>> Why bother ranking sites when bookmarking the results would achieve the same result (as well as be independent of a login).
We're not talking about the same result you get by bookmarking here, but about splitting the results into different sub-domains and having the search system automatically prioritize the sub-domain you search most often.
For example, if you are always searching for cars and always choose links to fuel-consumption statistics, it would make sense for you to have the statistics results among the first provided results, every time.
In this way, they're simply providing a better service (and probably intending to increase customer satisfaction).
In this case, neither geolocating nor logging is necessary: they could have simply looked at the phone numbers through the driver's phone, then asked the mobile company to tell them under which "Location Area Code" all the numbers in the list are located, at present (a LAC usually covers a small group of neighbouring cells in a network). Then it's down to routine police work ("find two wounded guys in this perimeter").
The location information is stored continuously by the network, in a database called HLR (Home Location Register) which has the purpose to decide through which cells to route call data, whenever somebody is trying to reach you (incoming call or SMS).
The alternative to the HLR would be that the network did a network-wide broadcast to locate you, every time you received a phone-call (which is not feasible).
I think you're (partially) missing the point: It still is illegal, the intention behind the law is ignored in lots of cases and you just compared "the land of the free" with a regime that was universally accepted as totalitarian (everything behind the iron curtain). That should tell you what "the land of the free" is on it's way to become.
Behind the iron curtain nobody was making much of a fuss about it, because those who did got in trouble, fast (the government was the biggest bully around and it didn't tolerate dissident voices).
Well... I've read what others wrote here and I don't think you got many actual answers (welcome to slashdot:) ).
While there were some (very) good points about both scalability and HA, they didn't tell you how to go about learning that; HA and HS are two areas where by reading the books or following case studies, you can understand the basic problems, but not see how you actually go about building a particularly scalable or HA system (because it's usually a system, not a single server).
I've worked in maintenance for a c++ server, where we gave the users guarantees of both low response time under stress and minimal down time.
While working with that, we've had to use different angles in attacking the appearing problems, and usually the solutions we used were particular for the problems couldn't be very well generalized.
For example, when different threads used the same input data, it was better in some situations to duplicate the data for each thread instead of using locks on it, so you didn't incur the delays involved in locking. In other places, we used locking as even with locking those places wouldn't create bottlenecks.
In some value-objects (large data collections), we used copy-on-write pattern with sharing objects values in the same thread (to minimize the allocations done and memory fragmentation), and deep copies when the data was needed to be sent in other threads.
We split the server in multiple different servers handling the different parts of processing (so we could throw hardware at the computation-heavy parts of the application logic) and use load-balancing.
We also used in-memory databases in one case, for storing some of the information.
Some IO operations (like logging) ran on different threads with messages passed to them (having multiple threads for logging for example).
For HA, we had a complete monitoring system, with processes listening for hartbeat and making load-statistics for the different modules, and every computation part had a backup server ready to take over at certain loads.
My point with these examples, is that neither of these solutions can be applied ad-hoc, but each possible bottleneck has to be studied separately and the solutions to avoid it can vary depending on a lot of factors (while you can get ideas, you won't learn scalability or HA from a book).
The best solution I can think of for learning about HS and HA is working in a product that needs them and getting direct experience in improving the scalability and availability.
And your legal dept. have a problem with that, why? Because you're relying on closed-source to make money instead of providing ongoing services for clients that they couldn't get otherwise get with just the sourcecode? Maybe you need a new businessmodel.
They're making over 2 billion dollars a year in profit (and have been for some years) and they're in the top 5 providers for their domain (though this position is self-proclaimed, so it might not be really that accurate). Considering their purpose is making money and not software (that is just a means of making money) I'd say their business model seems spot-on.
It may not be the moral ideal you (or I) are aspiring to, but good luck trying to explain that to the share-holders.
Free Software breeds more Free Software. It's not 'viral', as Microsoft likes to say - it's addictive.
That's not the same. Free software in itself is not 'viral'. The GNU (and similar licenses) could be viewed as viral though.
In working for a corporation, I've seen the "why not use open source?" question asked a few times, and the answers are the same everytime:
free open source is fine
stay away from GNU
I don't know, let us see the license .
We (for example) use the ACE framework with no problems, but the managers and the legal department around here are terrified that if GNU code entered our codebase someone would scream "show us the sourcecode!" for everything.
Because of this, I tend to agree with you that OSS is addictive, but can be viral for corporations if it comes with attached conditions (at least for corporations drawing their profit from closed source).
I don't mean to troll here, but "free with conditions" is not actually free after all (at least not for some).
>> Generally speaking hashes are very cpu and cache-inefficient beasts
Um... why you think hashes are inefficient? In a lot of languages (Perl, Python, Javascript, etc) the standard collection is the hash. In Javascript, even a simple array is a hash! Why you think it is inefficient?
Actually, the GP is right: hashing is can be an expensive operation, especially if you need collision-free hash values.
What you're referring to, is not hashing, but hash-bashed collections (hash-tables), where the indexing is done by hash-value. It's not the same thing: as far as algorithms go, hashing can be (but they're not always) expensive; as far as collections go, hashing is great for both read and write access. When you work with lots of data that's difficult to compare (xml or just strings usually) computing the hash only once and using it for comparing can indeed be a great optimization.
A week or so isn't going to be the end of the world. I'll wait for you guys to break your environments. I mean, if they are patching something - it HAS been broken all this time - since I installed the box. it didn't just break yesterday and then the patch came right out.
A week or so might be just enough. Even if something has been broken all this time, there is a big difference between your system having unpatched vulnerabilities and having unpatched and known vulnerabilities.
From your point of view, one might say that surely the latest patch is not the last one. If the system will continue having other (presently unknown) vulnerabilities, then patching one of them makes no difference, so, unless it's crashing, why patch at all?
I don't believe they will, for a simple reason: if they do that, their main revenue source (targeted advertising) is going to suffer. Let me expand on that:
If they start compromising their service, people will notice (It's enough for a couple of dozen websites publishing sought-for newsletters to tell you "don't use gmail as it classifies the newsletter you subscribed to as SPAM") and once people start to notice gmail starts being less desirable.
Once gmail starts being less desirable, google looses the targets of their targetted advertisement. In other words, asking for money to "uncensor" you (as a newsletter pubisher) should prove to be a long-term revenue strategy more reliable than targetted ads (And - as far as I can see - it doesn't look that way).
The problem is that it costs the US government money to destroy things, and costs the US government to rebuild them (US companies are not a part of the US government). Even if we pretend US companies and the US government are one entity, it's still a net loss. Look at it this way: imagine you are the US government, and you pay $1000 for your soldiers to destroy a building. Then, you pay $1000 for a US company to rebuild that building. That US company pays its employees $600 and pockets $400. Now, imagine that because you are "the US", you get to keep the profits from that US company (you don't get to, but pretend you do). That means that you paid-out $2000, and then put $400 back into your pocket. That's still a net loss of $1600. If you could get lots of foreign money for rebuilding, you might be able to turn a profit when you pay the US companies to rebuild - but there isn't much foreign money coming in. If the US was outright stealing the Iraqi oil, that's another way to make a profit, but the US is still paying for Iraqi oil. Besides, even if the US was outright stealing the oil, the profits from Iraqi oil still aren't enough to overturn the net loss.
Its a good argument, but it's missing one point: it is not "US money" to destroy things and "US money" to construct things, except in terms of it being spent and received in US Dollars. It's money US taxpayers (I guess) pay, which are then passing through someone's hands on contracts, to get things done. Somebody other than the taxpayers gets payed with any money the taxpayers pay, and it's not "the US" either. You cannot put a comparison sign between the money spent and money gained and speak of profit/loss for the US, simply because it's not gained and received by the same people.
On one hand, you will see the profiting parties arguing for war (*) - of course not for war's sake, nor for economical reasons, but for "war for piece", "war for stopping war" or "war on terror" (**) - and on the other hand you will see the people who actually see the profit made and realize it's an profiting war machine involved.
Actually, you will likely not see much of the second group at all, since it's "kind-of banned" in the US:(
__________________
(*) - Of course, it is not only profiting parties arguing for war, but other parties arguing for war play right into their hands.
(**) - It's been said many times (and attributed to Einstein I believe) that you cannot use the same energy that created a problem in attempting to solve it, but I guess that doesn't count when applying violence to stop violence.
One day, everybody will have a digital ID. You know, the kind used to digitally sign e-mail. If you had to digitally sign your request to create an account with a certificate issued from a trusted CA, then using a bot creates the potential of the user having his digital certificate revoked.
Well... maybe, but I think that's a long way in coming: First, this implies the existence of a global (or generally accepted) CA (certification authority), key issuer or something similar. Second, it involves massive changes for either HTTP/HTML (to make security closer to ubiquitous), the email protocols or at least the way we look at them and use them at present.
In short, I think its easier to evolve towards something alternative than to change/adopt said protocols to a large enough scale that *everyone* would have an assigned digital ID (maybe shared reputation systems?).
guess my whole point is that actually going into a test environment on a moments notice, donning your equipment and getting a first hand view of smoke/fire is a better experience, IMHO.
I don't think there is any theoretical training that can replace the hands-on experience. When I did the SCUBA Rescue diver course, we had to - first - get over the theoretical part (check for dangers, approach, evaluate, check for breathing etc.), and - second - go through with actual exercises.
I think all courses (should) cover at least these two parts, and I experienced hands-on what you're talking about: when my instructor went in 1m water and started splashing around and yelling "Help" I got blocked (with all the theoretical exercises I went through).
This program is related to theoretical exercises only though, so as far as I'm concerned, if it works, more power to them!:)
Interesting, those tactics look like spammer tactics. Maybe in the end the spammers will be a force for good?
Not a good idea.
Its enough for the major players to have a bad-enough label they can attach you (for the States for example, the terms "terrorist"/"terrorist simpatizer"/communist come to mind), so they can have the appearance of legality when censoring you. The appearance of legality matters a lot for governments that claim to represent the people.
By using spammer-like tactics, you'd just make it easier for them to place the "spammer" label on you (and easier for them to discredit your information by association which is an effective form of censorship in itself).
Second off, I apparently cannot spell "physicist" or the primary author's name, Calabrese. Guess I got caught up in the excitement of my first post ever:)
Yeah... it shows! You should have started your post with First p0st!
To the GP - I will feel able to criticise his lack of fluency in English the as soon as my Romanian is at a similar level. I can't even manage "I'd like a beer please" in Romanian. Can you?
That would be: "As vrea o bere va rog". (As vrea - I'd like, o bere - a beer, va rog - please).
Slashdot Mirror
>> Why bother ranking sites when bookmarking the results would achieve the same result (as well as be independent of a login).
We're not talking about the same result you get by bookmarking here, but about splitting the results into different sub-domains and having the search system automatically prioritize the sub-domain you search most often.
For example, if you are always searching for cars and always choose links to fuel-consumption statistics, it would make sense for you to have the statistics results among the first provided results, every time.
In this way, they're simply providing a better service (and probably intending to increase customer satisfaction).
In this case, neither geolocating nor logging is necessary: they could have simply looked at the phone numbers through the driver's phone, then asked the mobile company to tell them under which "Location Area Code" all the numbers in the list are located, at present (a LAC usually covers a small group of neighbouring cells in a network). Then it's down to routine police work ("find two wounded guys in this perimeter").
The location information is stored continuously by the network, in a database called HLR (Home Location Register) which has the purpose to decide through which cells to route call data, whenever somebody is trying to reach you (incoming call or SMS).
The alternative to the HLR would be that the network did a network-wide broadcast to locate you, every time you received a phone-call (which is not feasible).
So much for posting as AC :)
Posting as AC, as I have mod points.
I think you're (partially) missing the point: It still is illegal, the intention behind the law is ignored in lots of cases and you just compared "the land of the free" with a regime that was universally accepted as totalitarian (everything behind the iron curtain). That should tell you what "the land of the free" is on it's way to become.
Behind the iron curtain nobody was making much of a fuss about it, because those who did got in trouble, fast (the government was the biggest bully around and it didn't tolerate dissident voices).
You have chosen to cancel loading the libraries you will never use.
Are you sure you want to cancel?
Press OK to Cancel, or Cancel to Quit.
Well ... I've read what others wrote here and I don't think you got many actual answers (welcome to slashdot :) ).
While there were some (very) good points about both scalability and HA, they didn't tell you how to go about learning that; HA and HS are two areas where by reading the books or following case studies, you can understand the basic problems, but not see how you actually go about building a particularly scalable or HA system (because it's usually a system, not a single server).
I've worked in maintenance for a c++ server, where we gave the users guarantees of both low response time under stress and minimal down time.
While working with that, we've had to use different angles in attacking the appearing problems, and usually the solutions we used were particular for the problems couldn't be very well generalized.
For example, when different threads used the same input data, it was better in some situations to duplicate the data for each thread instead of using locks on it, so you didn't incur the delays involved in locking. In other places, we used locking as even with locking those places wouldn't create bottlenecks.
In some value-objects (large data collections), we used copy-on-write pattern with sharing objects values in the same thread (to minimize the allocations done and memory fragmentation), and deep copies when the data was needed to be sent in other threads.
We split the server in multiple different servers handling the different parts of processing (so we could throw hardware at the computation-heavy parts of the application logic) and use load-balancing.
We also used in-memory databases in one case, for storing some of the information.
Some IO operations (like logging) ran on different threads with messages passed to them (having multiple threads for logging for example).
For HA, we had a complete monitoring system, with processes listening for hartbeat and making load-statistics for the different modules, and every computation part had a backup server ready to take over at certain loads.
My point with these examples, is that neither of these solutions can be applied ad-hoc, but each possible bottleneck has to be studied separately and the solutions to avoid it can vary depending on a lot of factors (while you can get ideas, you won't learn scalability or HA from a book).
The best solution I can think of for learning about HS and HA is working in a product that needs them and getting direct experience in improving the scalability and availability.
They're making over 2 billion dollars a year in profit (and have been for some years) and they're in the top 5 providers for their domain (though this position is self-proclaimed, so it might not be really that accurate). Considering their purpose is making money and not software (that is just a means of making money) I'd say their business model seems spot-on.
It may not be the moral ideal you (or I) are aspiring to, but good luck trying to explain that to the share-holders.
That's not the same. Free software in itself is not 'viral'. The GNU (and similar licenses) could be viewed as viral though.
In working for a corporation, I've seen the "why not use open source?" question asked a few times, and the answers are the same everytime:
free open source is fine
stay away from GNU
I don't know, let us see the license .
We (for example) use the ACE framework with no problems, but the managers and the legal department around here are terrified that if GNU code entered our codebase someone would scream "show us the sourcecode!" for everything.
Because of this, I tend to agree with you that OSS is addictive, but can be viral for corporations if it comes with attached conditions (at least for corporations drawing their profit from closed source).
I don't mean to troll here, but "free with conditions" is not actually free after all (at least not for some).
... will a beowulf cluster of these run linux, or blend?
Actually, the GP is right: hashing is can be an expensive operation, especially if you need collision-free hash values.
What you're referring to, is not hashing, but hash-bashed collections (hash-tables), where the indexing is done by hash-value. It's not the same thing: as far as algorithms go, hashing can be (but they're not always) expensive; as far as collections go, hashing is great for both read and write access. When you work with lots of data that's difficult to compare (xml or just strings usually) computing the hash only once and using it for comparing can indeed be a great optimization.
UINTW (USA Is Not The World)
A week or so might be just enough. Even if something has been broken all this time, there is a big difference between your system having unpatched vulnerabilities and having unpatched and known vulnerabilities.
From your point of view, one might say that surely the latest patch is not the last one. If the system will continue having other (presently unknown) vulnerabilities, then patching one of them makes no difference, so, unless it's crashing, why patch at all?
Aha! Any second now your system will be shutti
Ummm ... No.
I don't believe they will, for a simple reason: if they do that, their main revenue source (targeted advertising) is going to suffer. Let me expand on that:
If they start compromising their service, people will notice (It's enough for a couple of dozen websites publishing sought-for newsletters to tell you "don't use gmail as it classifies the newsletter you subscribed to as SPAM") and once people start to notice gmail starts being less desirable.
Once gmail starts being less desirable, google looses the targets of their targetted advertisement. In other words, asking for money to "uncensor" you (as a newsletter pubisher) should prove to be a long-term revenue strategy more reliable than targetted ads (And - as far as I can see - it doesn't look that way).
True. Hell, We'd be pretty amazingly lucky to have intelligent live on earth!
Unfortunately, probably the BSA will...
Its a good argument, but it's missing one point: it is not "US money" to destroy things and "US money" to construct things, except in terms of it being spent and received in US Dollars. It's money US taxpayers (I guess) pay, which are then passing through someone's hands on contracts, to get things done. Somebody other than the taxpayers gets payed with any money the taxpayers pay, and it's not "the US" either. You cannot put a comparison sign between the money spent and money gained and speak of profit/loss for the US, simply because it's not gained and received by the same people.
The more you continue to keep this model, the more you create a niche market for war products and services, an interest in waging war for that money (among other things). At the moment, in some circles, there is the feeling that this market has specialized in a veritable "war machine" that's highly profitable.
On one hand, you will see the profiting parties arguing for war (*) - of course not for war's sake, nor for economical reasons, but for "war for piece", "war for stopping war" or "war on terror" (**) - and on the other hand you will see the people who actually see the profit made and realize it's an profiting war machine involved.
Actually, you will likely not see much of the second group at all, since it's "kind-of banned" in the US :(
__________________
(*) - Of course, it is not only profiting parties arguing for war, but other parties arguing for war play right into their hands.
(**) - It's been said many times (and attributed to Einstein I believe) that you cannot use the same energy that created a problem in attempting to solve it, but I guess that doesn't count when applying violence to stop violence.
Let me fix that for you:
I'd like that also :)
Exercise: Point to a period in human history devoid of oppression/abuse/war longer than, say ... 100 years?
Well ... maybe, but I think that's a long way in coming: First, this implies the existence of a global (or generally accepted) CA (certification authority), key issuer or something similar. Second, it involves massive changes for either HTTP/HTML (to make security closer to ubiquitous), the email protocols or at least the way we look at them and use them at present.
In short, I think its easier to evolve towards something alternative than to change/adopt said protocols to a large enough scale that *everyone* would have an assigned digital ID (maybe shared reputation systems?).
I don't think there is any theoretical training that can replace the hands-on experience. When I did the SCUBA Rescue diver course, we had to - first - get over the theoretical part (check for dangers, approach, evaluate, check for breathing etc.), and - second - go through with actual exercises.
I think all courses (should) cover at least these two parts, and I experienced hands-on what you're talking about: when my instructor went in 1m water and started splashing around and yelling "Help" I got blocked (with all the theoretical exercises I went through).
This program is related to theoretical exercises only though, so as far as I'm concerned, if it works, more power to them! :)
Not a good idea.
Its enough for the major players to have a bad-enough label they can attach you (for the States for example, the terms "terrorist"/"terrorist simpatizer"/communist come to mind), so they can have the appearance of legality when censoring you. The appearance of legality matters a lot for governments that claim to represent the people.
By using spammer-like tactics, you'd just make it easier for them to place the "spammer" label on you (and easier for them to discredit your information by association which is an effective form of censorship in itself).
Because I still have nightmares about that cat ...
Yeah ... it shows! You should have started your post with First p0st!
That would be: "As vrea o bere va rog". (As vrea - I'd like, o bere - a beer, va rog - please).