Unfortunately you've got your own idea of billion/trillion so that just makes it worse.
Alternatively, maybe it's you who has your own idea of billion/trillion:)
Being serious though, most people in the UK tend to use the US definitions of billion/trillion these days because it fits in better with the SI system - mega/giga/tera == million/billion/trillion
Broadcast is a stupid model for delivery in a world where you can just buy what you want, when you want. Even if you want to really "broadcast" something because you want people to see it live, multicast is a nice replacement.
I'm sorry, but multicast can't be used for "on demand" stuff because people want to start watching the stream at different times. It's only useful for live stuff (the same as broadcast infact). True on-demand services have to be done over unicast and we're only just starting to get close to the sorts of bandwidth we need (I'd love to know how many simultaneous viewers NTL's TV on demand thing supports - I'm guessing it's not that many). And unless you're receiving your data via cable, you're getting it over a broadcast medium anyway even if you encapsulate it in a multicast protocol, so no benefit here.
we are rapidly approaching critical mass of TV content online.
If everyone in the world was fetching their TV over the Internet using current technologies, your connection would quickly grind to a halt (and the average person wants their TV _now_ - pressing a button and waiting 6 hours for it do download isn't the experience they expect).
The problem could be eased by modifying P2P technology so it tries to only fetch blocks from people local to you where possible - i.e. if you can fetch a block from someone on your ISP then that saves congesting the peering between the ISPs.
but has anyone noticed how appalling modern (UK) cable boxes are? all the ones I've seen have had a second or two lag whenever you change channel
Not much you can do about this - when you change channel the box has to wait until a keyframe arrives before it can start rendering the video. Also, because you get a bit reserviour (to allow the signal to be encoded in a slightly VBR-like way == improved quality) you may need to buffer a bit of the data before you start encoding since the encoder won't be using a strictly constant bit rate but the input will be at a constant rate.
I guess the receiver could buffer the entire multiplex stream so that switching between channels in the same multiplex could be done quicker but without multiple tuners you can't improve the switching speed between channels in different multiplexes. The only possibility I can think of here is to have 2 tuners and have the decoder try and make intelligent guesses about which channel you're likley to change to next.
could you imagine expose-style simultaneous viewing of all your favourite channels
Assuming you had a meaty enough decoder you could decode all the channels in the current multiplex simultaneously. A possiblity is to provide a multiplex with scaled down low quality versions of all the channels in it so your decoder can decode the whole lot and display it at once.
and it would be nice to have full Tivo-like capabilities from my PC
Can I take this opportunity to point at MythTV. I've been using it for several years and it's still better than the systems the sat and cable operators are providing here in the UK. I was stuck using my parent's NTL Digital system at Christmas and realised just how much functionality I take for granted in Myth which just isn't there in the NTL system. Similarly I hear friends commenting about features they wished Sky+ had and they're always things that Myth has already been doing for ages.
Admittedly there are a few bugs that neet to be mopped up in Myth but on the whole it's a good project. I think the only thing I'd really like to be able to do that I can't already do is use a satellite tuner in my myth box instead of re-encoding the S-Video output from my Sky box. (Yes, Myth does support DVB-S cards, but getting a DVB-S card with a common interface is very pricey still and since Sky won't produce a VideoGuard CAM I'd be stuck using an illegal programmable CAM such as the DragonCAM - total cost, £100 for the card, £50 for the common interface and £80 for the CAM == way too pricey)
Channel numbers are obsolete, anyway, or at least they ought to be. Let me pick my channels from an onscreen menu. If I have favorites, let me memorize the shortcuts on my own terms, not as the arbitrary, unchangeable numbers that come from my cable provider.
Well, yes and no - I have Sky and the fact that every so often they change all the channel numbers without warning is a pain in the arse (suddenly my Myth box is changing to the wrong channel, etc). However, channel numbers are better than user defined shortcuts for 2 reasons:
1. Can I be bothered to set up short cuts for my favorite channels? Nope... my "favorites" list is infact empty, I've never got around to settin it up. A channel number is already there and I just have to remember it. 2. If I use someone else's Sky decoder then I can get to my favorite channels just as easilly - my "shortcuts" list is defined in my brain and so is portable, a shortcuts list on my decoder stays on my decoder. (Similarly, my favorite stations may not be the same as other people's in the same house so suddenly we need to allow for the definition of even more shortcuts to cope with multiple people).
Which brings me to a side point - maybe we need some central organisation handing out channel numbers so that all the cable/sat operators use the same numbers, kind of like telephone numbers. I don't have to dial a different phone number to talk to my mum if I dial from a different telco, why should I have to use a different channel number because my TV is provided by a different company?
I've often wondered why a truck license isn't required for commercial vehicles, since they fall under the same section of legislation for vehicle safety.
Well anyone (here in the UK) who got their driving licence before some time around 1997 (me included) is allowed to drive light trucks anyway... (up to 7.5 Tons ISTR)
While it is true more people have the opertunity to go through the source and check for problems, how many actually do? Did you read throught the source for any of the open source software you installed ? I certainly don't have the time and in many cases enough knowledge of the language and problem to properly review the code.
No, people don't generally review the source of everything they install. But the point is that there are more people fiddling with the code and therefore more people to spot bugs. On several occasions I have been hacking something into the Linux kernel and spotted bugs in existing code (fairly core stuff like netfilter, netdev, etc). When I spot these bugs I fix them where I can (and submit patches) and if I can't fix them myself I work with the relevent developers to help them develop a fix. As a result of this, a number of fairly serious core kernel bugs have been fixed as a result of me happening to spot them while hacking at something related - the same couldn't be said about Windows bugs (and infact because Windows is closed source I probably wouldn't have been able to hack up soemthing to do what I wanted anyway).
I never been that convinced by the Cathedral vs the Bazaar argument. Given fewer people use linux and of those users even less will have enough knowledge, time and expertise to bug hunt in any meaningful way is having the source open that much of an advantage?
Well, I wouldn't call most FOSS development a "Bazaar" - people don't just randomly apply code to projects, you submit stuff to a maintainer who will than decide whether or not to apply it (or indeed apply a re-coded version of your patch).
However, I don't think you should think about the percentage of users who will see the code - you want to think about the total number of people who will see the code and I'm willing to bet that there are a lot more people working on and reviewing any piece of core Linux code than Windows code.
5th Gear did a test a while back on 4x4's in near-miss conditions. They took a couple of normal cars, drove them along at motorway speeds and simulated a motorway incident with the front car slamming on it's brakes and the back car having to swerve to avoid it. Nomatter what they did they could only make the car spin, not roll over. They then re-ran the test but this time at only 40mph and with a 4x4 as the second car - it rolled right over as soon as the driver swerved to avoid the stopped car. Certainly very eye opening - the driver of the 4x4 would definately not be in very good shape if he wasn't wearing a racing harness, etc.
Increases in that sort of safety may decrease insurance premiums
Or it may increase them. More survivors of road accidents == more people alive to claim long term personal injury payments. (No, I'm not advocating killing people to cut insurance, but it _is_ a fact that in some cases it's cheaper for an insurance company to pay out a large sum to the familly of someone who was killed rather than paying out for life long treatment of an injury).
The point is that if all that extra weight is put into engineering stronger cars that absorb the impact and slow you down more gradually then you're going to be more likely to survive a crash.
Well, lets assume that both vehicles involved in a collision are of identical design... actually, lets make it even easier - you're crashing a car into a solid object. If you've got a 1 metre crumple-zone at the front of the car, it means that you _will_ accellerate from whatever speed you were doing to 0mph within the distance of 1 metre. It doesn't matter how heavy or strong the car is, if you're doing 30mph and you hit something solid then you're going to go from 30mph to 0mph in the space of 1 metre. The only thing you can do to improve the performance in a crash is to make sure the accelleration is constant over the length of the whole crumple-zone. I.e. if you're doing 30mph, hit something solid and 90cm later you're doing 20mph then your problem is that you have to lose that 20mph in a distance of only 10cm. Ideally the crumple-zone of a car would be made out of some kind of smart material that changed it's strength depending on the size of an impact, so a 30mph and a 60mph collision would both result in the whole crumple zone being used and the energy distributed over the whole length of it.
Now, going back to having 2 cars colliding, this is only different to hitting something solid if there is an imbalance in the equation - i.e. one of the cars is slower, lighter or weaker. So assuming you already have a perfect crumple-zone, the only way you can engineer the car to dissipate it's energy over a longer period of time (and therefore make the crash more survivable) is at the expense of the other car in the crash - e.g. make your car massively heavier than the car you're going to hit.
Is a major software manufacturer, or in the mentioned case of Linux, OS type *REQUIRED* to provide support for a patented technology? Especially without royalties being paid?
IMHO, while patents may be acceptable on how a medium works, they shouldn't (be allowed to) cover how the medium is _read_ (actually, I don't really believe in patents at all but...). Although my original complaint wasn't actually intended to be about patents - whilest it may be illegal to play DVDs under Linux because of a patent, it's also illegal to play them because in doing so you are defeating a DRM system and doing that is outlawed by the EUCD.
You are quire right. However, it was the *GOVERNMENT* who allowed it to happen by implementing such broad laws with the idea that everyone can be trusted to have their own moral compass.
I think only reason he got an apology at all was because he's 80 odd and was a holocaust surviver.
And the fact that the government shouldn't be in the business of preventing people's free speech - he disagreed with the government and was ejected _because_ of that. Isn't it a wonderful country we live in where you're not allowed to voice your disagreement with the people in power - I seem to remember we invaded Iraq to free the Iraqi people from that kind of oppression... oh no, wait - it was because they've got WMDs, yes, that was it.
Also, I'd like to insert a cliche: I've got nothing to hide.
Until you get pulled in by the police on a murder charge because you happened to be near a murder scene...
I keep seeing broad laws being passed with people saying "well it's ok for them to be really broad because noone will ever abuse them" and then they get abused _every time_.
For example: does shouting "nonsense" in a political debate make you a terrorist? The government seem to think so. Just days before that happened, the Prime Minister argued that it was ok for the anti-terrorism laws (the same ones used to detain someone for shouting "nonsense") to be so broad because the police would never use them inappropriately.
There are similar examples of abuse of the DMCA, EUCD, PATRIOT Act, etc. I've got nothing to hide either... oh wait, yes I do - I play legally purchased DVDs under Linux and that's illegal.
Except the equipment to make plates has been heavily regulated for several years now. Long gone are the days when you could just walk into a shop and ask for a random license plate to be made. Now you need to produce the vehicle license documentation, and a drivers license at least.
I rather suspect the people working in the shops producing licence plates wouldn't need to produce any documentation to knock up a licence plate for themselves in their lunch break. Infact I'm sure a suitably sized brown envelope with cash in it would do the trick too.
And besides, do you _really_ think the cameras can tell the difference between a real licence plate and one made out of cardboard?
No questions asked, no running diag software, nada.
Strange, when we did (many many) Maxtor warranty returns at my last job they required us to send the results from their (Windows only) diagnostic software in order to get an RMA number... which was a PITA since we were a Linux-only company.
Now, when ever I go out and buy a drive, I'm leaning towards Maxtor simply because I have a lot of them and one hasn't failed me with crucial data on it. I'm a lot better prepared to deal with that now as I'm older and wiser so maybe I won't ever feel that level of pain again.
Well there's a lot of anecdotal "evidence" against all the manufacturers - people who buy a very small number of drives will scream loudly when one fails, making that manufacturer seem bad despite it only being a single failure. My _personal_ score so far is:
2 Western Digital Caviars still running fine 11 years after they were bought 1 Western Digital Caviar still running 4 years after it was bought 1 Seagate Baracuda still running 2 years after it was bought 1 Quantum Furrball died after 10 power-on-hours 1 Quantum Furrball died after 3 years 1 Maxtor died after 8 months And various 2.5" hard drives which are all working fine.
That said, my _professional_ experience is a lot different - my old employer used to ship lots of Maxtors and Seagates in servers. Obviously there were failures in both camps, but where we would have a few failures a year from Seagates we had well over a 50% failure rate from Maxtors (and this wasn't just 1 batch, this was drives shipped over a reasonable length of time, and I'm ignoring failures outside of the warranty.) Anecdotal evidence from other people I know who have dealt with a reasonable number of Maxtors also suggests a very high failure rate (again, well over 50% in under a year of purchase, many dieing within a few weeks).
So personally I wouldn't touch Maxtor drives these days. Oh, and compare the noise output of Maxtors against Seagates - the Maxtors are _really_ bad and the Seagates are some of the quietest drives you can get.
What I'd like to ask slashdot readers is for a good way to measure drive quality other than throwing down chicken bones and looking at them or reading tea leaves?
I'm not sure there is a "good" way of measuring drive quality - I've never seen any league tables of drive failure rates and IMHO MTBF figures are next to useless. The best thing to do is RAID them since then if one fails at least you can keep running... of course that also increases the cost of your storage space. Of course, by the time you've been running a particular model of drive for long enough to get reasonable statistics about failure rates then it's obsolete so league tables probably wouldn't be that useful.
If you put in the SPF record in your DNS, AOL will accept your mail.
I do publish SPF records and I think SPF is a Good Thing (although sending bounces to mail you think is spam is terminally stupid since it will almost always end up at some poor innocent sole who's being Joe Jobbed).
I don't know what AOL's current policy is (I haven't tried sending mail to an AOL address recently) but a couple of years ago they had a policy of only accepting mail from mail servers that were registered with them, so all the big ISPs had to run around telling AOL the addresses of their mail servers. As you can imagine, that's a reasonable amount of effort for other ISPs and many of the smaller ones couldn't be bothered to go through the registration process. Especially worrying is what happens if all the other ISPs decide to implement a similar policy to AOL - now every ISP has to manually inform every other ISP what the addresses of their mail servers are. Clearly a crazy idea.
You know, I gotta say that AOL is the best ISP for the computer unsavvy user ever since they've tagged on their anti virus and anti-spyware packages
Unfortunately AOL have a history of implementing some pretty over-zealous anti-spam systems - there have certainly been a number of times when I have tried to reply to an email sent from an AOL address only to have my reply blocked and being told I have to jump through hoops to be allowed to email AOL - the result is that I just don't bother replying because I don't have time to bend over for their brain damaged "anti-spam" policies. That's one reason why I would never recommend someone use AOL. (For the record, this isn't just me - at my last job I frequently had to deal with customers (sysadmins) who were getting mail bounced by AOL. After telling them what hoops they would have to jump through to get AOL to accept the mail the majority of them said it was too much work and they wouldn't bother.)
Unfortunately, whatever protocol AOL uses it's still completely useless to the majority of XMPP users unless Google ditches their plans for a "federation".
AOL has a reputation of being a bad ISP, and also creating bad software for it's users. Will this move help AOL, or hurt Google?
I'm not sure they have a reputation of being a "bad ISP" - they have a reputation of having a very large clueless userbase, which is not the same thing (if anything it might demonstrate their software is easier for clueless people to understand).
Admittedly they've made some fundamentally stupid decisions which has probably driven away a proportion of clueful users whilest making clueless users think the service is "better" (for example, their over-agressive spam filtering. Clueful people will be pissed off that it's overagressive the the clueless will think it's "better" because they're getting less spam).
Personally I wouldn't use any of the ISPs run by massive companies - I don't think any of them are any good:
NTL run an ISP with a terrible quality of service (they do things like run transparent proxies which break all the time and being transparent you can't just tell your browser not to use them). Also have a habit of completely ignoring abuse reports.
BT have a history of doing some fairly stupid things such as NATting their dialup customers, etc. Their technical abilities also seem pretty variable when things go wrong. If you read NANOG for long enough you will see complaints about BT ignoring technical requests from other ISPs too, which is rather bad form.
Demon were a great ISP until they were bought by Thus, at which point the quality of service went downhill and it appears the directive came from management to never admit something was their fault. Before they were bought they were happy to tell people there was a problem with their network but after Thus acquired them they would always deny there was any problem leaving the customers spending hours believing the fault was on their own equipment.
I suppose smaller ISPs manage to pick a small group of very capable employees whereas large companies seem to have a higher proportion of employees who really aren't qualified to do their job.
If Linux is made by hardware vendors to rely on binary drivers, where exactly does that leave its stability advantage?
I think I need to clarify my point: binary drivers are a Bad Thing (it's debatable whether they're better/worse than _no_ drivers, but I'll leave that discussion for now). My point was that if you're boycotting a manufacturer because you disagree with their business practices then why should it just be that one manufacturer - surely you have the same problem with other manufacturers employing the same business practices?
And if you are going to boycott a single manufacturer it might make more sense to do it in an area of the market where there's more choice between closed drivers and open drivers rather than aiming for a market where there aren't many open devices to use _instead_ of the device you're boycotting.
Another problem is manufacturers removing perfectly good well-supported devices from the market (whether they are supported through openness or reverse engineering) and replacing them with closed devices for cost reasons - it seems very difficult to put pressure on the manufacturers to keep making the old devices. A good example of this is Intersil stopping manufacture of the well supported Prism GT 802.11g chipset and replacing it with the Prism Javalin (softmac) chipset which isn't supported at all. Reverse engineering hardware is a lot of work and it's wasted if the hardware you reverse engineered is obsolete by the time you've got a working driver. (It should be noted that the hardware I'm talking about isn't obsolete because it lacked functionality, it's obsolete because the manufacturer invented something marginally cheaper).
Actually their binary drivers are really crashy historically
I'm aware of the problems with the nVidia drivers (I've been waiting on them fixing a bug with interlaced modes for about 2 years), but there's not a lot of hardware choice if you want good performance 3D, and nVidia chipsets are one of the few that do happen to work under Linux. Buggy support is better than no support.
Slashdot Mirror
Unfortunately you've got your own idea of billion/trillion so that just makes it worse.
:)
Alternatively, maybe it's you who has your own idea of billion/trillion
Being serious though, most people in the UK tend to use the US definitions of billion/trillion these days because it fits in better with the SI system - mega/giga/tera == million/billion/trillion
In the UK: $1.000.000,50
Sorry, wrong - we use exactly the same system in the UK as the US does: £1,000,000.50
Broadcast is a stupid model for delivery in a world where you can just buy what you want, when you want. Even if you want to really "broadcast" something because you want people to see it live, multicast is a nice replacement.
I'm sorry, but multicast can't be used for "on demand" stuff because people want to start watching the stream at different times. It's only useful for live stuff (the same as broadcast infact). True on-demand services have to be done over unicast and we're only just starting to get close to the sorts of bandwidth we need (I'd love to know how many simultaneous viewers NTL's TV on demand thing supports - I'm guessing it's not that many). And unless you're receiving your data via cable, you're getting it over a broadcast medium anyway even if you encapsulate it in a multicast protocol, so no benefit here.
we are rapidly approaching critical mass of TV content online.
If everyone in the world was fetching their TV over the Internet using current technologies, your connection would quickly grind to a halt (and the average person wants their TV _now_ - pressing a button and waiting 6 hours for it do download isn't the experience they expect).
The problem could be eased by modifying P2P technology so it tries to only fetch blocks from people local to you where possible - i.e. if you can fetch a block from someone on your ISP then that saves congesting the peering between the ISPs.
but has anyone noticed how appalling modern (UK) cable boxes are? all the ones I've seen have had a second or two lag whenever you change channel
Not much you can do about this - when you change channel the box has to wait until a keyframe arrives before it can start rendering the video. Also, because you get a bit reserviour (to allow the signal to be encoded in a slightly VBR-like way == improved quality) you may need to buffer a bit of the data before you start encoding since the encoder won't be using a strictly constant bit rate but the input will be at a constant rate.
I guess the receiver could buffer the entire multiplex stream so that switching between channels in the same multiplex could be done quicker but without multiple tuners you can't improve the switching speed between channels in different multiplexes. The only possibility I can think of here is to have 2 tuners and have the decoder try and make intelligent guesses about which channel you're likley to change to next.
could you imagine expose-style simultaneous viewing of all your favourite channels
Assuming you had a meaty enough decoder you could decode all the channels in the current multiplex simultaneously. A possiblity is to provide a multiplex with scaled down low quality versions of all the channels in it so your decoder can decode the whole lot and display it at once.
and it would be nice to have full Tivo-like capabilities from my PC
Can I take this opportunity to point at MythTV. I've been using it for several years and it's still better than the systems the sat and cable operators are providing here in the UK. I was stuck using my parent's NTL Digital system at Christmas and realised just how much functionality I take for granted in Myth which just isn't there in the NTL system. Similarly I hear friends commenting about features they wished Sky+ had and they're always things that Myth has already been doing for ages.
Admittedly there are a few bugs that neet to be mopped up in Myth but on the whole it's a good project. I think the only thing I'd really like to be able to do that I can't already do is use a satellite tuner in my myth box instead of re-encoding the S-Video output from my Sky box. (Yes, Myth does support DVB-S cards, but getting a DVB-S card with a common interface is very pricey still and since Sky won't produce a VideoGuard CAM I'd be stuck using an illegal programmable CAM such as the DragonCAM - total cost, £100 for the card, £50 for the common interface and £80 for the CAM == way too pricey)
Channel numbers are obsolete, anyway, or at least they ought to be. Let me pick my channels from an onscreen menu. If I have favorites, let me memorize the shortcuts on my own terms, not as the arbitrary, unchangeable numbers that come from my cable provider.
Well, yes and no - I have Sky and the fact that every so often they change all the channel numbers without warning is a pain in the arse (suddenly my Myth box is changing to the wrong channel, etc). However, channel numbers are better than user defined shortcuts for 2 reasons:
1. Can I be bothered to set up short cuts for my favorite channels? Nope... my "favorites" list is infact empty, I've never got around to settin it up. A channel number is already there and I just have to remember it.
2. If I use someone else's Sky decoder then I can get to my favorite channels just as easilly - my "shortcuts" list is defined in my brain and so is portable, a shortcuts list on my decoder stays on my decoder. (Similarly, my favorite stations may not be the same as other people's in the same house so suddenly we need to allow for the definition of even more shortcuts to cope with multiple people).
Which brings me to a side point - maybe we need some central organisation handing out channel numbers so that all the cable/sat operators use the same numbers, kind of like telephone numbers. I don't have to dial a different phone number to talk to my mum if I dial from a different telco, why should I have to use a different channel number because my TV is provided by a different company?
I've often wondered why a truck license isn't required for commercial vehicles, since they fall under the same section of legislation for vehicle safety.
Well anyone (here in the UK) who got their driving licence before some time around 1997 (me included) is allowed to drive light trucks anyway... (up to 7.5 Tons ISTR)
While it is true more people have the opertunity to go through the source and check for problems, how many actually do? Did you read throught the source for any of the open source software you installed ? I certainly don't have the time and in many cases enough knowledge of the language and problem to properly review the code.
No, people don't generally review the source of everything they install. But the point is that there are more people fiddling with the code and therefore more people to spot bugs. On several occasions I have been hacking something into the Linux kernel and spotted bugs in existing code (fairly core stuff like netfilter, netdev, etc). When I spot these bugs I fix them where I can (and submit patches) and if I can't fix them myself I work with the relevent developers to help them develop a fix. As a result of this, a number of fairly serious core kernel bugs have been fixed as a result of me happening to spot them while hacking at something related - the same couldn't be said about Windows bugs (and infact because Windows is closed source I probably wouldn't have been able to hack up soemthing to do what I wanted anyway).
I never been that convinced by the Cathedral vs the Bazaar argument. Given fewer people use linux and of those users even less will have enough knowledge, time and expertise to bug hunt in any meaningful way is having the source open that much of an advantage?
Well, I wouldn't call most FOSS development a "Bazaar" - people don't just randomly apply code to projects, you submit stuff to a maintainer who will than decide whether or not to apply it (or indeed apply a re-coded version of your patch).
However, I don't think you should think about the percentage of users who will see the code - you want to think about the total number of people who will see the code and I'm willing to bet that there are a lot more people working on and reviewing any piece of core Linux code than Windows code.
5th Gear did a test a while back on 4x4's in near-miss conditions. They took a couple of normal cars, drove them along at motorway speeds and simulated a motorway incident with the front car slamming on it's brakes and the back car having to swerve to avoid it. Nomatter what they did they could only make the car spin, not roll over. They then re-ran the test but this time at only 40mph and with a 4x4 as the second car - it rolled right over as soon as the driver swerved to avoid the stopped car. Certainly very eye opening - the driver of the 4x4 would definately not be in very good shape if he wasn't wearing a racing harness, etc.
Increases in that sort of safety may decrease insurance premiums
Or it may increase them. More survivors of road accidents == more people alive to claim long term personal injury payments. (No, I'm not advocating killing people to cut insurance, but it _is_ a fact that in some cases it's cheaper for an insurance company to pay out a large sum to the familly of someone who was killed rather than paying out for life long treatment of an injury).
The point is that if all that extra weight is put into engineering stronger cars that absorb the impact and slow you down more gradually then you're going to be more likely to survive a crash.
Well, lets assume that both vehicles involved in a collision are of identical design... actually, lets make it even easier - you're crashing a car into a solid object. If you've got a 1 metre crumple-zone at the front of the car, it means that you _will_ accellerate from whatever speed you were doing to 0mph within the distance of 1 metre. It doesn't matter how heavy or strong the car is, if you're doing 30mph and you hit something solid then you're going to go from 30mph to 0mph in the space of 1 metre. The only thing you can do to improve the performance in a crash is to make sure the accelleration is constant over the length of the whole crumple-zone. I.e. if you're doing 30mph, hit something solid and 90cm later you're doing 20mph then your problem is that you have to lose that 20mph in a distance of only 10cm. Ideally the crumple-zone of a car would be made out of some kind of smart material that changed it's strength depending on the size of an impact, so a 30mph and a 60mph collision would both result in the whole crumple zone being used and the energy distributed over the whole length of it.
Now, going back to having 2 cars colliding, this is only different to hitting something solid if there is an imbalance in the equation - i.e. one of the cars is slower, lighter or weaker. So assuming you already have a perfect crumple-zone, the only way you can engineer the car to dissipate it's energy over a longer period of time (and therefore make the crash more survivable) is at the expense of the other car in the crash - e.g. make your car massively heavier than the car you're going to hit.
Is a major software manufacturer, or in the mentioned case of Linux, OS type *REQUIRED* to provide support for a patented technology? Especially without royalties being paid?
IMHO, while patents may be acceptable on how a medium works, they shouldn't (be allowed to) cover how the medium is _read_ (actually, I don't really believe in patents at all but...). Although my original complaint wasn't actually intended to be about patents - whilest it may be illegal to play DVDs under Linux because of a patent, it's also illegal to play them because in doing so you are defeating a DRM system and doing that is outlawed by the EUCD.
The GOVERNMENT doen't see to think so at all.
You are quire right. However, it was the *GOVERNMENT* who allowed it to happen by implementing such broad laws with the idea that everyone can be trusted to have their own moral compass.
I think only reason he got an apology at all was because he's 80 odd and was a holocaust surviver.
And the fact that the government shouldn't be in the business of preventing people's free speech - he disagreed with the government and was ejected _because_ of that. Isn't it a wonderful country we live in where you're not allowed to voice your disagreement with the people in power - I seem to remember we invaded Iraq to free the Iraqi people from that kind of oppression... oh no, wait - it was because they've got WMDs, yes, that was it.
Also, I'd like to insert a cliche: I've got nothing to hide.
Until you get pulled in by the police on a murder charge because you happened to be near a murder scene...
I keep seeing broad laws being passed with people saying "well it's ok for them to be really broad because noone will ever abuse them" and then they get abused _every time_.
For example: does shouting "nonsense" in a political debate make you a terrorist? The government seem to think so. Just days before that happened, the Prime Minister argued that it was ok for the anti-terrorism laws (the same ones used to detain someone for shouting "nonsense") to be so broad because the police would never use them inappropriately.
There are similar examples of abuse of the DMCA, EUCD, PATRIOT Act, etc. I've got nothing to hide either... oh wait, yes I do - I play legally purchased DVDs under Linux and that's illegal.
Except the equipment to make plates has been heavily regulated for several years now. Long gone are the days when you could just walk into a shop and ask for a random license plate to be made. Now you need to produce the vehicle license documentation, and a drivers license at least.
I rather suspect the people working in the shops producing licence plates wouldn't need to produce any documentation to knock up a licence plate for themselves in their lunch break. Infact I'm sure a suitably sized brown envelope with cash in it would do the trick too.
And besides, do you _really_ think the cameras can tell the difference between a real licence plate and one made out of cardboard?
This cant stop "terrorists", they can go and buy a car for £1000 from any used car dealer whenever they like, or OMG they could get a bus or train.
:)
Clearly you haven't used the public transport system in the UK
Were you dealing with the online RMA tool, or calling direct?
I think we were using the online tool...
No questions asked, no running diag software, nada.
Strange, when we did (many many) Maxtor warranty returns at my last job they required us to send the results from their (Windows only) diagnostic software in order to get an RMA number... which was a PITA since we were a Linux-only company.
Now, when ever I go out and buy a drive, I'm leaning towards Maxtor simply because I have a lot of them and one hasn't failed me with crucial data on it. I'm a lot better prepared to deal with that now as I'm older and wiser so maybe I won't ever feel that level of pain again.
Well there's a lot of anecdotal "evidence" against all the manufacturers - people who buy a very small number of drives will scream loudly when one fails, making that manufacturer seem bad despite it only being a single failure. My _personal_ score so far is:
2 Western Digital Caviars still running fine 11 years after they were bought
1 Western Digital Caviar still running 4 years after it was bought
1 Seagate Baracuda still running 2 years after it was bought
1 Quantum Furrball died after 10 power-on-hours
1 Quantum Furrball died after 3 years
1 Maxtor died after 8 months
And various 2.5" hard drives which are all working fine.
That said, my _professional_ experience is a lot different - my old employer used to ship lots of Maxtors and Seagates in servers. Obviously there were failures in both camps, but where we would have a few failures a year from Seagates we had well over a 50% failure rate from Maxtors (and this wasn't just 1 batch, this was drives shipped over a reasonable length of time, and I'm ignoring failures outside of the warranty.) Anecdotal evidence from other people I know who have dealt with a reasonable number of Maxtors also suggests a very high failure rate (again, well over 50% in under a year of purchase, many dieing within a few weeks).
So personally I wouldn't touch Maxtor drives these days. Oh, and compare the noise output of Maxtors against Seagates - the Maxtors are _really_ bad and the Seagates are some of the quietest drives you can get.
What I'd like to ask slashdot readers is for a good way to measure drive quality other than throwing down chicken bones and looking at them or reading tea leaves?
I'm not sure there is a "good" way of measuring drive quality - I've never seen any league tables of drive failure rates and IMHO MTBF figures are next to useless. The best thing to do is RAID them since then if one fails at least you can keep running... of course that also increases the cost of your storage space. Of course, by the time you've been running a particular model of drive for long enough to get reasonable statistics about failure rates then it's obsolete so league tables probably wouldn't be that useful.
If you put in the SPF record in your DNS, AOL will accept your mail.
I do publish SPF records and I think SPF is a Good Thing (although sending bounces to mail you think is spam is terminally stupid since it will almost always end up at some poor innocent sole who's being Joe Jobbed).
I don't know what AOL's current policy is (I haven't tried sending mail to an AOL address recently) but a couple of years ago they had a policy of only accepting mail from mail servers that were registered with them, so all the big ISPs had to run around telling AOL the addresses of their mail servers. As you can imagine, that's a reasonable amount of effort for other ISPs and many of the smaller ones couldn't be bothered to go through the registration process. Especially worrying is what happens if all the other ISPs decide to implement a similar policy to AOL - now every ISP has to manually inform every other ISP what the addresses of their mail servers are. Clearly a crazy idea.
You know, I gotta say that AOL is the best ISP for the computer unsavvy user ever since they've tagged on their anti virus and anti-spyware packages
Unfortunately AOL have a history of implementing some pretty over-zealous anti-spam systems - there have certainly been a number of times when I have tried to reply to an email sent from an AOL address only to have my reply blocked and being told I have to jump through hoops to be allowed to email AOL - the result is that I just don't bother replying because I don't have time to bend over for their brain damaged "anti-spam" policies. That's one reason why I would never recommend someone use AOL. (For the record, this isn't just me - at my last job I frequently had to deal with customers (sysadmins) who were getting mail bounced by AOL. After telling them what hoops they would have to jump through to get AOL to accept the mail the majority of them said it was too much work and they wouldn't bother.)
Or AOL switches to Jabber?
Unfortunately, whatever protocol AOL uses it's still completely useless to the majority of XMPP users unless Google ditches their plans for a "federation".
AOL has a reputation of being a bad ISP, and also creating bad software for it's users. Will this move help AOL, or hurt Google?
I'm not sure they have a reputation of being a "bad ISP" - they have a reputation of having a very large clueless userbase, which is not the same thing (if anything it might demonstrate their software is easier for clueless people to understand).
Admittedly they've made some fundamentally stupid decisions which has probably driven away a proportion of clueful users whilest making clueless users think the service is "better" (for example, their over-agressive spam filtering. Clueful people will be pissed off that it's overagressive the the clueless will think it's "better" because they're getting less spam).
Personally I wouldn't use any of the ISPs run by massive companies - I don't think any of them are any good:
NTL run an ISP with a terrible quality of service (they do things like run transparent proxies which break all the time and being transparent you can't just tell your browser not to use them). Also have a habit of completely ignoring abuse reports.
BT have a history of doing some fairly stupid things such as NATting their dialup customers, etc. Their technical abilities also seem pretty variable when things go wrong. If you read NANOG for long enough you will see complaints about BT ignoring technical requests from other ISPs too, which is rather bad form.
Demon were a great ISP until they were bought by Thus, at which point the quality of service went downhill and it appears the directive came from management to never admit something was their fault. Before they were bought they were happy to tell people there was a problem with their network but after Thus acquired them they would always deny there was any problem leaving the customers spending hours believing the fault was on their own equipment.
I suppose smaller ISPs manage to pick a small group of very capable employees whereas large companies seem to have a higher proportion of employees who really aren't qualified to do their job.
If Linux is made by hardware vendors to rely on binary drivers, where exactly does that leave its stability advantage?
I think I need to clarify my point: binary drivers are a Bad Thing (it's debatable whether they're better/worse than _no_ drivers, but I'll leave that discussion for now). My point was that if you're boycotting a manufacturer because you disagree with their business practices then why should it just be that one manufacturer - surely you have the same problem with other manufacturers employing the same business practices?
And if you are going to boycott a single manufacturer it might make more sense to do it in an area of the market where there's more choice between closed drivers and open drivers rather than aiming for a market where there aren't many open devices to use _instead_ of the device you're boycotting.
Another problem is manufacturers removing perfectly good well-supported devices from the market (whether they are supported through openness or reverse engineering) and replacing them with closed devices for cost reasons - it seems very difficult to put pressure on the manufacturers to keep making the old devices. A good example of this is Intersil stopping manufacture of the well supported Prism GT 802.11g chipset and replacing it with the Prism Javalin (softmac) chipset which isn't supported at all. Reverse engineering hardware is a lot of work and it's wasted if the hardware you reverse engineered is obsolete by the time you've got a working driver. (It should be noted that the hardware I'm talking about isn't obsolete because it lacked functionality, it's obsolete because the manufacturer invented something marginally cheaper).
Actually their binary drivers are really crashy historically
I'm aware of the problems with the nVidia drivers (I've been waiting on them fixing a bug with interlaced modes for about 2 years), but there's not a lot of hardware choice if you want good performance 3D, and nVidia chipsets are one of the few that do happen to work under Linux. Buggy support is better than no support.