of files across NFS. The machine (running RedHat 8) crashed five times. In fact I wrote a little script that echoed the directories copied, so when it crashed I knew where to start it after the machine came back up. A little while after that I happened to read the Beowulf network card notes (and/or Don Becker's opinions), and they matched me Realtek experiences, and suggested that the Intel EEPro 1000 was the one that worked. (Dang, I couldn't find that page after a quick googling.) I decided that although Intel is evil, I decided I'd switch all the machines over to eepros. After that I tried the same test, of transferring that 20GB via NFS. No crashes. Although I'm sure the temperature, the phase of the moon etc. probably had more to do with it, it was good enough for me.
The Realtek chips might do OK for light loads like web browsing, but try transferring large amounts of data with them. What's the point of a 100 Mbps full-duplex connection when the chip craps out (often crashing the machine)? Go read the Beowulf guys' recommendations. Dirt cheap is right -- you get what you pay for.
Free maps from the TIGER data, as well as the (free) software that draws them. Here's Gregg Townsend's package in Icon. (Icon is a free VHLL -- very high level language -- of which Unicon is the current development extension.
Or, if you don't want to be called pedantic, just use schemas. If we borrow a word from another language there isn't really a good reason to follow its rules.
(And virii is never correct. It would be the plural of virius, not virus. cf. radius, radii. Just say viruses.)