Seagate drives still have problems with the sata_sil kernel module. Some models just stop working after a week or two of use, requiring a reboot. Kernel bug #5047
I know it's a lot to ask to have support for such obscure hardware as Silicon Image SATA controllers and Seagate drives.
Stay ahead of the load, but don't build in capacity you don't need.
CPUs are especially expensive right now for very little gain on the high-end.
Make sure to use a PHP accelerator if anything is CPU bound. (assuming you are using the real LAMP and not one of those redefinitions of the P that anti-PHP revisionists like to use).
The next step is caching of static content, this can be as simple as making image links point to a different server, or you could take the next step with an inwardly facing proxy.
But really, don't build in capacity you don't need. When you actually do need it, your bang for the buck will be better because of the advancement of technology.
The review process shouldn't disclose the patent at first. It should disclose the problem the patent solves.
Then challenge a some people well versed in the field to come up with realistic solutions to the problem.
If one of their solutions is very much like the patent being applied for, then deny the patent.
This isn't rocket science.
Re:Top search engine name searches
on
Google Zeitgeist '05
·
· Score: 2, Informative
If you are really interested in more information, sign up for an Adsense advertisers account (free), which will give you unlimited statistics on any particular search term.
After all you need to know what sorts of traffic levels which keywords will generate.
Try an ImageStream rebel router. It can serve as a direct termination point for your T3/T1 with no other equipment, it routes obviously, and it runs linux so you can ssh into it and configure iptables just as any other linux box.
No hard disk, it's flash based for reliability.
With a T3 card it'll be about $7000 so it's not cheap, but if it replaces some overpriced cisco crap along with the firewall, it could be a real money saver.
It's just an option to use the original controls. I doubt it's buggy, it's probably just an extra if statement and a reused older object or a bunch of copy-paste code from the old version.
I remember when slashdot changed from small banners to the huge banners and the midstory ads.
Back then they guaranteed they'd not take flash ads or anything obnoxious. That didn't last long. Once you start to sell out to ad people it always seems to go downhill.
Not necessarily, if you had replied like you just did to my challenging of those particular questions I'd probably get over being pissed off at the initial wordings. Nonverbal communication would be important in judging zealotry, that would indicate a lot, and it's something I can't judge over a medium like this one.
Of course, if you are really of that opinion about unit testing, you might not want to work for me, since we don't use a formal unit testing framework, but rather ad-hoc "unit" testing of writing small test programs for each non-trivial function, and manual regression testing after each change that might have side effects. We do heavily use incremental development, I encourage testing of some sort after each small change or addition.
I wouldn't be opposed to more formal unit testing, but I've not seen any frameworks for doign it with anything other than OOP so we'd have to look at how much development it would take to implement.
I do think most of Joel's questions are important questions that do indicate the maturity of the development process at a company, and I would answer "Yes" truthfully to most of them.
I donno about that Joel guy. It looks like "hallway usability testing" is just some neologism he made up.
I'd be tempted not to hire someone just for making such an obscure reference, thinking they just wanted to sound smarter than me or something.
"The best tools money can buy" would piss me off too, since I'd be hiring them to develop solely on open source software.
Also, your attitude toward unit testing is insulting toward people that do not use OOP. Our codebase does not use OOP, and I have no room for some OOP zealot that would try to migrate our codebase.
I think it would be safer to ask about regression testing, something that is applicable to any codebase.
Careful, that might get you hired to maintain a bunch of spaghetti Javascript code that is totally unmaintainable because they bought into the latest buzzwords.
That doesn't mean that Timmy doesn't want that XBox 360 more than anyone who has the money to lay out. It also doesn't mean that Timmy's parent's can afford to lay out $50 chunks here and there for games either.
Timmy isn't the consumer, his parents are.
They bought food and paid the electric bill right? It's simply that they wanted those things more than they wanted the X-Box.
because some people were willing to pay a ridiculous amount for a 360 all of the consoles should have been sold at a ridiculous price.
He never said that. The higher prices would generate a relatively small amount of profit in the long run, yes. The point of them wouldn't be mainly to generate profit, it would be to ensure fair allocation of a limited resource. Making more profit would be a nice side effect.
Those willing to pay more would do so, and those people are the ones that wanted it the most. People that didn't want it as badly wouldn't pay as much. This is economics 101.
A hard-core fan would be happier paying $700 than not getting one at all. That's the point of economics, the people that want it the most will pay the most.
Higher prices ensure that only the hard-core fans get the console, and other people will just have to wait until the price drops to a point they are willing to pay.
Even on the first page, they act like all these companies were run by idiots, ignoring the possibility of a PC that was supposedly right under their noses.
It wasn't that the technology wasn't ready. Intel, at the time primarily a manufacturer of memory chips, had invented the first microprocessor (the 4-bit 4004) in 1971. This was followed up with the 8-bit 8008 in 1972 and the more-capable 8080 chip in 1974. However, Intel didn't see the potential of its own product, considering it to be useful mainly for calculators, traffic lights, and other embedded applications
That's because that's all it was good for. SMPS technology was in its infancy. Storage technology involved huge platters or huge tapes. RAM was damn expensive.
So what did they think Intel should have done? Released a "PC" in 1971 that weighed 200 pounds with a linear power supply, came with a mini-fridge sized persistant storage unit that held 100k, had 4k RAM and cost $20,000?
The technology indeed wasn't ready. The PC came when it did because technology allowed it to come, not because of lack of vision.
You will run out of manufacturers with that attitude. EVERY hard disk manufacturer has turned out batches of duds.
Here's another bit of news, every hard disk fails, it's only a question of when. 6 years is an excellent run, and 2 years isn't too bad, a little low unless you were thrashing it to hell 24/7.
We run dozens of disks in RAID configuration at work (used to be over 100 disks, see below), we only keep any particular set of disks in service 3 years max. We've found that the statistical failure rate starts rising sharply around 3 years for most models.
3 years also represents about the amount of time for disk space to double lately, so each time we upgrade a RAID we make it more reliable, because we can use about half as many disks.
At this pace we'll have all our storage on single-disk RAID 1 in about 9 years.:)
Anyway, I think individual users have so much "brand superstition" because of the slightly weird statistical distribution of failures with hard disks that can vary so much from model to model and production run to production run.
If I did, I doubt I'd stake my future on the availability of a web service from another company.
If you buy software and the parent company stops offering it or raises the price a lot, OK, no big deal, you can still run it and just not get upgrades.
If Google turns off maps or makes you pay a ton, or goes out of business, you are fucked if you used it for anything important.
In other words, only a fool would use these Google "applications" for anything important to their business.
Other people have posted useful things they have done with them for business, etc, but not one of those applications was anything important, they were more like the "cherry on top".
It's the same old "Application Service Provider" paradigm warmed over. No one wanted to pay then, and no one will want to pay now. And anyone smart won't do anything important with the free services.
I was responding to the asserting that they could be used as disposable product labels are used today, since that's what this article is about, not about some reusable label.
Re:Is Opera Google's doorway to beating Microsoft?
on
Google to Buy Opera?
·
· Score: 1, Flamebait
Google Earth, Google Maps, Google SMS, Google Blogsearch, they're all applications
They aren't applications, they are toys. Yeah they are neat to play with, but they aren't exactly something that businesses will use everyday for anything that matters.
Slashdot Mirror
Seagate drives still have problems with the sata_sil kernel module. Some models just stop working after a week or two of use, requiring a reboot. Kernel bug #5047
I know it's a lot to ask to have support for such obscure hardware as Silicon Image SATA controllers and Seagate drives.
Is that even legal?
I'm pretty sure encrypting phone conversations is still illegal. The Encrypted Communications Privacy Act of 1996 didn't pass.
Stay ahead of the load, but don't build in capacity you don't need.
CPUs are especially expensive right now for very little gain on the high-end.
Make sure to use a PHP accelerator if anything is CPU bound. (assuming you are using the real LAMP and not one of those redefinitions of the P that anti-PHP revisionists like to use).
The next step is caching of static content, this can be as simple as making image links point to a different server, or you could take the next step with an inwardly facing proxy.
But really, don't build in capacity you don't need. When you actually do need it, your bang for the buck will be better because of the advancement of technology.
That's easy to prevent.
The review process shouldn't disclose the patent at first. It should disclose the problem the patent solves.
Then challenge a some people well versed in the field to come up with realistic solutions to the problem.
If one of their solutions is very much like the patent being applied for, then deny the patent.
This isn't rocket science.
If you are really interested in more information, sign up for an Adsense advertisers account (free), which will give you unlimited statistics on any particular search term.
After all you need to know what sorts of traffic levels which keywords will generate.
Try an ImageStream rebel router. It can serve as a direct termination point for your T3/T1 with no other equipment, it routes obviously, and it runs linux so you can ssh into it and configure iptables just as any other linux box.
No hard disk, it's flash based for reliability.
With a T3 card it'll be about $7000 so it's not cheap, but if it replaces some overpriced cisco crap along with the firewall, it could be a real money saver.
The forged signature wouldn't check out as Ebay's though. It wouldn't be there for looks, it would be there to actually use.
how could they?
:)
A browser plugin could do it easily without exposing your private key. Start writing!
It's just an option to use the original controls. I doubt it's buggy, it's probably just an extra if statement and a reused older object or a bunch of copy-paste code from the old version.
I remember when slashdot changed from small banners to the huge banners and the midstory ads.
Back then they guaranteed they'd not take flash ads or anything obnoxious. That didn't last long. Once you start to sell out to ad people it always seems to go downhill.
See, you'd "fail" my employer interview
Not necessarily, if you had replied like you just did to my challenging of those particular questions I'd probably get over being pissed off at the initial wordings. Nonverbal communication would be important in judging zealotry, that would indicate a lot, and it's something I can't judge over a medium like this one.
Of course, if you are really of that opinion about unit testing, you might not want to work for me, since we don't use a formal unit testing framework, but rather ad-hoc "unit" testing of writing small test programs for each non-trivial function, and manual regression testing after each change that might have side effects. We do heavily use incremental development, I encourage testing of some sort after each small change or addition.
I wouldn't be opposed to more formal unit testing, but I've not seen any frameworks for doign it with anything other than OOP so we'd have to look at how much development it would take to implement.
I do think most of Joel's questions are important questions that do indicate the maturity of the development process at a company, and I would answer "Yes" truthfully to most of them.
I donno about that Joel guy. It looks like "hallway usability testing" is just some neologism he made up.
I'd be tempted not to hire someone just for making such an obscure reference, thinking they just wanted to sound smarter than me or something.
"The best tools money can buy" would piss me off too, since I'd be hiring them to develop solely on open source software.
Also, your attitude toward unit testing is insulting toward people that do not use OOP. Our codebase does not use OOP, and I have no room for some OOP zealot that would try to migrate our codebase.
I think it would be safer to ask about regression testing, something that is applicable to any codebase.
In Rotterdam: How about Ajax eh?
Careful, that might get you hired to maintain a bunch of spaghetti Javascript code that is totally unmaintainable because they bought into the latest buzzwords.
The question is, who would you vote for?
I'd say if you are voting for a democrat or republican, you aren't accomplishing much.
A joule is one watt for one second.
The original poster's objection is correct, 10kw doesn't tell you anything about the amount of power it produces. Your math is bunk.
If it's 10kw it could be 1 joule for 1/10000 of a second.
That doesn't mean that Timmy doesn't want that XBox 360 more than anyone who has the money to lay out. It also doesn't mean that Timmy's parent's can afford to lay out $50 chunks here and there for games either.
Timmy isn't the consumer, his parents are.
They bought food and paid the electric bill right? It's simply that they wanted those things more than they wanted the X-Box.
dump fois gras, Chateau Latour, beluga caviar and a savoy truffle into a blender
:)
This is just like sesame street. Which one of these doesn't belong?
(Hint, Savoy Truffle isn't a fungus).
because some people were willing to pay a ridiculous amount for a 360 all of the consoles should have been sold at a ridiculous price.
He never said that. The higher prices would generate a relatively small amount of profit in the long run, yes. The point of them wouldn't be mainly to generate profit, it would be to ensure fair allocation of a limited resource. Making more profit would be a nice side effect.
Those willing to pay more would do so, and those people are the ones that wanted it the most. People that didn't want it as badly wouldn't pay as much. This is economics 101.
A hard-core fan would be happier paying $700 than not getting one at all. That's the point of economics, the people that want it the most will pay the most.
Higher prices ensure that only the hard-core fans get the console, and other people will just have to wait until the price drops to a point they are willing to pay.
That doesn't mean there wasn't a market for low-spec'd, expensive machines
I'm not so sure about that. The early PCs in the late 70s were only a few hundred bucks and bought by hobbyists.
I don't know how many hobbyists could have swung a lets say $3000 "PC" in 1972. Keep in mind a brand new car was about $1500 back then.
This article sucks.
Even on the first page, they act like all these companies were run by idiots, ignoring the possibility of a PC that was supposedly right under their noses.
It wasn't that the technology wasn't ready. Intel, at the time primarily a manufacturer of memory chips, had invented the first microprocessor (the 4-bit 4004) in 1971. This was followed up with the 8-bit 8008 in 1972 and the more-capable 8080 chip in 1974. However, Intel didn't see the potential of its own product, considering it to be useful mainly for calculators, traffic lights, and other embedded applications
That's because that's all it was good for. SMPS technology was in its infancy. Storage technology involved huge platters or huge tapes. RAM was damn expensive.
So what did they think Intel should have done? Released a "PC" in 1971 that weighed 200 pounds with a linear power supply, came with a mini-fridge sized persistant storage unit that held 100k, had 4k RAM and cost $20,000?
The technology indeed wasn't ready. The PC came when it did because technology allowed it to come, not because of lack of vision.
but with my data you only get one chance
:)
You will run out of manufacturers with that attitude. EVERY hard disk manufacturer has turned out batches of duds.
Here's another bit of news, every hard disk fails, it's only a question of when. 6 years is an excellent run, and 2 years isn't too bad, a little low unless you were thrashing it to hell 24/7.
We run dozens of disks in RAID configuration at work (used to be over 100 disks, see below), we only keep any particular set of disks in service 3 years max. We've found that the statistical failure rate starts rising sharply around 3 years for most models.
3 years also represents about the amount of time for disk space to double lately, so each time we upgrade a RAID we make it more reliable, because we can use about half as many disks.
At this pace we'll have all our storage on single-disk RAID 1 in about 9 years.
Anyway, I think individual users have so much "brand superstition" because of the slightly weird statistical distribution of failures with hard disks that can vary so much from model to model and production run to production run.
If I did, I doubt I'd stake my future on the availability of a web service from another company.
If you buy software and the parent company stops offering it or raises the price a lot, OK, no big deal, you can still run it and just not get upgrades.
If Google turns off maps or makes you pay a ton, or goes out of business, you are fucked if you used it for anything important.
In other words, only a fool would use these Google "applications" for anything important to their business.
Other people have posted useful things they have done with them for business, etc, but not one of those applications was anything important, they were more like the "cherry on top".
It's the same old "Application Service Provider" paradigm warmed over. No one wanted to pay then, and no one will want to pay now. And anyone smart won't do anything important with the free services.
I was responding to the asserting that they could be used as disposable product labels are used today, since that's what this article is about, not about some reusable label.
Google Earth, Google Maps, Google SMS, Google Blogsearch, they're all applications
They aren't applications, they are toys. Yeah they are neat to play with, but they aren't exactly something that businesses will use everyday for anything that matters.