Sandra overestimates Intel CPUs relative to AMD CPUs by a large margin, when the vast majority of real life tests are exactly opposite. I don't trust its benchmarking much at all.
We also, for the most part, didn't have the absolute poverty that we have today. That's the real reason people didn't leave.
Imagine: it's 2 days before a big hurricane hits. You're a single mom (bear with me, I realize this is Slashdot:) ), have a Buick that's not running well, three kids, and $20 to last through the end of the week. How the hell are you supposed to pick up and go to Houston?
The ones that had the money and didn't go.... they were dumb, and deserved the later problems. But an awful, awful lot of those folks just didn't have many options.
You'll still be able to get it for free... in fact, the more it's distributed for free, the more Apple will make.
They're not really selling the bits, although they're pretending to. What they're selling is convenient, automated delivery, and super-convenient playback. It blends many of the best elements of the computer and a VCR. So the more available it is online, the more people will be interested, and the more will sign up for the automated delivery service.
This is the first really definite step toward the Holy Grail of convergence.
I might even subscribe. It'd take more than 10 bucks' worth of time to find and download these episodes anyway.
"The most exciting phrase to hear in science, the one that heralds new discoveries, is not Eureka, but rather 'Hmm, that's funny...'"
-- Isaac Asimov
This is potentially a very, very big deal. The temperature is NOT the most important thing... that's the headline for dummies.
The important part: they're getting out more energy than they're putting in, and they don't understand why.
Re:I had been looking forward to the B5 game.
on
Cut Down In Their Prime
·
· Score: 2, Interesting
Yes! I'm not sure where I got the idea from, but I've been thinking that's what space combat would be like. It would be sneaky, not flashy. Space is enormous, and it's hard to detect things at long range. It would be all about the sensors.
If your sensors (or your tactical person) was better than the enemy's, and you detected him before he detected you... you'd do some kind of a springloaded launch that didn't release any radiative energy. You'd wait for the missile to get a nice long, long way from you, and then activate it, so that the enemy couldn't discern where you were. And ideally, the missile would have a very low-power burn to start with.... even a few minutes at 0.01g acceleration will result in a very large absolute velocity. Whisper the engine a few minutes, get up a good amount of velocity, and then coast in silently, hoping for the explosion to be a total surprise.
If you actually got to the point of needing short-range guns, it would be a massive tactical failure by both parties. Most of the time, space combat would be over before you had ever even figured out where the enemy was. It would be vast stretches of boredom interspersed with short periods of absolute terror.
Unfortunately, this would make _very_ boring games, I think.
Further, you only have to give the code to the people you distribute the binaries to, NOT the public at large. You don't owe the public a _damn thing_ under the GPL, just the people to whom you give binaries.
The GPL doesn't require source distribution for 'internal usage', but in large companies, what's internal and what's external can get fuzzy. The safe thing to do is to always distribute the source with the binary.
I'm a little unclear on what would happen if a rogue employee distributed your internal source code without permission... I believe that the GPL requires a chain of distribution permission. If you authorize Subsidiaries A, B, and C to have your GPL code, they definitely have the right to give that code to anyone they want, and you cannot legally restrict them from doing so. (you could punish them in other ways, I imagine, but you couldn't sue them for copyright infringement.) But if a rogue employee of one of those subsidiaries released the code, that transfer would be unauthorized, and I don't think anyone further down the transfer chain would be entitled to use it.
Practically, you might not be able to stop copying, but as far as I know, you'd have the same rights to punish copyright infringement that Microsoft would. (this actually happened to Microsoft, and they've sued at least one person distributing their code.)
Basically, by using the GPL, any 'customer' can become a competitor. If you choose your customers carefully, that doesn't *have* to happen. You don't owe the public at large anything at all.
SQLite doesn't seem to be very fast. I know the SlimDevices people are having some trouble with it. They write SlimServer, an open-source Perl server that indexes music and drives the company's (excellent) Squeezebox players.
The problem seems to happen when people have very large collections, greater than 10,000 tracks... updates become slow, and the whole system gets a little sluggish. Apparently, when using MySQL, the problem goes away completely... or at least until someone gets to 100k tracks or something.:)
Perhaps the Slim team is doing something wrong, but they're definitely seeing some performance issues with SQLite.
Actually, in Exchange, replies and CCs don't matter much. If you have forty people with the same 100mb attachment, it takes up only 100mb in the store, plus forty pointers. (tiny). And if 35 of those people 'delete' their attachment, the 100mb will still be used; your database size will barely shrink. Only if all references to an object are deleted will the space be auto-reclaimed. You can run into a problem when it's forwarded out of the company and then forwarded back IN, but as long as it stays within Exchange, it's just a bunch of pointers, not a bunch of 100mb attachments.
Limiting attachment sizes seems to curb the worst of the problems... but a lot of non-technical people will scream and kick about having to upload files to a server. When you explain to them that email storage is extremely, extremely expensive (because it has to be hyper-reliable), and website storage can be very cheap, they're often more accommodating. And you can usually automate it fairly well with a good client, like VanDyke's stuff.
I usually offer to set up a cron job to wipe a web transfer directory every day... this means the user doesn't need to remove the files they've uploaded. (so they don't give today's files to tomorrow's recipient by accident.) Some people like that: some people don't. Some want both a temporary and a permanent site, which is easy to set up.
Routine external-user password changes are a very good idea in this kind of setup. Fortunately, it's easy to script. It can run with the file-wipe.... autogenerate a new http auth password for the day and email it to the user. If there were no files to wipe, don't make a new password.
Whatever they like is cool with me, as long as they don't use Exchange for file storage.:) Once upon a time, I liked having people be able to email everything... but files have gotten so huge, and storage and backup for a big Exchange server is so obscenely expensive, that I regretfully discourage it now.
Re:Only real answer is free character transfer
on
World of Queuecraft
·
· Score: 1
There's a simple way to fix any economic problems. Transferred characters can bring only soulbound items, crafting tools, and 1g per level. (ie, at level 60, you can bring 60 gold and all your soulbound stuff.) That's enough to get them a start, without supporting farmers.
Most people would switch early in their careers, maybe shopping around, looking for a good server. Net economic effect: nearly zero. High-level characters wouldn't switch much, because they would lose a lot, but they probably wouldn't _anyway_.
Perhaps crafting supplies could be brought too, but it'd take only a few people transferring in, each with a bankful of dark iron bars, to knock the economy off kilter for awhile.
They probably do, but they must have a lot of admins for 1200 servers. If they're all logging in as root (standard practice, albeit suboptimal), then they'd all be sharing a.bash_profile. Probably, management's fear of an alias messing up a real admin is higher than their desire to please a help-desk guy.
Submitter: SecureCRT from VanDyke Software (about $90 if I remember correctly), will support login scripts. That would solve your problem nicely.
If you can't afford that, or aren't allowed to install it, you could download your aliases over a network connection each time you connect to a machine.
It would be better to run it over ssh copy than HTTP, so that it's not traveling in plaintext. A command like:
would do what you needed. This command would simply cat a remote file over an SSH connection.... which eval then executes. You'd have to host it on a machine that all the servers could access. Depending on your security setup, that could potentially be difficult.
You could get fancy and set up agent forwarding if you wanted to save typing the extra password. Sadly, agent forwarding can be something of a security risk, so it may be disabled on your servers. I'd suggest not doing that without getting explicit permission from the more senior people.
Remember that you need to secure the machine hosting the script at least as well as the most secure machine from which you'll be running it. If the script is compromised, any machine on which you run it will be also.
The biggest issue you're going to have is that laptops are inherently unreliable. MAKE SURE that upper management realizes that you're going to need more people in the IT department if they want to do laptops.
Since the students have to pay for them, rather than the University covering the expense, you might want to push them toward specifying better-quality gear, like Thinkpads. This will raise the tuition somewhat for the students, but it'll decrease your load quite a bit. The students will undboutedly squawk about it, but good laptops won't break as much, and should last four years. Students won't know to be happy about this... all they'll do is complain about the cost. In my opinion, it's better to have them squawk once when school starts, instead of many times over their school career.
If management specifies lower build-quality stuff (Dells, for instance, are really not very good), then you will need even more help in IT, and you'll need to be prepared to deal with unhappy students. Lower-cost laptops are probably more expensive, over the long run.
If that happens, remember that the students don't know any better, they're getting screwed as much as you are... try to stress backups and the expectation of machine failure. If the students are taught to expect the laptop to break at any time, it will likely decrease your stress one heck of a lot.
If you have the infrastructure and space for it, take regular images of their systems.... that way, when systems fail, you can slap the last backup onto a loaner machine while you get them back up and running. If you use smaller drives (20 or 40 gigs), it'll be much less expensive to offer that service.
And give them a way to make images onto their home computers, whether or not you also offer net backups. Symantec's Ghost, for instance, has changed from a system-image tool to a system-backup tool. The new version sucks as a system imager, but it's a good backup utility.
Again, this needs to be a whole mindset change. It's more of a big deal than you may realize. With desktops and servers, the expectation is that things will work almost all the time, and failures are treated as unusual exceptions. You take backups, but you don't have to rebuild systems too often.
With laptops, especially cheap ones, the whole organization has to switch to failure-expected mode. You're not just in backups-as-insurance mode anymore... it's more backups-as-daily-necessity. And you will need to become very efficient at repair... you're going to be doing a lot of it.
While your code is very, very important to you, to that bank, it's just a program they want to buy. To the auditor (if it's a separate organization, as it probably legally has to be), it's just a bunch of code that the bank wants checked out.
Your company, virtually certainly, isn't even vaguely important enough to them to mess with. If that code leaked and the word got out, the reputation of both the bank and the auditor would be badly damaged...a financial loss greatly in excess of the net worth of your entire company.
If for some reason they wanted to leak the code, it would be a lot cheaper for them to just buy you out, lock, stock and barrel.
Use your brain. Give them the damn code. They'll probably treat it better than you guys do. They have a lot more to lose if they don't.
Back in the days of the front-panel LCDs that announced how fast the CPU inside was, I was pretty jazzed to get my first full-three-digit one. They were originally two digits, and then added a '1' in front, and finally went to three.
I set mine to 666 and kept it as long as I kept that case. If anyone asked, I told them the machine was beastly fast.
That's the downside to the on-die memory controller. AMD has to make certain you don't mix the DDR-only chips with the DDR2-only motherboards, and vice versa.
Had the memory controller been on the Northbridge, a la Intel, they could have kept the old sockets, but then they wouldn't have had as much of a performance advantage.
Of course they're denying the delay...
on
Sony Denies PS3 Delay
·
· Score: 3, Interesting
Of course they're denying the delay. Sony, if you'll remember, lies. They'll do anything to blunt momentum of the 360. If they keep people thinking that the PS3 is just around the corner, they'll hold out just a little longer. And as long as they don't release specs, people can fantasize.
They are NOT shipping in Spring. No way. Not happening. Blu-Ray isn't even finished yet. There's no specs, no launch titles, nothing but empty shells and promises. They need time to work out their manufacturing process and build up stock for the retailers, and you can't exactly go into full-scale production when you don't even have finalized hardware.
They could take a page from the NVidia and ATI playbooks and paper-launch it, I suppose, but that would be the worst of all possible outcomes.... people will finally be able to make an intelligent choice whether the 360 or the PS3 is better, and by then, the 360 is likely to actually be available.
If Microsoft can execute and get some really good, next-gen games out, they could build up some momentum. The games that are out now are, to my perception, competent but not very exciting. They have good graphics, but there's not much new, gameplay-wise. Kameo has some good moments, but it's too short, and Perfect Dark feels very unpolished and unfinished. Project Gotham Racing is probably the best game so far, and it's just not that much different from the old games... the graphics are great, sure, but that's about all there is to it. (I've heard CoD2 is very good on the 360, but I already have it on the PC and it seemed dumb to buy it again.)
The Fight Night demo looks promising, I love how there's no on-screen gauges, just two fighters in a ring. And Oblivion may be truly next-gen content... we'll have to see.
So far, the best game I've played on the 360 is the $5 Geometry Wars. A $400 console to play a $5 game... and it's the best game on the system. Nintendo may be on to something.....
Oh, and to the AC who called me a 'good liddle fanboy' for saying that Sony wouldn't ship in Spring the last time we had this conversation.... here's a big middle finger just for you.
I didn't like SS2 as well, mostly because I found it impossible to finish.:( I'm sure it's great, but I've never had the full experience.... with NOLF, using your brain and being sneaky, you could avoid most conflict and come through most levels nearly unscathed.
What I've read suggests that the Psi track in SS2 was super-hard. That's what I was trying to do both times I've played it, so that may be part of the problem. I suppose a trainer or a cheat might have worked, but in a survival-horror type game, that just seems so contrary to the fundamental goal that I never tried one.
And the respawning monsters really pissed me off, so I eventually gave up on it completely.
Oh, I need to take that back.... SECOND best game of its generation. "No One Lives Forever" shipped right around the same time, and it was, in my opinion, absolutely the best FPS ever done. Superb voice acting, incredible variety, good story, top-notch level design. If Sacrifice knocked one out of the park, NOLF made low orbit.
NOLF2, on the other hand, just wasn't that good. Buy the first one, but skip the second. It has moments, but it's really not worth hunting down... a pale shadow of its predecessor.
Sacrifice is probably the best game of its generation. It's interesting how much it looks like Giants:Citizen Kabuto.... it shipped at very nearly the same time, and it has much of the same look, at least in spots. It's almost eerie.
Sadly, Giants got all the press, even though it wasn't nearly as good a game. People liked the idea of playing the Giant more than they liked the idea of a really good game. And Walmart didn't like it.
Sacrifice would have been better if the multiplayer were stronger... the battles ended up mostly being "my clump of critters versus your clump of critters". It was very difficult to figure out just what was going on, or why you'd won or lost, because there was so much happening in such a small space... not to mention your limited perspective, stuck down in the world itself.
As a single-player game, though, it was truly excellent, with great replay value. It's a classic game in the old sense... with many paths to follow, many things to see, and a great deal of gameplay, but with the graphics of a fairly modern game.
Sacrifice also woke gaming up to the power of Wal-Mart. They refused to carry this game because it involved sacrificing souls on altars. (that's how you gain power: you kill your enemies' creatures, and convert their souls to your cause, bringing them back as your own creature types afterward.) And that, as far as I know, killed its sales. There have been no further games involving sacrifices, if you'll notice.
If you're tired of the latest trend, '8 hours of gameplay on rails for $50' (ie, F.E.A.R), and you're willing to invest some time learning a real classic, hunt down Sacrifice on EBay.
It's pretty annoying that they put the NVidia card in for some benchmarks, and then didn't just LEAVE IT IN. They explicitly mention in one test that they think ATI graphics are having trouble with OpenGL (no shock, ATI drivers have sucked rat fur in this area for many years). Instead of twigging to 'hey, let's set this to be as fair as we can', they just accept the screwed up results! That's really dumb... they're not thinking it through. They claim to be testing the CHIPS, not the LAPTOPS.
I get so frustrated with benchmarks in general... they so often miss really obvious stuff like this. If you're trying to test a CPU, then you do your best to remove as many other variables as possible. Use the same damn video card. Test what you SAY you are testing. Sheesh.
I think it would have been interesting to see power consumption scores both with and without the NVidia card, too. It'd be nice to try to separate the video power requirements from the CPU/chipset requirements.
Re:"Non-hard-core gamers" aren't playing anymore
on
Mario All Grown Up?
·
· Score: 1
Considering you didn't read his post, it sounds more like short attention span syndrome to m... ooh, shiny!
Basically, they point out that Intel's dual core processors are already starved on the FSB, and loading two more cores isn't going to do very much. He seems to expect that, until Intel gets their FSB in order(which won't happen until 2008), AMD is going to stomp all over them. He says that Intel's cores are excellent, but without CSI (their new FSB), it may not matter much.
My own projection is that the extra contention may end up imposing a net speed _penalty_ for many workloads. That is, however, pure speculation from an amateur, based mostly on the dismal performance of the first dual-CPU G4 Macs.
Dude, the government is now saying that they can search anything, anytime. They're putting together a system to track the entire Internet. They can just disappear people they don't like, and can hold them indefinitely without even access to a lawyer. White House counsel Steve Bradbury now says Bush can kill 'terrorists' on US soil without a warrant or any judicial review.
In that kind of a climate, do you REALLY think Google is going to give its all in defense of freedom? Because their 'all' might literally mean body count.
Slashdot Mirror
Sandra overestimates Intel CPUs relative to AMD CPUs by a large margin, when the vast majority of real life tests are exactly opposite. I don't trust its benchmarking much at all.
We also, for the most part, didn't have the absolute poverty that we have today. That's the real reason people didn't leave.
:) ), have a Buick that's not running well, three kids, and $20 to last through the end of the week. How the hell are you supposed to pick up and go to Houston?
Imagine: it's 2 days before a big hurricane hits. You're a single mom (bear with me, I realize this is Slashdot
The ones that had the money and didn't go.... they were dumb, and deserved the later problems. But an awful, awful lot of those folks just didn't have many options.
Decades ago, Americans weren't this poor.
You'll still be able to get it for free... in fact, the more it's distributed for free, the more Apple will make.
They're not really selling the bits, although they're pretending to. What they're selling is convenient, automated delivery, and super-convenient playback. It blends many of the best elements of the computer and a VCR. So the more available it is online, the more people will be interested, and the more will sign up for the automated delivery service.
This is the first really definite step toward the Holy Grail of convergence.
I might even subscribe. It'd take more than 10 bucks' worth of time to find and download these episodes anyway.
This is potentially a very, very big deal. The temperature is NOT the most important thing... that's the headline for dummies.
The important part: they're getting out more energy than they're putting in, and they don't understand why.
Yes! I'm not sure where I got the idea from, but I've been thinking that's what space combat would be like. It would be sneaky, not flashy. Space is enormous, and it's hard to detect things at long range. It would be all about the sensors.
If your sensors (or your tactical person) was better than the enemy's, and you detected him before he detected you... you'd do some kind of a springloaded launch that didn't release any radiative energy. You'd wait for the missile to get a nice long, long way from you, and then activate it, so that the enemy couldn't discern where you were. And ideally, the missile would have a very low-power burn to start with.... even a few minutes at 0.01g acceleration will result in a very large absolute velocity. Whisper the engine a few minutes, get up a good amount of velocity, and then coast in silently, hoping for the explosion to be a total surprise.
If you actually got to the point of needing short-range guns, it would be a massive tactical failure by both parties. Most of the time, space combat would be over before you had ever even figured out where the enemy was. It would be vast stretches of boredom interspersed with short periods of absolute terror.
Unfortunately, this would make _very_ boring games, I think.
Further, you only have to give the code to the people you distribute the binaries to, NOT the public at large. You don't owe the public a _damn thing_ under the GPL, just the people to whom you give binaries.
The GPL doesn't require source distribution for 'internal usage', but in large companies, what's internal and what's external can get fuzzy. The safe thing to do is to always distribute the source with the binary.
I'm a little unclear on what would happen if a rogue employee distributed your internal source code without permission... I believe that the GPL requires a chain of distribution permission. If you authorize Subsidiaries A, B, and C to have your GPL code, they definitely have the right to give that code to anyone they want, and you cannot legally restrict them from doing so. (you could punish them in other ways, I imagine, but you couldn't sue them for copyright infringement.) But if a rogue employee of one of those subsidiaries released the code, that transfer would be unauthorized, and I don't think anyone further down the transfer chain would be entitled to use it.
Practically, you might not be able to stop copying, but as far as I know, you'd have the same rights to punish copyright infringement that Microsoft would. (this actually happened to Microsoft, and they've sued at least one person distributing their code.)
Basically, by using the GPL, any 'customer' can become a competitor. If you choose your customers carefully, that doesn't *have* to happen. You don't owe the public at large anything at all.
SQLite doesn't seem to be very fast. I know the SlimDevices people are having some trouble with it. They write SlimServer, an open-source Perl server that indexes music and drives the company's (excellent) Squeezebox players.
:)
The problem seems to happen when people have very large collections, greater than 10,000 tracks... updates become slow, and the whole system gets a little sluggish. Apparently, when using MySQL, the problem goes away completely... or at least until someone gets to 100k tracks or something.
Perhaps the Slim team is doing something wrong, but they're definitely seeing some performance issues with SQLite.
Actually, in Exchange, replies and CCs don't matter much. If you have forty people with the same 100mb attachment, it takes up only 100mb in the store, plus forty pointers. (tiny). And if 35 of those people 'delete' their attachment, the 100mb will still be used; your database size will barely shrink. Only if all references to an object are deleted will the space be auto-reclaimed. You can run into a problem when it's forwarded out of the company and then forwarded back IN, but as long as it stays within Exchange, it's just a bunch of pointers, not a bunch of 100mb attachments.
:) Once upon a time, I liked having people be able to email everything... but files have gotten so huge, and storage and backup for a big Exchange server is so obscenely expensive, that I regretfully discourage it now.
Limiting attachment sizes seems to curb the worst of the problems... but a lot of non-technical people will scream and kick about having to upload files to a server. When you explain to them that email storage is extremely, extremely expensive (because it has to be hyper-reliable), and website storage can be very cheap, they're often more accommodating. And you can usually automate it fairly well with a good client, like VanDyke's stuff.
I usually offer to set up a cron job to wipe a web transfer directory every day... this means the user doesn't need to remove the files they've uploaded. (so they don't give today's files to tomorrow's recipient by accident.) Some people like that: some people don't. Some want both a temporary and a permanent site, which is easy to set up.
Routine external-user password changes are a very good idea in this kind of setup. Fortunately, it's easy to script. It can run with the file-wipe.... autogenerate a new http auth password for the day and email it to the user. If there were no files to wipe, don't make a new password.
Whatever they like is cool with me, as long as they don't use Exchange for file storage.
There's a simple way to fix any economic problems. Transferred characters can bring only soulbound items, crafting tools, and 1g per level. (ie, at level 60, you can bring 60 gold and all your soulbound stuff.) That's enough to get them a start, without supporting farmers.
Most people would switch early in their careers, maybe shopping around, looking for a good server. Net economic effect: nearly zero. High-level characters wouldn't switch much, because they would lose a lot, but they probably wouldn't _anyway_.
Perhaps crafting supplies could be brought too, but it'd take only a few people transferring in, each with a bankful of dark iron bars, to knock the economy off kilter for awhile.
They probably do, but they must have a lot of admins for 1200 servers. If they're all logging in as root (standard practice, albeit suboptimal), then they'd all be sharing a .bash_profile. Probably, management's fear of an alias messing up a real admin is higher than their desire to please a help-desk guy.
/path/to/scriptfile"`
Submitter: SecureCRT from VanDyke Software (about $90 if I remember correctly), will support login scripts. That would solve your problem nicely.
If you can't afford that, or aren't allowed to install it, you could download your aliases over a network connection each time you connect to a machine.
It would be better to run it over ssh copy than HTTP, so that it's not traveling in plaintext. A command like:
eval `ssh user@scripthost "cat
would do what you needed. This command would simply cat a remote file over an SSH connection.... which eval then executes. You'd have to host it on a machine that all the servers could access. Depending on your security setup, that could potentially be difficult.
You could get fancy and set up agent forwarding if you wanted to save typing the extra password. Sadly, agent forwarding can be something of a security risk, so it may be disabled on your servers. I'd suggest not doing that without getting explicit permission from the more senior people.
Remember that you need to secure the machine hosting the script at least as well as the most secure machine from which you'll be running it. If the script is compromised, any machine on which you run it will be also.
The biggest issue you're going to have is that laptops are inherently unreliable. MAKE SURE that upper management realizes that you're going to need more people in the IT department if they want to do laptops.
Since the students have to pay for them, rather than the University covering the expense, you might want to push them toward specifying better-quality gear, like Thinkpads. This will raise the tuition somewhat for the students, but it'll decrease your load quite a bit. The students will undboutedly squawk about it, but good laptops won't break as much, and should last four years. Students won't know to be happy about this... all they'll do is complain about the cost. In my opinion, it's better to have them squawk once when school starts, instead of many times over their school career.
If management specifies lower build-quality stuff (Dells, for instance, are really not very good), then you will need even more help in IT, and you'll need to be prepared to deal with unhappy students. Lower-cost laptops are probably more expensive, over the long run.
If that happens, remember that the students don't know any better, they're getting screwed as much as you are... try to stress backups and the expectation of machine failure. If the students are taught to expect the laptop to break at any time, it will likely decrease your stress one heck of a lot.
If you have the infrastructure and space for it, take regular images of their systems.... that way, when systems fail, you can slap the last backup onto a loaner machine while you get them back up and running. If you use smaller drives (20 or 40 gigs), it'll be much less expensive to offer that service.
And give them a way to make images onto their home computers, whether or not you also offer net backups. Symantec's Ghost, for instance, has changed from a system-image tool to a system-backup tool. The new version sucks as a system imager, but it's a good backup utility.
Again, this needs to be a whole mindset change. It's more of a big deal than you may realize. With desktops and servers, the expectation is that things will work almost all the time, and failures are treated as unusual exceptions. You take backups, but you don't have to rebuild systems too often.
With laptops, especially cheap ones, the whole organization has to switch to failure-expected mode. You're not just in backups-as-insurance mode anymore... it's more backups-as-daily-necessity. And you will need to become very efficient at repair... you're going to be doing a lot of it.
While your code is very, very important to you, to that bank, it's just a program they want to buy. To the auditor (if it's a separate organization, as it probably legally has to be), it's just a bunch of code that the bank wants checked out.
Your company, virtually certainly, isn't even vaguely important enough to them to mess with. If that code leaked and the word got out, the reputation of both the bank and the auditor would be badly damaged...a financial loss greatly in excess of the net worth of your entire company.
If for some reason they wanted to leak the code, it would be a lot cheaper for them to just buy you out, lock, stock and barrel.
Use your brain. Give them the damn code. They'll probably treat it better than you guys do. They have a lot more to lose if they don't.
Back in the days of the front-panel LCDs that announced how fast the CPU inside was, I was pretty jazzed to get my first full-three-digit one. They were originally two digits, and then added a '1' in front, and finally went to three.
I set mine to 666 and kept it as long as I kept that case. If anyone asked, I told them the machine was beastly fast.
That's the downside to the on-die memory controller. AMD has to make certain you don't mix the DDR-only chips with the DDR2-only motherboards, and vice versa.
Had the memory controller been on the Northbridge, a la Intel, they could have kept the old sockets, but then they wouldn't have had as much of a performance advantage.
Of course they're denying the delay. Sony, if you'll remember, lies. They'll do anything to blunt momentum of the 360. If they keep people thinking that the PS3 is just around the corner, they'll hold out just a little longer. And as long as they don't release specs, people can fantasize.
They are NOT shipping in Spring. No way. Not happening. Blu-Ray isn't even finished yet. There's no specs, no launch titles, nothing but empty shells and promises. They need time to work out their manufacturing process and build up stock for the retailers, and you can't exactly go into full-scale production when you don't even have finalized hardware.
They could take a page from the NVidia and ATI playbooks and paper-launch it, I suppose, but that would be the worst of all possible outcomes.... people will finally be able to make an intelligent choice whether the 360 or the PS3 is better, and by then, the 360 is likely to actually be available.
If Microsoft can execute and get some really good, next-gen games out, they could build up some momentum. The games that are out now are, to my perception, competent but not very exciting. They have good graphics, but there's not much new, gameplay-wise. Kameo has some good moments, but it's too short, and Perfect Dark feels very unpolished and unfinished. Project Gotham Racing is probably the best game so far, and it's just not that much different from the old games... the graphics are great, sure, but that's about all there is to it. (I've heard CoD2 is very good on the 360, but I already have it on the PC and it seemed dumb to buy it again.)
The Fight Night demo looks promising, I love how there's no on-screen gauges, just two fighters in a ring. And Oblivion may be truly next-gen content... we'll have to see.
So far, the best game I've played on the 360 is the $5 Geometry Wars. A $400 console to play a $5 game... and it's the best game on the system. Nintendo may be on to something.....
Oh, and to the AC who called me a 'good liddle fanboy' for saying that Sony wouldn't ship in Spring the last time we had this conversation.... here's a big middle finger just for you.
I didn't like SS2 as well, mostly because I found it impossible to finish. :( I'm sure it's great, but I've never had the full experience.... with NOLF, using your brain and being sneaky, you could avoid most conflict and come through most levels nearly unscathed.
What I've read suggests that the Psi track in SS2 was super-hard. That's what I was trying to do both times I've played it, so that may be part of the problem. I suppose a trainer or a cheat might have worked, but in a survival-horror type game, that just seems so contrary to the fundamental goal that I never tried one.
And the respawning monsters really pissed me off, so I eventually gave up on it completely.
No arguments about the cut dungeons, but out of curiosity, what did you think was wrong with the sailing?
Oh, I need to take that back.... SECOND best game of its generation. "No One Lives Forever" shipped right around the same time, and it was, in my opinion, absolutely the best FPS ever done. Superb voice acting, incredible variety, good story, top-notch level design. If Sacrifice knocked one out of the park, NOLF made low orbit.
NOLF2, on the other hand, just wasn't that good. Buy the first one, but skip the second. It has moments, but it's really not worth hunting down... a pale shadow of its predecessor.
Sacrifice is probably the best game of its generation. It's interesting how much it looks like Giants:Citizen Kabuto.... it shipped at very nearly the same time, and it has much of the same look, at least in spots. It's almost eerie.
Sadly, Giants got all the press, even though it wasn't nearly as good a game. People liked the idea of playing the Giant more than they liked the idea of a really good game. And Walmart didn't like it.
Sacrifice would have been better if the multiplayer were stronger... the battles ended up mostly being "my clump of critters versus your clump of critters". It was very difficult to figure out just what was going on, or why you'd won or lost, because there was so much happening in such a small space... not to mention your limited perspective, stuck down in the world itself.
As a single-player game, though, it was truly excellent, with great replay value. It's a classic game in the old sense... with many paths to follow, many things to see, and a great deal of gameplay, but with the graphics of a fairly modern game.
Sacrifice also woke gaming up to the power of Wal-Mart. They refused to carry this game because it involved sacrificing souls on altars. (that's how you gain power: you kill your enemies' creatures, and convert their souls to your cause, bringing them back as your own creature types afterward.) And that, as far as I know, killed its sales. There have been no further games involving sacrifices, if you'll notice.
If you're tired of the latest trend, '8 hours of gameplay on rails for $50' (ie, F.E.A.R), and you're willing to invest some time learning a real classic, hunt down Sacrifice on EBay.
Doesn't say THEY have to label it, though.
A Dymo Labelmaker is one heck of a lot cheaper than an Intel iMac.
That's fine, but then say 'we're testing the laptops'... and skip the tests with the NVidia card. As is, it's just an inconsistent mess.
It's pretty annoying that they put the NVidia card in for some benchmarks, and then didn't just LEAVE IT IN. They explicitly mention in one test that they think ATI graphics are having trouble with OpenGL (no shock, ATI drivers have sucked rat fur in this area for many years). Instead of twigging to 'hey, let's set this to be as fair as we can', they just accept the screwed up results! That's really dumb... they're not thinking it through. They claim to be testing the CHIPS, not the LAPTOPS.
I get so frustrated with benchmarks in general... they so often miss really obvious stuff like this. If you're trying to test a CPU, then you do your best to remove as many other variables as possible. Use the same damn video card. Test what you SAY you are testing. Sheesh.
I think it would have been interesting to see power consumption scores both with and without the NVidia card, too. It'd be nice to try to separate the video power requirements from the CPU/chipset requirements.
Considering you didn't read his post, it sounds more like short attention span syndrome to m... ooh, shiny!
As usual, if you want good CPU info, Ars Technica is the place to look. They have a blurb on Intel's 4-way core plans here.
Basically, they point out that Intel's dual core processors are already starved on the FSB, and loading two more cores isn't going to do very much. He seems to expect that, until Intel gets their FSB in order(which won't happen until 2008), AMD is going to stomp all over them. He says that Intel's cores are excellent, but without CSI (their new FSB), it may not matter much.
My own projection is that the extra contention may end up imposing a net speed _penalty_ for many workloads. That is, however, pure speculation from an amateur, based mostly on the dismal performance of the first dual-CPU G4 Macs.
Dude, the government is now saying that they can search anything, anytime. They're putting together a system to track the entire Internet. They can just disappear people they don't like, and can hold them indefinitely without even access to a lawyer. White House counsel Steve Bradbury now says Bush can kill 'terrorists' on US soil without a warrant or any judicial review.
In that kind of a climate, do you REALLY think Google is going to give its all in defense of freedom? Because their 'all' might literally mean body count.