Well, on the other hand, everything is expected to be much more efficient today. As a result, design often suffers.
When I say efficient, I guess you could substitute the word cheap. A simple example is office furniture. Many of the desks at my job are from the 60s. All heavy duty metal, and holding up pretty well after almost 50 years, except for the poured vinyl-sort of surfaces on some which are getting mushy (your monitor will sink in a litttle after a year or so).
I hear the company paid something like $500 a piece for the desks in the 60s. That's way more than the company would pay for a desk these days, if you scale for inflation.
Was it worth it? Maybe. It's a shift in attitudes, but it's not necessarily a bad one.
A counterpoint, we have quickly growing mass storage needs at work. For a long time, we suffered with SCSI RAIDs that were consistantly running out of space because the money required to replace them was mind-boggling. I was key in pushing for replacing them with cheap ATA and SATA arrays as soon as the technology to do so was available.
We replace the arrays on a much faster schedule (every 2 years or so), but for perspective, the replacements cost the same as one year of just the support contract on the old arrays, and the space grows by at least 50% with each replacement.
So in some cases, cheap is just cheap, but on the other hand, cheap can mean agile, it means you can keep up with cutting edge without breaking the bank.
So I guess the bottom line is "They sure don't build things like they used to, but that's not always a bad thing".
Human 1.0 is a buggy piece of crap. Apparently there's a hard coded uptime limit of somewhere around 16-48 hours, and rebooting takes up to 12 hours, but usually 8.
There are hundreds of DDoS attacks, including something as trivial as a potassium injection attack.
All in all, I can't recommend Human 1.0 for production use yet.
First off, what's a PSP, and second, what the hell is a UMD? Is that some kind of "unit of mass destruction"?
Re:The smallest possible reason but big enough for
on
Mozilla 1.7.5 Released
·
· Score: 2, Informative
I think you may be thinking of an old behaviour in really old firebird/phoenix/whatever.
It used to be if you clicked in the URL bar, it would highlight the URL without copying it. This would allow you to "paste over" an existing URL without hassle, something that is not easy in X because the way the clipboard works normally.
They changed it back in later version to have a behavior more consistant with normal X behavior.
Power users like big bloated applications that do bunch of different functions?
I don't know where you're from, but the power users I know like small efficient applications that do one thing well, and interface well with other applications. It's the entire UNIX philosophy.
Communicator suite was always a big bloated bad idea inspired by Microsoft's and Lotus's move toward making everything a "suite" around 1996 or so. A very outdated and obselete idea these days.
Sure they fucked up and overly crippled the standalone versions in some really annoying ways, but that doesn't make the fat, bloated, obselete version better.
The air is great here and I have no problems breathing. It's not my fault if people want to crowd together in dirty cities. No matter what some assert, air pollution is a very local problem.
It would be worse. Much worse. At least we can still build our computers from commodity components on a competitive market. Apple would have us tied up in proprietary hardware and software.
It's not really up to the court to judge whether software is "protected enough already"... even the supreme court really can't do that. The constitution gives congress fairly open ended powers to promote science and engineering, and the supreme court has been very reluctatant to put an explicit limit on them.
He's making a due process challenge to this, basically saying that the laws are too vague.
We might not want to dismiss him out of hand as a corporate shill, he does seem to want to get the ambiguity out of the way copyrights are applied to software.
This guy enbodies the hacker spirit. He didn't "invent" anything in the strictest sense, but it's a damn clever hack using the teconology available to him.
You totally missed my point. Very few people need to learn calculus because computers can do it. Automatically.
So let a few well paid mathematician/programmers learn it, and the rest of us can just use the programs they wrote.
No other specialized field forces masses of people to learn the field's skills, why is math somehow different? Why do we need 20million people in the US that know calculus?
That server you linked to has hacked me and put all my accounts on it and has all my files! I better delete them quick before they use them for some nefarious purpose!
Of course, if you are going to go as high as 16% overhead, you could just compute XORs for 20% slices, which gives you 20% overhead (like RAID5), but you only need to download 5 of the 6 slices. If something is corrupt (using a normal overall MD5), download the 6th parity slice and repair.
So you not only get verification, you also get error recovery.
Yeah, that's the problem with kids these days, they give them calculators. Imagine if when the pencil and modern paper came out, they allowed kids to use those! If they had done that, no one would know how to chisel a proper character into stone anymore!
If you asked any of those users if they wanted a program that intercepted and redirected searches and URLs, popped up random ads, etc.. I'm sure they would say no.
It's blaringly obvious that most people don't want spyware.
Same as the other poster, I also advocate a company being able to hire and fire at will for whatever reason...
But a reasonable company will have a reasonable reason, not blindly follow some policy.. to do otherwise gives network admin types too much power over getting someone fired.
I agree with your general sentiment, but not with the idea that companies "grow into" big iron.
I'm also confused that you compare "big iron" with clusters, in my head those are two different things, one is when you need massive I/O with fairly low end massively parallel CPUs, and the other is when you need lots of CPU without much internode communications.
In any case, I can cite many examples of the low-end market driving the high end... Current high performance video cards are completely consumer driven. SCSI innovation has stagnated now that ATA/SATA is eating up the low to mid RAID markets, and massive file storage RAIDs are rarely SCSI/FC anymore unless the company is rolling in cash. The IBM PC itself even, was a low-end offering that has totally revolutionized almost all aspects of modern computing.
In a related story to this, I heard IBM is selling off their PC division. It will be interesting to see how far they take that, I think it would be folly for them to abandon commodity platforms completely, relegating themselves to the same fate as SGI or Sun, a slow shrink into irrelevance.
Slashdot Mirror
Well, on the other hand, everything is expected to be much more efficient today. As a result, design often suffers.
When I say efficient, I guess you could substitute the word cheap. A simple example is office furniture. Many of the desks at my job are from the 60s. All heavy duty metal, and holding up pretty well after almost 50 years, except for the poured vinyl-sort of surfaces on some which are getting mushy (your monitor will sink in a litttle after a year or so).
I hear the company paid something like $500 a piece for the desks in the 60s. That's way more than the company would pay for a desk these days, if you scale for inflation.
Was it worth it? Maybe. It's a shift in attitudes, but it's not necessarily a bad one.
A counterpoint, we have quickly growing mass storage needs at work. For a long time, we suffered with SCSI RAIDs that were consistantly running out of space because the money required to replace them was mind-boggling. I was key in pushing for replacing them with cheap ATA and SATA arrays as soon as the technology to do so was available.
We replace the arrays on a much faster schedule (every 2 years or so), but for perspective, the replacements cost the same as one year of just the support contract on the old arrays, and the space grows by at least 50% with each replacement.
So in some cases, cheap is just cheap, but on the other hand, cheap can mean agile, it means you can keep up with cutting edge without breaking the bank.
So I guess the bottom line is "They sure don't build things like they used to, but that's not always a bad thing".
Human 1.0 is a buggy piece of crap. Apparently there's a hard coded uptime limit of somewhere around 16-48 hours, and rebooting takes up to 12 hours, but usually 8.
There are hundreds of DDoS attacks, including something as trivial as a potassium injection attack.
All in all, I can't recommend Human 1.0 for production use yet.
First off, what's a PSP, and second, what the hell is a UMD? Is that some kind of "unit of mass destruction"?
I think you may be thinking of an old behaviour in really old firebird/phoenix/whatever.
It used to be if you clicked in the URL bar, it would highlight the URL without copying it. This would allow you to "paste over" an existing URL without hassle, something that is not easy in X because the way the clipboard works normally.
They changed it back in later version to have a behavior more consistant with normal X behavior.
Power users like big bloated applications that do bunch of different functions?
I don't know where you're from, but the power users I know like small efficient applications that do one thing well, and interface well with other applications. It's the entire UNIX philosophy.
Communicator suite was always a big bloated bad idea inspired by Microsoft's and Lotus's move toward making everything a "suite" around 1996 or so. A very outdated and obselete idea these days.
Sure they fucked up and overly crippled the standalone versions in some really annoying ways, but that doesn't make the fat, bloated, obselete version better.
The air is great here and I have no problems breathing. It's not my fault if people want to crowd together in dirty cities. No matter what some assert, air pollution is a very local problem.
Heh, you probably already know this, but for the ones that might not, skittles here refers to what we would call bowling.
Thousands of people die from car accidents too, and yet I don't see a push to ban cars.
It's just the cost of progress.
It's not my responsibility to ensure the place I buy my music pays all royalties due. That's between them and the record labels.
allofmp3 is about 10 times less.
It would be worse. Much worse. At least we can still build our computers from commodity components on a competitive market. Apple would have us tied up in proprietary hardware and software.
There are music stores that don't use DRM period. You could just use those, instead of paying 10 times more for crippled files.
It's not really up to the court to judge whether software is "protected enough already"... even the supreme court really can't do that. The constitution gives congress fairly open ended powers to promote science and engineering, and the supreme court has been very reluctatant to put an explicit limit on them.
He's making a due process challenge to this, basically saying that the laws are too vague.
We might not want to dismiss him out of hand as a corporate shill, he does seem to want to get the ambiguity out of the way copyrights are applied to software.
Windows 95 wasn't out.
This guy enbodies the hacker spirit. He didn't "invent" anything in the strictest sense, but it's a damn clever hack using the teconology available to him.
San Francisco... Flaming... suck off... don't turn your back... hehe is it just me? :)
You totally missed my point. Very few people need to learn calculus because computers can do it. Automatically.
So let a few well paid mathematician/programmers learn it, and the rest of us can just use the programs they wrote.
No other specialized field forces masses of people to learn the field's skills, why is math somehow different? Why do we need 20million people in the US that know calculus?
OMGOMGOMG
That server you linked to has hacked me and put all my accounts on it and has all my files! I better delete them quick before they use them for some nefarious purpose!
Of course, if you are going to go as high as 16% overhead, you could just compute XORs for 20% slices, which gives you 20% overhead (like RAID5), but you only need to download 5 of the 6 slices. If something is corrupt (using a normal overall MD5), download the 6th parity slice and repair.
So you not only get verification, you also get error recovery.
Yeah, that's the problem with kids these days, they give them calculators. Imagine if when the pencil and modern paper came out, they allowed kids to use those! If they had done that, no one would know how to chisel a proper character into stone anymore!
If you asked any of those users if they wanted a program that intercepted and redirected searches and URLs, popped up random ads, etc.. I'm sure they would say no.
It's blaringly obvious that most people don't want spyware.
Actually your sig pretty much explains the picture. :)
Which is a fine reason to fire them.
Same as the other poster, I also advocate a company being able to hire and fire at will for whatever reason...
But a reasonable company will have a reasonable reason, not blindly follow some policy.. to do otherwise gives network admin types too much power over getting someone fired.
:) ...
.. SCSI is harder isn't it?
We'll I assumed it was something trivial, but I've seen it sold as a sort of device that has IDE/SCSI/etc
I agree with your general sentiment, but not with the idea that companies "grow into" big iron.
I'm also confused that you compare "big iron" with clusters, in my head those are two different things, one is when you need massive I/O with fairly low end massively parallel CPUs, and the other is when you need lots of CPU without much internode communications.
In any case, I can cite many examples of the low-end market driving the high end... Current high performance video cards are completely consumer driven. SCSI innovation has stagnated now that ATA/SATA is eating up the low to mid RAID markets, and massive file storage RAIDs are rarely SCSI/FC anymore unless the company is rolling in cash. The IBM PC itself even, was a low-end offering that has totally revolutionized almost all aspects of modern computing.
In a related story to this, I heard IBM is selling off their PC division. It will be interesting to see how far they take that, I think it would be folly for them to abandon commodity platforms completely, relegating themselves to the same fate as SGI or Sun, a slow shrink into irrelevance.