Valve made the original half life, half-life 2 and the episodes. But all the expansions to the original half life (opposing force, blue shift and decay) were made by gearbox software.
At the user code level ARM is somewhat standardised. The core instruction set is standardised (though comes in several different versions each adding extra instructions). There are serveral different FPUs but afaict most current phones are using the same one.
At the OS level things get a lot messier. On a PC there is a LOT of standard hardware in standard places and the non-standard hardware tends to be on a PCI bus (with a standardised method for accessing PCI configuration space) which makes it easy to enumerate. There is also a bios which knows how to initialise the hardware, read the code from the boot sector on a drive and run it while providing it with services that let it easily access the keyboard, mouse, screen and hdd. On an arm system you might have a PCI bus but you might not, the hardware may be mapped anywhere. ARM linux distros often have multiple kernels for different hardware and the kernels they do have have many special cases in them.
Electronics are designed to run at temperatures well above normal outdoor air tempreature in most places and so an efficiant cooling system should be able to cool them with nothing more than moving the working fluid arround yet we still end up actively pumping heat using phase change refridgeration to cool air below ambiant. To me this sounds like a MASSIVE WTF.
Sorry that should have said it is at best as secure as the LEAST secure CA. In practice it is likely to be worse than the worst individual CA because different CAs may be susceptible to different hackers/blackmailers/etc.
There is a fundamental problem in cryptography that users don't want to remember keys. So cryptosystems become only as secure as the master server that checks the users login and manages the mapping from user accounts to temporary keys that are generated for each session . Afaict skype falls into this category and as such it would be pretty trivial for the owners of skype to MITM their users communications.
If you really want security you need to use a cryptosystem where you manage the keys yourself and take appropriate precauations in storing and managing those keys. Most peope though don't want to go to that effort.
SSL as typically implemented has even worse problems it is only as secure as the LEAST secure CA that the browser trusts (either directly or through delegation).
TFA says the hard drives have to be sprayed with a coating, presumably to make the housing oil-tight as well as airtight.
This should IMMEDIATELY ring alarm bells. Hard drives are NOT airtight. They have a filtered air hole. They would never get away with such flimsy construction on an airtight product.
Plus does this system REALLY offer that much advantage over conventional "waterblocks" which keep the cooling fluid seperate from the electronics. I very much doubt it. The major heat generators in a PC are designed to pass out their heat by contact conduction anyway.
Windows people are gonna expect Windows apps to work which of course they most assuredly WON'T, not without a recompile that most companies simply won't do.
I'm pretty sure MS can make them run, dynamic recompilation is hardly a new concept.
The question is can they make them run with tolerable performance. Apple got away with using this method to migrate from PPC to x86 but afaict they only got away with it because the x86 boxes were much faster than the PPC boxes they were replacing. ARM boxes tend to be slow in comparison to x86 boxes.
Or are you suggesting that low-income people are just too stupid to read a bill, and can't help but be hoodwinked?
Either that or they have bad credit ratings (Either through legitimate misfortune, lack of discipline, location of residense or some combination of the preceeding).
Why do you think there is a market for payday loans with some people using them every month?
With Usage Based Billing, I will always be charged the same amount.
Most usage based plans do charge a single rate for simplicity but there is no reason all usage based plans have to be like that. Just as phone and electricity tarrifs can have an off peak rate there is no reason not to have them for internet plans too.
I don't see how charging a data user based on data volume is any different from charging a phone user based on time spent on the phone. In both cases you have an imperfect but practical system for splitting the charges for a communication resource among the users of that resource in rough proportion to the ammount of that communication resource they are using.
In the past we have always paid for internet via the size of the pipe, not how much data comes through.
With consumer broadband what you are really paying for is the ability to use part of a very large pipe with mobile internet which pipe you are using part of is even changing continuously.
For a while ISPs just charged all home users with the same speed of final connection to the same ammount, buisness users paid a bit more. This worked ok for a while sure some users used more of it than others but it was within what the ISPs had planned their networks to deal with and TCP dealt reasonably well with congestion. Then bittorrrent came along and became the bane of the ISPs life by simultaneously allowing users to maintain high data rates continuously AND using large numbers of connections at once which meant it didn't play very nice when congestion did happen. The ISPs had three choices
1: spend a lot of money on building infrastructure to satisfy the small proportion of customers who are heavy torrenters 2: kick the heavy torrenters off the network 3: resort to dirty tricks to try and deprioritise, throttle or even outright block bittorrent 4: move to a charging model that makes the heavy torrenters pay in proportion to their higher usage thereby either stopping them from being heavy torrenters, driving them off the ISP or bringing in more money to spend on infrastructure.
It's very difficult to measure quality of a product before purchase so we have to rely on reputation.
Unfortunately this means the benifit to a company of reducting cost/quality is immediate and the cost is years down the line. This makes it very tempting for companies to reduce quality either to get through bad times or just plain to make their figures look good this quarter so the CEO gets his bonus. Further big retailers (who don't really care about their suppliers long term) are constantly trying to push suppliers to reduce prices.
I would add that the wholesale providers must not be allowed to discriminate based on the type of device the end user is using. Only based on the total number of devices and total usage of each type of service (voice, data, sms etc). Ideally they should also provide a way for retail providers who use multiple wholesale providers to easilly change their "preffered" wholesale provider.
Ideally one would also get rid of the technical compatibility problems in the US phone market (though those may be going away anyway with the move to LTE).
Yes the syntax and rules are new but all the wordy parts of the language are generally based on english. I wonder how well you would do at following a C program where all the identifiers and and comments were in a foreign language?
One strong point of modern college is that language classes have be depreciated for fields they have no bearing. A robotics or CS major will have zero use for latin or greek or really any language other than english.
I'd agree on latin and greek but i'd think some modern foreign languages could be quite useful if only to increase the variety of places one could potentially work/study. While I know some institutions in other countries will let you take a PHD without speaking their language (becuase I saw a poster from a greek university advertising it) I wouldn't imagine doing so is much fun. It's bad enough being in a british university and being able to understand what the chinese are talking to each other about.
Actually 10 year old computers are probablly the worst in many ways. Older computers are actually better.
10 years old would put a computer arround the time when capacitor plauge was coming in. It would also put it arround the time when power consuption was getting high enough to seriously burn stuff up but before AMD and it's partners had got a handle on how to stop stuff burning up (intel to their credit got a handle on this much earlier).
Even if it still 'works', every day it is in your office increases your risk of a business-destroying fire.
Do you have any stats to back that up or is it just conjecture. I'd think the combination of flame retardent PCB substrates (i'm pretty sure these were in place by 10 years ago, please correct me if i'm wrong) and metal cases should make the risk of a fire spreading out of the PC pretty damn low.
Co-worker spilled an entire litre of water directly onto the keyboard of a company owned running laptop. Hell to pay if laptop ruined. Unplug, pull out battery, pull disk drive, dry face-down on cookie rack for 24 hours.
So you spilt plain water on it and then did all the right things to maximise your chance of survival, no wonder it survived.
If I wanted to (not that I would) fuck up a computer by "accidently" spilling a drink on it I wouldn't chose plain water. I'd chose something sticky and acidic like coca-cola. I would also pretend not to know how best to deal with the situation.
If you do end up spilling something like coca-cola, fruit juice, coffee etc into an electronic device I'd reccomend rinsing it under the tap ASAP (after removing power/batteries/hdds etc). Distilled water would probablly be better than tap water but isn't likely to be available quickly and in sufficiant quantities. Time is of the essense because once the drink dries into a sticky mess it is likely to be much harder to rinse off.
Not sure why you IT types bitch about this so much.
Because unfortunately the two are mutally exclusive. Especially if IT are overworked (as they often are because they don't directly generate revenue) and the users are incompetant with computers.
and can't tell you how often we get in piece of equipment where the software that drives the thing requires admin rights for no apparent reason. Everything from disc duplicators, to a pneumatic driven matt cutter. Utter nonsense!
What the hardware vendor should really do is allow permissions to use the hardware to be allocated to users but afaict that is a lot of work so unfortunately most drivers end up defaulting to either wide open or admins only neither of which is ideal.
To make matters worse our IT dept. has a no admin rights policy in effect, so those machines that require admin rights have to run with a script that uses the help desk login no matter who is logged onto the machine.
The thing with locking things down is it works provided one of two assumptions are met
1: everyone is doing the same thing with the same set of software 2: IT (or whoever is entrusted with admin privilages) are responsive and deal with requests to install software or change other configurations quickly and efficiantly.
If neither of those are the case then it can be hell (i've seen this recently where the local guy in our research group who had admin on a bunch of machines left and the guy from IT he passed the credentials on to is very difficult to get hold of). Personally I support a benefit of the doubt policy, let people have control over their machines the first time but be prepared to start locking things down for people who break the rules and/or are incapable of keeping their machines clean.
Having said that while I do like method pointers and there are certainly cases where it would be nice to be able to arbiterally wire up events and/or rewire them on the fly most of the time an inner class descending from an adaptor class is actually a pretty neat method of event handing.
Personally I found java's lack of properties* and it's lack of an official GUI builder (I belive there is one that is at least semi-official now but I don't think it was arround back when I was coding in JAVA and it's still tied to a particualr IDE that is different from the one I use) far more painful than it's lack of method pointers.
As I see it both sides had a lot more to lose in this case than they had to gain.
If sony won the lawsuit then they would just drive the next guy further underground yeah that is a victory for them but most people with a sense of self preservation undertaking such activities are likely to be somewhat underground about it anyway. If they lose then the case would essentially legitmise such hacking attempts.
If geohot lost the lawsuit then he would likely be bankrupted, lose most of his possesions and have financial problems for however long it takes a bankrupcy to come off your record. If he won then he would be winning a huge victory for the hacking community but it wouldn't personally gain him much. There are plenty of other products to hack on that don't have such litigous companies behind them.
Before he could read out the key he had to pick apart the rom to find out where it was stored and how to decode it that sounds like reverse engineering to me. Further it seems pretty clear he did it for the purposes of interoperability.
Of course just because it's reverse engineering doesn't nesacerally make it legal. In particular apple could try to invoke the DMCA by claiming what was cracked was a copy protection scheme.
Hardly any electricity comes from oil at least in the west (developing countries that are lacking in infrastructure may be a different matter), afaict the main sources of electricity are coal, nuclear, natural gas and hydro.
The first thing I note is that while the title says Europe the proposal talks cooperation stretching far beyond the stable democratic area of europe into sibera at one end and into the sahara at the other.
Would the EU countries really be willing to rely to an even greater extent than they do now (gas can at least be stored for a while building up a buffer and can also be shipped in by sea) on stability in north africa and the former soviet states? Especially given the current situation in north africa (libya is obviously the worst but it remains to be seen what the political landscape in all those countries will be like if and when democracy rises there).
Slashdot Mirror
Valve made the original half life, half-life 2 and the episodes. But all the expansions to the original half life (opposing force, blue shift and decay) were made by gearbox software.
At the user code level ARM is somewhat standardised. The core instruction set is standardised (though comes in several different versions each adding extra instructions). There are serveral different FPUs but afaict most current phones are using the same one.
At the OS level things get a lot messier. On a PC there is a LOT of standard hardware in standard places and the non-standard hardware tends to be on a PCI bus (with a standardised method for accessing PCI configuration space) which makes it easy to enumerate. There is also a bios which knows how to initialise the hardware, read the code from the boot sector on a drive and run it while providing it with services that let it easily access the keyboard, mouse, screen and hdd. On an arm system you might have a PCI bus but you might not, the hardware may be mapped anywhere. ARM linux distros often have multiple kernels for different hardware and the kernels they do have have many special cases in them.
Electronics are designed to run at temperatures well above normal outdoor air tempreature in most places and so an efficiant cooling system should be able to cool them with nothing more than moving the working fluid arround yet we still end up actively pumping heat using phase change refridgeration to cool air below ambiant. To me this sounds like a MASSIVE WTF.
it is only as secure as the LEAST secure CA
Sorry that should have said it is at best as secure as the LEAST secure CA. In practice it is likely to be worse than the worst individual CA because different CAs may be susceptible to different hackers/blackmailers/etc.
There is a fundamental problem in cryptography that users don't want to remember keys. So cryptosystems become only as secure as the master server that checks the users login and manages the mapping from user accounts to temporary keys that are generated for each session . Afaict skype falls into this category and as such it would be pretty trivial for the owners of skype to MITM their users communications.
If you really want security you need to use a cryptosystem where you manage the keys yourself and take appropriate precauations in storing and managing those keys. Most peope though don't want to go to that effort.
SSL as typically implemented has even worse problems it is only as secure as the LEAST secure CA that the browser trusts (either directly or through delegation).
TFA says the hard drives have to be sprayed with a coating, presumably to make the housing oil-tight as well as airtight.
This should IMMEDIATELY ring alarm bells. Hard drives are NOT airtight. They have a filtered air hole. They would never get away with such flimsy construction on an airtight product.
Plus does this system REALLY offer that much advantage over conventional "waterblocks" which keep the cooling fluid seperate from the electronics. I very much doubt it. The major heat generators in a PC are designed to pass out their heat by contact conduction anyway.
Windows people are gonna expect Windows apps to work which of course they most assuredly WON'T, not without a recompile that most companies simply won't do.
I'm pretty sure MS can make them run, dynamic recompilation is hardly a new concept.
The question is can they make them run with tolerable performance. Apple got away with using this method to migrate from PPC to x86 but afaict they only got away with it because the x86 boxes were much faster than the PPC boxes they were replacing. ARM boxes tend to be slow in comparison to x86 boxes.
Or are you suggesting that low-income people are just too stupid to read a bill, and can't help but be hoodwinked?
Either that or they have bad credit ratings (Either through legitimate misfortune, lack of discipline, location of residense or some combination of the preceeding).
Why do you think there is a market for payday loans with some people using them every month?
With Usage Based Billing, I will always be charged the same amount.
Most usage based plans do charge a single rate for simplicity but there is no reason all usage based plans have to be like that. Just as phone and electricity tarrifs can have an off peak rate there is no reason not to have them for internet plans too.
I don't see how charging a data user based on data volume is any different from charging a phone user based on time spent on the phone. In both cases you have an imperfect but practical system for splitting the charges for a communication resource among the users of that resource in rough proportion to the ammount of that communication resource they are using.
In the past we have always paid for internet via the size of the pipe, not how much data comes through.
With consumer broadband what you are really paying for is the ability to use part of a very large pipe with mobile internet which pipe you are using part of is even changing continuously.
For a while ISPs just charged all home users with the same speed of final connection to the same ammount, buisness users paid a bit more. This worked ok for a while sure some users used more of it than others but it was within what the ISPs had planned their networks to deal with and TCP dealt reasonably well with congestion. Then bittorrrent came along and became the bane of the ISPs life by simultaneously allowing users to maintain high data rates continuously AND using large numbers of connections at once which meant it didn't play very nice when congestion did happen. The ISPs had three choices
1: spend a lot of money on building infrastructure to satisfy the small proportion of customers who are heavy torrenters
2: kick the heavy torrenters off the network
3: resort to dirty tricks to try and deprioritise, throttle or even outright block bittorrent
4: move to a charging model that makes the heavy torrenters pay in proportion to their higher usage thereby either stopping them from being heavy torrenters, driving them off the ISP or bringing in more money to spend on infrastructure.
It's very difficult to measure quality of a product before purchase so we have to rely on reputation.
Unfortunately this means the benifit to a company of reducting cost/quality is immediate and the cost is years down the line. This makes it very tempting for companies to reduce quality either to get through bad times or just plain to make their figures look good this quarter so the CEO gets his bonus. Further big retailers (who don't really care about their suppliers long term) are constantly trying to push suppliers to reduce prices.
I know of the "sam vimes boot" theory but I must have missed the "ramkin extension" care to fill me on on what that was.
I would add that the wholesale providers must not be allowed to discriminate based on the type of device the end user is using. Only based on the total number of devices and total usage of each type of service (voice, data, sms etc). Ideally they should also provide a way for retail providers who use multiple wholesale providers to easilly change their "preffered" wholesale provider.
Ideally one would also get rid of the technical compatibility problems in the US phone market (though those may be going away anyway with the move to LTE).
Not even Java, or C++ or....
Yes the syntax and rules are new but all the wordy parts of the language are generally based on english. I wonder how well you would do at following a C program where all the identifiers and and comments were in a foreign language?
One strong point of modern college is that language classes have be depreciated for fields they have no bearing. A robotics or CS major will have zero use for latin or greek or really any language other than english.
I'd agree on latin and greek but i'd think some modern foreign languages could be quite useful if only to increase the variety of places one could potentially work/study. While I know some institutions in other countries will let you take a PHD without speaking their language (becuase I saw a poster from a greek university advertising it) I wouldn't imagine doing so is much fun. It's bad enough being in a british university and being able to understand what the chinese are talking to each other about.
Actually 10 year old computers are probablly the worst in many ways. Older computers are actually better.
10 years old would put a computer arround the time when capacitor plauge was coming in. It would also put it arround the time when power consuption was getting high enough to seriously burn stuff up but before AMD and it's partners had got a handle on how to stop stuff burning up (intel to their credit got a handle on this much earlier).
Even if it still 'works', every day it is in your office increases your risk of a business-destroying fire.
Do you have any stats to back that up or is it just conjecture. I'd think the combination of flame retardent PCB substrates (i'm pretty sure these were in place by 10 years ago, please correct me if i'm wrong) and metal cases should make the risk of a fire spreading out of the PC pretty damn low.
Co-worker spilled an entire litre of water directly onto the keyboard of a company owned running laptop. Hell to pay if laptop ruined. Unplug, pull out battery, pull disk drive, dry face-down on cookie rack for 24 hours.
So you spilt plain water on it and then did all the right things to maximise your chance of survival, no wonder it survived.
If I wanted to (not that I would) fuck up a computer by "accidently" spilling a drink on it I wouldn't chose plain water. I'd chose something sticky and acidic like coca-cola. I would also pretend not to know how best to deal with the situation.
If you do end up spilling something like coca-cola, fruit juice, coffee etc into an electronic device I'd reccomend rinsing it under the tap ASAP (after removing power/batteries/hdds etc). Distilled water would probablly be better than tap water but isn't likely to be available quickly and in sufficiant quantities. Time is of the essense because once the drink dries into a sticky mess it is likely to be much harder to rinse off.
Not sure why you IT types bitch about this so much.
Because unfortunately the two are mutally exclusive. Especially if IT are overworked (as they often are because they don't directly generate revenue) and the users are incompetant with computers.
and can't tell you how often we get in piece of equipment where the software that drives the thing requires admin rights for no apparent reason. Everything from disc duplicators, to a pneumatic driven matt cutter. Utter nonsense!
What the hardware vendor should really do is allow permissions to use the hardware to be allocated to users but afaict that is a lot of work so unfortunately most drivers end up defaulting to either wide open or admins only neither of which is ideal.
To make matters worse our IT dept. has a no admin rights policy in effect, so those machines that require admin rights have to run with a script that uses the help desk login no matter who is logged onto the machine.
That is just plain fucked up.
The thing with locking things down is it works provided one of two assumptions are met
1: everyone is doing the same thing with the same set of software
2: IT (or whoever is entrusted with admin privilages) are responsive and deal with requests to install software or change other configurations quickly and efficiantly.
If neither of those are the case then it can be hell (i've seen this recently where the local guy in our research group who had admin on a bunch of machines left and the guy from IT he passed the credentials on to is very difficult to get hold of). Personally I support a benefit of the doubt policy, let people have control over their machines the first time but be prepared to start locking things down for people who break the rules and/or are incapable of keeping their machines clean.
Having said that while I do like method pointers and there are certainly cases where it would be nice to be able to arbiterally wire up events and/or rewire them on the fly most of the time an inner class descending from an adaptor class is actually a pretty neat method of event handing.
Personally I found java's lack of properties* and it's lack of an official GUI builder (I belive there is one that is at least semi-official now but I don't think it was arround back when I was coding in JAVA and it's still tied to a particualr IDE that is different from the one I use) far more painful than it's lack of method pointers.
*No what javabeans calls properties don't count.
As I see it both sides had a lot more to lose in this case than they had to gain.
If sony won the lawsuit then they would just drive the next guy further underground yeah that is a victory for them but most people with a sense of self preservation undertaking such activities are likely to be somewhat underground about it anyway. If they lose then the case would essentially legitmise such hacking attempts.
If geohot lost the lawsuit then he would likely be bankrupted, lose most of his possesions and have financial problems for however long it takes a bankrupcy to come off your record. If he won then he would be winning a huge victory for the hacking community but it wouldn't personally gain him much. There are plenty of other products to hack on that don't have such litigous companies behind them.
Before he could read out the key he had to pick apart the rom to find out where it was stored and how to decode it that sounds like reverse engineering to me. Further it seems pretty clear he did it for the purposes of interoperability.
Of course just because it's reverse engineering doesn't nesacerally make it legal. In particular apple could try to invoke the DMCA by claiming what was cracked was a copy protection scheme.
9.5ML is an SI unit
No it isn't, it's a "non SI unit accepted for use with SI". The SI equivalent of the megalitre is the cubic decametre.
Hardly any electricity comes from oil at least in the west (developing countries that are lacking in infrastructure may be a different matter), afaict the main sources of electricity are coal, nuclear, natural gas and hydro.
Oil is mostly used to make transport fuel.
The first thing I note is that while the title says Europe the proposal talks cooperation stretching far beyond the stable democratic area of europe into sibera at one end and into the sahara at the other.
Would the EU countries really be willing to rely to an even greater extent than they do now (gas can at least be stored for a while building up a buffer and can also be shipped in by sea) on stability in north africa and the former soviet states? Especially given the current situation in north africa (libya is obviously the worst but it remains to be seen what the political landscape in all those countries will be like if and when democracy rises there).