It sounds to me like you're pining for the days before people realized that object destruction costs, and that therefore it's best left to a garbage collector that can operate at idle times.
Sometimes, sometimes not. C++ has an idom "resource allocation is object creation" which when followed lets you use hevyweight OS/Toolkit objects without worrying that the garbage collector might delay "too long" in running destructors. It also lets you use it for things like locking object use in multi-threaded code. Of corse to avoid pain you either need to use stack locals for this, or reference counted pointers (or even the odd auto_ptr beastie).
It would be much nicer if a language let the author of the class decide if the class should be referece counted and get deterministic destruction times, or if garbage collection is "good enough" (modern GC is frequently faster then reference counting, so you don't want to use counts when exact destruction times is not a big deal).
Yeah it seems to me that people only watch game shows when nothing else is on.
I'm not so sure, Millionare and the Weakest Link both did pretty good in primetime where the most good stuff seems to be on (i.e. the times I'm most likely to run out of tuners!)
So some people like gameshows a fair bit.
If I had a TiVo, I'd never watch game shows again!
So go buy one already! Seriously I have very few consumer eletronics purchases that worked out far better then I thought they would, and the TiVo is the biggest of those (Vonage coming in second, my Mac laptop the third, and the iPod a distant distant 4th).
Most people who have a TiVo record everything they watch, or really watch only stuff they have recorded, so they have the TiVo set up to record everything they might want to watch.
I tend not to watch things I record on the TiVo more then once (after all it tends to pick up all sorts of new stuff I want to watch more then stuff I have just seen), in fact there are things it records that I just don't get around to watching (sure the Simpsons are great, and when I'm in the mood to watch them it is nice to have some on the TiVo, but if I don't get around to seeing it before it is a week ot two old and vanishes, it's no great loss, ditto for Good Eats, and a bunch of stuff on the History channel, and...).
So the real question is "who watches game shows"? (and the answer is "not me")
If I don't expect my electricity bill to be subsidized by neighbors who use less electricity that me, then why should I expect my bandwidth usage to be subsidized by neighbors who use less bandwidth than me?
Most people have a good idea how to keep their eletricity bill down. Most people don't have to worry about some virus coming in and making their eletricity bill go from $40 one month to $400.
So I think ISPs will try to find a way to keep the P2P traffic mostly on their network so it is a whole lot cheaper (but not free), which is better (faster) for their end users anyway. Failing that they may rate limit off-net P2P traffic (when it uses known port numbers), which would not be better for their customers. Metered rates are likely to be only a last resort.
Actually it does, in the USA DSL providers tend to buy copper from the RBOC and provide their own DSL service and arrange to send the IP traffic to their own peers or transit providors. In the UK DSL providers buy pre-provisioned DSL service and transit from BT (or at least that was how it looked when I was visiting, maybe it can be done other ways but the DSL providers I looked at all did it that way).
Technically DSL doesn't get slower if lots of people use it, but economically the way BT sells it to the people that sell to end users you get the same effect. Even in the USA you can get that effect if your DSL provider under provisions their transit/peering circuits, but then you just leave them and buy form someone else. In the UK "someone else" has the same crappy connection to the internet as the bloke you just fired.
They say in the article that they will have an "antenna stabilisation system" that keeps the balloon in place even if its getting blow around, but could it really keep it in place within a few feet considering it's on a 1.5km tether?
I doubt it keeps the balloon in place, just angles the balloon's antenna (or they have solid state aiming systems). I don't know how long it would take to re-aim after a gust though, is it 30ms worth of lost packets? A half second? Eight seconds? (if it is a second or two, and a "re-xmit last two seconds" message is sent it won't be a big deal...)
Customer sites would need tight beam directional antennas to go that far that would lose signal as soon as the balloon moved much at all.
I doubt it. I expect they would choose a frequency they can just blast "up", and not disturb much else. No super fancy aiming needed.
I bought a 900Mhz, 14dBi yagi (HG914Y) from them for making cell phones work at my mom's no-cell-reception house. It works extremely well, and I love the welded aluminium construction and white powder-coating
As in you installed this in her house an now she has to hook her phone up to it to get reception, or you installed it in her house and now cell phones in her house get better reception? If so, how does that work? (and... I must get one....or the 1.9Ghz version actually...)
Correct me if I'm wrong, but since the second-rev G3 machines (blue and white towers), hasn't the Apple BIOS been unnecessary?
With the Apple "new world" machines the OS isn't in the BIOS, but the BIOS is still needed to load the OS into RAM in the first place, and to do some inital set up of the hardware and a few other pre-OS jobs. Much like the BIOS in a PC, or even more like the Sun OpenFirmware (since it is in fact a FORTH boot enviroment maybe licenced from Sun or just extreamly closely modeled after it).
It should be noted that Berkley was pondering a countersuit, claiming that Novel's code lifted large portions from BSD without copyrights or attributions.
They actually entered a plea for releif during the Novel vs. BSDI suit. I don't recall if it was independent or only in the context of the other suit, but I found it in with documents on that suit.
Isn't the DFA pipeline schedular much better at producing fast code for VLIW and wide issue super scaler CPUs? (and if the P4 isn't "wide issue", it at least deals with knowing which instructions use which pipelines better then the old scheduler).
I would expect at least some speedups on most superscaler CPUs (i.e. the P1, P2, P3, P4, the AMD K7, the modern and somewhat old PowerPC and SPARCs). Not huge ones since there was a superscaler scheduler before, but the new one is better.
In the US cell phone theft isn't a big problem at all due to the fact that our phones don't use SIMs and therefore are tied to a specific network
Some providers don't use SIMs, some do. AT&T's old service didn't, their new one does, T-Mobile does, Cingular does. Verizon doesn't, NexTel doesn't, Sprint doesn't.
The ones that do use phones with SIMs (GSM at 1900Mhz) lock the phones to only work with their SIMs, which most UK cell providers also do. You can get the phones unlocked though (oddly enough I got my US phone unlocked to use with a UK SIM while I was there, the UK SIM cost £12, unlocking was £10; unlocking in the USA would have been $35, so good deal).
GSM phones do have a unique per phone ID just like the CDMA (and I think AMPS) phones, and one could make a global "stolen phone" list, but I don't think that is done. I havn't had a phone stolen though, so I don't know.
The original complaint was "99% of apps don't need more then 4G". The "multiple programs" argument is talking about the OS handling more then 4G, not apps. You can easily do this on a OS that assigns a new address space per app (i.e. Unix) without the application programs being aware of it. In fact you can do it without even supporting 64-bit apps (the Intel XEON for example), if you are willing to make the OSes life hell.
As for the original complaint, yes I do buy it, 99% of apps don't need more then 4G. However the 1% of apps that do need it (or at least can make good use of it!) are very important to some people. And are a growing set of apps anyway. So if you can get to 64 bits for free it is a no brainer to do it. Since it isn't free it is a bit of a challenge to decide when to do it, but for Apple "before Microsoft" seems like a pretty good time...
Do any of the apps I use need more then 64 bits? Nope, I may have a digital SLR, but it is only 3Mpixels. I may have a DV cam, but I don't think iMovie will really get a huge kick out of 64-bit address spaces, at least not until I get more then 1G of RAM! I sure hope gcc doesn't need a 64-bit address space any time soon!!! But I recognize there are other people that use computers, and that many of those people have money.
Its called a higher cost of living and is very much related to the value of their currency
On the other hand $1.20 of flour costs 7p in the UK, and some other food staples also have that alarming sort of low price that might make someone from the USA think they are getting floor scrapings rather then real food.
The execution of instructions from Thread A and Thread B can be interleaved an any random order whether there 1-100 CPUs. The bugs are a little more likely in the SMP case, but I can't imagine a user mode bug that can happen in the SMP case, but not the single.
In theory they can be interleved. In practice things are frequently not so random. For example when you call fork on a non-SMP system the parent process gets to use up it's time slice before the new child process gets it's first chance to run. There might be some (buggy!) code that accidentally does the wrong thing and gets away with it because it does some stuff that needs a lock it will normally be ok. Once in a while it will fail. Not real frequently, just once in a while. On a SMP system if there isn't anything else running the child process and parent process will both return from the fork at the same time and that "once in a great while it goes bad" thing becomes "most of the time it blows up".
That's what happened when Sun became the first major Unix-like system that did real SMP.
I expect there is other stuff that can go wrong going from UP to SMP, and while lots of OSes have done it before, none have the same security bent that OpenBSD does, so there may be security bugs that still are unfixed...
But seriously, what is the going price for the cheapest CD-Rs in the US?
Last I checked it was free. Really. Well, almost. It was $20 plus 7% sales tax for a spool of 50, and I got a $20 rebate form, kick in another $0.34 for a stamp and it wasn't quite free, but it was around that price.
I see a lot of similar rebates, I guess the componies doing that count on a low rate of replies since CD-R's can't be cheap enough to stamp 'em out just for the float on $20 a spool!
I have developed a large system that deals with end-to-end running of a large supplier of outdoor leather goods. Including B2B transactions, custom querying, post-sales tracking.
And lots of stuff is written in Cobol (still!!). The article went out of it's way to say Java isn't dead as in not used, "just" dead as future languages won't steal ideas from Java.
Depending on how you look it at, that's right or wrong. I see future languages as having exception handling, name spaces, (compiler and runtime linker enforced) security features, big libraires, lots of OS independence, OO, garbage collection, and lots of other stuff Java has.
Then again you can find all those features in past languages. Even the (compiler and runtime linker enforced) security stuff was on, um, er, some sort of stack based mainframe set of languages. Many languages had the other stuff, Modula-3 for example had almost all of it.
So if a future language has the same kind of GC and exception handling that both Modula-3 and Java has, is it decending from Java, or from Modula-3?
So if the momentum has built for DSL and Cable, why push for the third option too?
Because there are lots of places that don't have DSL, or IP over Cable TV...and even more places that don't have both. It's nice to have something avilable, but even nicer to have more then one thing so they can compete...
Well to the extent that I think the goverment can ever help the "free" market out...yes, yes it is. DirecTV was beign sold, and the choices were to sell it to the only other USA satalite TV system, or to sell it to some media content conglomarate. One of those radically reduces the consumers avilable choices of satalite TV delevery (from "not much choice" to "no choice"), and even if you accept DISH/DTV's viewpoint that they compete with Cable TV systems then the choice goes from 3 to 2 in most places (some places from 4 to 2, a few places from 2 to 1 still).
So from the consumers viewpoint at least in the short to mid term DirecTV being bought by Sky, er, Fox, er, Murdoch is better then DISH. In the long term? Well who knows, it might have been better to let the 2 satalite componies merge and ovver more channels or something.
So it could be merely a temporary anomaly in mobile pricing.
It might be a temporary anomaly in mobile pricing, but it is a permeant feature of the underlying mobile cost. If you are an orange customer calling another orange customer the call can (and most likely does) run entirely over a network that orange has payed for, and doesn't cost orange anything. If you are an orange customer and call an O2 customer then you have to pay O2 to put the call through, and you may have to pay someone other then O2 to bring the voice traffic to O2.
So if the "anomaly in mobile pricing" goes away, it will vanish because same network call prices will go up.
One USA moble provider has yet another issue. Their cell phones can actually make direct handset to handset "intercom" calls. I think they make those calls directly like walkie-talkies (there is a distance limit). They charge $0.00 per minute for this.
Re:Why is this so hard to understand?
on
BusinessWeek on Wi-Fi
·
· Score: 2, Insightful
There is no business model for wireless, that's why it's so great!
Yeah, but if there is a business model that includes profit there will be a lot more wireless sites to use (if you are willing to pay), and it shouldn't have a big impact on the number of free ones.
Having a model where some folks "give it away" is great, the lack of a for-profit model isn't great.
specially since Comcast and Cox (2 of the top cable companies [tracetools.com]... especially since Comcast now owns AT&T) are all Tivo investors
Sure, but the way AT&T cable was built means it is 40+ different cable systems, right down to the CPE. So they need 40+ different TiVo boxes to get the same sort of features that DirecTV gets (high level of integration so you don't need to tell the TiVo what channes you do/don't get, and so it never misses a channel change, and so it can record mroe then one thing at once...)
Cable companies like TiVo, they see the much much lower churn rate DTV gets from it's DirecTV w/ TiVo users then the normal ones. They just can't get the same product because they screwed themselves in the past by making sure no 2 cable companies used the exact same broadcast system.
But do you honestly think that people who think about lighting in a store can do something about that?
Well Jobs (apparently) spent 4 hours a day (presumabbably not for life, but for more then a few days) working with the lighting people. One assumes that is 4 hours a day he can't spend on the phone to Moto saying "get me a faster CPU!", or to IBM saying "get me POWER4's I can afford to sell!"
I just started the Hyperion series, and after that, the Foundation trilogy. Would you put the Ringworld series in this kind of class of a whole universe unfolded for the reader?
And life just gets more and more like TV: Now, I have to consider whether my family/friends/coworkers are "gettin' paid" before I take them up on that recommendation to see "Master Of Disguise II".
You aready do. There are a lot of products that have referal programs (for example the phone company I use will give you and I both $40 worth of service if you mention my name when you sign up -- p.s. it's a great service, you should buy it, I'm sure it will be just perfect for you. what are you waiting for? Buy! Buy! Buy!). Other products that I just happen to like, but don't seem to get as much traction in the market place as I think they need, I might go out and shill for them in the hopes that they get enough busness to stay afloat (you need a TiVo, really, go get one, or another one as appropriate, hurry! Stores might run out! Don't get cought in a shortage!).
Really, for most non-warez (and related) people, a 20GB harddrive would be more than enough. Of course I'm aware of servers, datacenters, people working in film production, the music industry, et al, but these are hardly the majority of harddrive buyers.
Add photographers to that list. I have shot close to 1G worth of still images in a day, and that is using lossy compression and a "mere" 3Mpixel camera.
Back in the 80s, the drives in my computers never died, and I can still boot up that ol' Macintosh SE, and the harddrive works.
Lots of older drives failed too. The drive in the Lisa (aka Mac XL...well, once new ROMs were put in) was prone to failure. I remeber the double eagle drives in the late 80s being failure prone too.
I want my harddrives to be as reliable as my RAM.
I want my car to be able to drive to mars, but it doesn't seem likely, nor does it seem likely that a highly complex mechinical device with exceptionally tight tolerences and moving parts will be as reliable as a solid state device.
I quite want a reliable drive too. Fast would be nice, very arge storage capacity would be very very nice. Affordable woud also be nice. In fact it doesn't have to be a drive, it could be something that acts like one, but FLASH ROM is still too slow (to erase, reads are fast if you design right) and too costly.
Slashdot Mirror
Sometimes, sometimes not. C++ has an idom "resource allocation is object creation" which when followed lets you use hevyweight OS/Toolkit objects without worrying that the garbage collector might delay "too long" in running destructors. It also lets you use it for things like locking object use in multi-threaded code. Of corse to avoid pain you either need to use stack locals for this, or reference counted pointers (or even the odd auto_ptr beastie).
It would be much nicer if a language let the author of the class decide if the class should be referece counted and get deterministic destruction times, or if garbage collection is "good enough" (modern GC is frequently faster then reference counting, so you don't want to use counts when exact destruction times is not a big deal).
I'm not so sure, Millionare and the Weakest Link both did pretty good in primetime where the most good stuff seems to be on (i.e. the times I'm most likely to run out of tuners!)
So some people like gameshows a fair bit.
So go buy one already! Seriously I have very few consumer eletronics purchases that worked out far better then I thought they would, and the TiVo is the biggest of those (Vonage coming in second, my Mac laptop the third, and the iPod a distant distant 4th).
Most people who have a TiVo record everything they watch, or really watch only stuff they have recorded, so they have the TiVo set up to record everything they might want to watch.
I tend not to watch things I record on the TiVo more then once (after all it tends to pick up all sorts of new stuff I want to watch more then stuff I have just seen), in fact there are things it records that I just don't get around to watching (sure the Simpsons are great, and when I'm in the mood to watch them it is nice to have some on the TiVo, but if I don't get around to seeing it before it is a week ot two old and vanishes, it's no great loss, ditto for Good Eats, and a bunch of stuff on the History channel, and...).
So the real question is "who watches game shows"? (and the answer is "not me")
Most people have a good idea how to keep their eletricity bill down. Most people don't have to worry about some virus coming in and making their eletricity bill go from $40 one month to $400.
So I think ISPs will try to find a way to keep the P2P traffic mostly on their network so it is a whole lot cheaper (but not free), which is better (faster) for their end users anyway. Failing that they may rate limit off-net P2P traffic (when it uses known port numbers), which would not be better for their customers. Metered rates are likely to be only a last resort.
Actually it does, in the USA DSL providers tend to buy copper from the RBOC and provide their own DSL service and arrange to send the IP traffic to their own peers or transit providors. In the UK DSL providers buy pre-provisioned DSL service and transit from BT (or at least that was how it looked when I was visiting, maybe it can be done other ways but the DSL providers I looked at all did it that way).
Technically DSL doesn't get slower if lots of people use it, but economically the way BT sells it to the people that sell to end users you get the same effect. Even in the USA you can get that effect if your DSL provider under provisions their transit/peering circuits, but then you just leave them and buy form someone else. In the UK "someone else" has the same crappy connection to the internet as the bloke you just fired.
I doubt it keeps the balloon in place, just angles the balloon's antenna (or they have solid state aiming systems). I don't know how long it would take to re-aim after a gust though, is it 30ms worth of lost packets? A half second? Eight seconds? (if it is a second or two, and a "re-xmit last two seconds" message is sent it won't be a big deal...)
I doubt it. I expect they would choose a frequency they can just blast "up", and not disturb much else. No super fancy aiming needed.
As in you installed this in her house an now she has to hook her phone up to it to get reception, or you installed it in her house and now cell phones in her house get better reception? If so, how does that work? (and... I must get one....or the 1.9Ghz version actually...)
With the Apple "new world" machines the OS isn't in the BIOS, but the BIOS is still needed to load the OS into RAM in the first place, and to do some inital set up of the hardware and a few other pre-OS jobs. Much like the BIOS in a PC, or even more like the Sun OpenFirmware (since it is in fact a FORTH boot enviroment maybe licenced from Sun or just extreamly closely modeled after it).
They actually entered a plea for releif during the Novel vs. BSDI suit. I don't recall if it was independent or only in the context of the other suit, but I found it in with documents on that suit.
Isn't the DFA pipeline schedular much better at producing fast code for VLIW and wide issue super scaler CPUs? (and if the P4 isn't "wide issue", it at least deals with knowing which instructions use which pipelines better then the old scheduler).
I would expect at least some speedups on most superscaler CPUs (i.e. the P1, P2, P3, P4, the AMD K7, the modern and somewhat old PowerPC and SPARCs). Not huge ones since there was a superscaler scheduler before, but the new one is better.
Some providers don't use SIMs, some do. AT&T's old service didn't, their new one does, T-Mobile does, Cingular does. Verizon doesn't, NexTel doesn't, Sprint doesn't.
The ones that do use phones with SIMs (GSM at 1900Mhz) lock the phones to only work with their SIMs, which most UK cell providers also do. You can get the phones unlocked though (oddly enough I got my US phone unlocked to use with a UK SIM while I was there, the UK SIM cost £12, unlocking was £10; unlocking in the USA would have been $35, so good deal).
GSM phones do have a unique per phone ID just like the CDMA (and I think AMPS) phones, and one could make a global "stolen phone" list, but I don't think that is done. I havn't had a phone stolen though, so I don't know.
The original complaint was "99% of apps don't need more then 4G". The "multiple programs" argument is talking about the OS handling more then 4G, not apps. You can easily do this on a OS that assigns a new address space per app (i.e. Unix) without the application programs being aware of it. In fact you can do it without even supporting 64-bit apps (the Intel XEON for example), if you are willing to make the OSes life hell.
As for the original complaint, yes I do buy it, 99% of apps don't need more then 4G. However the 1% of apps that do need it (or at least can make good use of it!) are very important to some people. And are a growing set of apps anyway. So if you can get to 64 bits for free it is a no brainer to do it. Since it isn't free it is a bit of a challenge to decide when to do it, but for Apple "before Microsoft" seems like a pretty good time...
Do any of the apps I use need more then 64 bits? Nope, I may have a digital SLR, but it is only 3Mpixels. I may have a DV cam, but I don't think iMovie will really get a huge kick out of 64-bit address spaces, at least not until I get more then 1G of RAM! I sure hope gcc doesn't need a 64-bit address space any time soon!!! But I recognize there are other people that use computers, and that many of those people have money.
On the other hand $1.20 of flour costs 7p in the UK, and some other food staples also have that alarming sort of low price that might make someone from the USA think they are getting floor scrapings rather then real food.
In theory they can be interleved. In practice things are frequently not so random. For example when you call fork on a non-SMP system the parent process gets to use up it's time slice before the new child process gets it's first chance to run. There might be some (buggy!) code that accidentally does the wrong thing and gets away with it because it does some stuff that needs a lock it will normally be ok. Once in a while it will fail. Not real frequently, just once in a while. On a SMP system if there isn't anything else running the child process and parent process will both return from the fork at the same time and that "once in a great while it goes bad" thing becomes "most of the time it blows up".
That's what happened when Sun became the first major Unix-like system that did real SMP.
I expect there is other stuff that can go wrong going from UP to SMP, and while lots of OSes have done it before, none have the same security bent that OpenBSD does, so there may be security bugs that still are unfixed...
Last I checked it was free. Really. Well, almost. It was $20 plus 7% sales tax for a spool of 50, and I got a $20 rebate form, kick in another $0.34 for a stamp and it wasn't quite free, but it was around that price.
I see a lot of similar rebates, I guess the componies doing that count on a low rate of replies since CD-R's can't be cheap enough to stamp 'em out just for the float on $20 a spool!
And lots of stuff is written in Cobol (still!!). The article went out of it's way to say Java isn't dead as in not used, "just" dead as future languages won't steal ideas from Java.
Depending on how you look it at, that's right or wrong. I see future languages as having exception handling, name spaces, (compiler and runtime linker enforced) security features, big libraires, lots of OS independence, OO, garbage collection, and lots of other stuff Java has.
Then again you can find all those features in past languages. Even the (compiler and runtime linker enforced) security stuff was on, um, er, some sort of stack based mainframe set of languages. Many languages had the other stuff, Modula-3 for example had almost all of it.
So if a future language has the same kind of GC and exception handling that both Modula-3 and Java has, is it decending from Java, or from Modula-3?
Because there are lots of places that don't have DSL, or IP over Cable TV...and even more places that don't have both. It's nice to have something avilable, but even nicer to have more then one thing so they can compete...
Well to the extent that I think the goverment can ever help the "free" market out...yes, yes it is. DirecTV was beign sold, and the choices were to sell it to the only other USA satalite TV system, or to sell it to some media content conglomarate. One of those radically reduces the consumers avilable choices of satalite TV delevery (from "not much choice" to "no choice"), and even if you accept DISH/DTV's viewpoint that they compete with Cable TV systems then the choice goes from 3 to 2 in most places (some places from 4 to 2, a few places from 2 to 1 still).
So from the consumers viewpoint at least in the short to mid term DirecTV being bought by Sky, er, Fox, er, Murdoch is better then DISH. In the long term? Well who knows, it might have been better to let the 2 satalite componies merge and ovver more channels or something.
It might be a temporary anomaly in mobile pricing, but it is a permeant feature of the underlying mobile cost. If you are an orange customer calling another orange customer the call can (and most likely does) run entirely over a network that orange has payed for, and doesn't cost orange anything. If you are an orange customer and call an O2 customer then you have to pay O2 to put the call through, and you may have to pay someone other then O2 to bring the voice traffic to O2.
So if the "anomaly in mobile pricing" goes away, it will vanish because same network call prices will go up.
One USA moble provider has yet another issue. Their cell phones can actually make direct handset to handset "intercom" calls. I think they make those calls directly like walkie-talkies (there is a distance limit). They charge $0.00 per minute for this.
Yeah, but if there is a business model that includes profit there will be a lot more wireless sites to use (if you are willing to pay), and it shouldn't have a big impact on the number of free ones.
Having a model where some folks "give it away" is great, the lack of a for-profit model isn't great.
Sure, but the way AT&T cable was built means it is 40+ different cable systems, right down to the CPE. So they need 40+ different TiVo boxes to get the same sort of features that DirecTV gets (high level of integration so you don't need to tell the TiVo what channes you do/don't get, and so it never misses a channel change, and so it can record mroe then one thing at once...)
Cable companies like TiVo, they see the much much lower churn rate DTV gets from it's DirecTV w/ TiVo users then the normal ones. They just can't get the same product because they screwed themselves in the past by making sure no 2 cable companies used the exact same broadcast system.
Well Jobs (apparently) spent 4 hours a day (presumabbably not for life, but for more then a few days) working with the lighting people. One assumes that is 4 hours a day he can't spend on the phone to Moto saying "get me a faster CPU!", or to IBM saying "get me POWER4's I can afford to sell!"
Yes, without reservations of any kind.
You aready do. There are a lot of products that have referal programs (for example the phone company I use will give you and I both $40 worth of service if you mention my name when you sign up -- p.s. it's a great service, you should buy it, I'm sure it will be just perfect for you. what are you waiting for? Buy! Buy! Buy!). Other products that I just happen to like, but don't seem to get as much traction in the market place as I think they need, I might go out and shill for them in the hopes that they get enough busness to stay afloat (you need a TiVo, really, go get one, or another one as appropriate, hurry! Stores might run out! Don't get cought in a shortage!).
Add photographers to that list. I have shot close to 1G worth of still images in a day, and that is using lossy compression and a "mere" 3Mpixel camera.
Lots of older drives failed too. The drive in the Lisa (aka Mac XL...well, once new ROMs were put in) was prone to failure. I remeber the double eagle drives in the late 80s being failure prone too.
I want my car to be able to drive to mars, but it doesn't seem likely, nor does it seem likely that a highly complex mechinical device with exceptionally tight tolerences and moving parts will be as reliable as a solid state device.
I quite want a reliable drive too. Fast would be nice, very arge storage capacity would be very very nice. Affordable woud also be nice. In fact it doesn't have to be a drive, it could be something that acts like one, but FLASH ROM is still too slow (to erase, reads are fast if you design right) and too costly.