I know this is a trite argument, but it is appropriate (and no, I'm not trying to flame anybody, or even single you out or anything):
If you really think another way is the way to go, please start doing it. You don't even need to be a programmer yourself; write up a paper detailing the failings of the current efforts and propose a better way. Disseminate this text, and persuade other coders to join in and implement it. Even if you do not succeed in getting your project started, your work will not be wasted as your analysis will be helpful in guiding the current projects.
The people working on Mono, DotGNU and Portable.NET are all doing it because they believe their project is the right way to go about it. Any productive feedback - in the form of a design document or a competing project - is very helpful for all involved. A random 'I don't like this', on the other hand, is likely to be ignored.
He isn't asking about the moral issues, he wants to know the legal aspects - these are not always congruent, you know.
I vaguely seem to remember that where I live (sweden) you are not free to redistribute or publish stuff that's gotten into your hands by mistake if the stuff is clearly sent to you by mistake or is obviously confidential. We've had some incidents where hospitals or social services have faxed journals and other files to private citizens by mistake, and I think that was the result of those incidents. Note that you are not required to destroy the documents, or alert anybody that the information's got astray, you just aren't allowed to spread it around.
Yes, sometimes you do need JavaScript (my bank is the same). I've found, however, that very few of my regular or semi-regular sites need Javascript, and those few I can easily add to the exception list for Mozilla.
Most sites that require a login will work quite happily if you log in as usual and allow them to set a couple of cookies. Then you can turn everything off, and the site will continue to work fien everytime you go there, as long as you make sure to keep the cookies it set that first time.
I agree with the above, and would suggest the following to supplement these:
Tanenbaum, "Structured Computer Organisation" - a very well written, accessible volume explaining how computers are actually organized.
Rich, Knight, "Artificial Intelligence" - a good introductory text to this field.
Glassner, et. al., "Graphics Gems" - a series of volumes that (if you can afford them) collects everything you need graphics-wise in an encyclopedic, rather than narrative, format.
Boolos, Jeffrey, "Computabgility and Logic" - Another introductory text that is great reading for anybody interested in the field of computability (and you even get to prove Gödel's theorem!).
There are others that are true classics (the Cinderella book comes to mind) but as you are asking for titles for a general-interst library they might be a bit too obscure.
Your point about the oppressed Gnome being the secret reason the Linux community does not want to throw its weight behind a Microsoft developer standard is... a strange one.
You misunderstood, perhaps. It's not the adoption of.NET or not that I'm reacting to, it's the _very_ harsh language used, aimed at.NET, mono, Ximian, Miguel, and Gnome (it's been even worse at LinuxToday). Everybody involved has been painted as craven traitors, and the language used has at times been just shy of character assassination. I've been trying to understand where this reaction stems from, and it seems it is for some people an opportunity to attack a project they see as an "enemy".
The OpenPrivacy project would like the announce the creation and initial release (0.0.1) of Reptile . Reptile is an Open Source/Free Software, Peer-to-Peer, content syndication engine (think RSS/OCS), which is driven by Java/XML and is privacy protection/Reputation enabled. Reptile nodes can publish to each other (everything is driven by XML based subscriptions) and provides a decentralized authentication model based on public/private key crypto (and Reputation).
To paraphrase a classic line:
"Your Honour, that sentence should be taken out and shot."
A group of open-source developers have started a project to implement a set of programs and protocols that have been submitted to a standards comittee. Like so many other open-source projects, others are free to use it - or to help developing it - or not as they choose. This is nothing strange; this has happened many times before. So why this sudden outpouring of emotions, and even, in a few cases, hate?
Well, is it becuse the protocols are under another entities control? No, it can't be that; after all, Java Run Time environments, and GNU:s gcc Java compiler, have not received this reaction, even though Java is not only under the control of a corporation, but isn't even submitted as an open standard.
Is it because the controlling entity happens to be Microsoft? No, or the SAMBA developers would have been boiled alive a long time ago.
Are people perhaps afraid that this will giveMicrosoft a foothold to subvert the open-source movement? Maybe, but not likely; the existence of Samba, for instance, have only resulted in more Linux servers being deployed than would have been the case otherwise. It is likely that, by allowing other services to run on Linux, we will increase its prevalence further - and this time on the desktop.
Is it because it happens to be developed by some of the people that also work on Gnome? Well, it does seem that way, unfortunately. We have not seen the same backlash against FSF for their DotGNU, after all. What seems to be happening here is that some people have taken an ABG (Anything But Gnome) approach to their lives, and will take any opportunity to help their desktop of choice "win". Of course, the same immature personalities exist among Gnome users and among any even remotely divisive application/framework/whatever - just look at how long and bitter the Emacs/Vi flamefests were before a realisation crept in that they can peacefully coexist.
This would not have blown up like this, of course, had not Petreley written a very inflammatory, factually erroneous piece in Inforworld last week. For those of us who read him every week (I've stopped since that last piece), he has, ever since he started to use KDE, used every opportunity to attck Gnome for any reason he has been able to come up with. He is, in fact, a good example of the ABG crowd that seems incapable of lifting his eyes a bit and realize that the choice of desktop isn't really important (after all, you can run either, both or none without problems).
To summarize: no,.NET is not a danger; it might be beneficial to Linux; it's not connected to Gnome other than being another Gnome application; even if you hate.NET, don't connect it with GNome; it won't hurt KDE or other desktops; you don't have to use it if you don't want.
The moment that we become dependent on software with restrictive licensing is the moment that this whole party goes down the toilet.
You mean, like depending on Java, as some people suggest as an alternative?
The.NET specification is submitted as a standard. Mono is GPL. The people implementing it _are_ a part of the GNU/Linux community. What more do you want?
If the worst-case scenario happens, and MS finds a way to block implementation of mono, then what? Well, we are back at square one, without having lost anything. And as Miguel said, the only way would be to patent specifica ways to implement it, which still leaves developers free rein to implement it in a patent-free manner.
Add *that* to the fact that all most no open-source developer would even consider using Gnome...
OK, I was going to stay out of this flamefest, but I just can't.
To begin with, the only connection.NET has to Gnome is that it's initially being developed on Gnome as a platform - there is nothing inherently Gnome:ish to it.
Second, statements like the one above are rude, inflammatory, and - in this case - patently wrong. I suggest you do a comparison (pool data from Freshmeat and the Gnome and KDE sites), and you will find more applications for Gnome than KDE, and _much_ more apps written using GTK than Qt. I know that checking of facts before posting here at/. seems to be against peoples religion or something...
Unfortunately, the next step is probably to force ISP:s to filter out peer-to-peer traffic. No way to connect = kill the P2P networks... And even if you're not directly affected (not living in the US, say), you still get a far smaller selection of stuff on the networks.
The logical next step for P2P networks is then to tunnel stuff through other services. Gnutella/HTML, anybody?
Let's say I write a paper (which I'm supposed to do anyway, because it's part of my job). I send it in to a journal - without getting paid anything by the journal - to get the result out, and to get another line in my CV. They farm that paper out to a reviewer. That reviewer is sometimes paid, but more often than not, they only get _their_ datapoint in their resume. If it's accepted, I edit the paper and resend it. After a few iterations it is accepted in its final form. Then I often have to format the paper according to the standards of the journal (sometimes the journal does this step) and sign over all rights to that text to the publisher. Eventually it gets printed.
After all this, where most of the work is done unpaid and outside of the publisher, they can still charge enough for the journal that my department cannot afford to actually get a copy... And today, prices are getting high enough that not even the university library will take in the more expensive journals without a massive interest among employees.
Only problem is that the FAQ dates from sometimes last year, and others have indicated that this development has stalled, so there's really no knowing when (or even if) it will really happen.
I've seen the Delphi environment under Windows, and it's fairly sweet. Object Pascal is a much nicer language than C++ for development and prototyping. And don't forget that you can code GTK+ stuff as well if you want to.
Now, if only they'd ship the C++Builder stuff as well so Linux will get the NeverWinter Nights toolset...
Nothing to be upset about. The very reason that Yellowstone is such a biologically and geologically interesting place is because of these periodic eruptions. If the activity wasn't there, neither would Yellowstone.
It's the same way with earthquakes, really. Sometimes people wonder aloud what possessed people to live in places with frequent earthquakes and/or volcanoes, and the answer is that the geological activity - or, rather the result of that activity - is the very reason it's a good place to live. You get a varied seafloor with lots of marine life, natural deep harbours, very fertile volcanic soil and varied terrain where it's easy to find secluded, protected sites on which to build.
If it's not in the database, it doesn't exist, and will thus not need to be disposed of.
"MS announced today that the next version of SQL Server will have a tenfold increase in its lauded 'virtual disposal' capability as compared to previous versions. 'This is a huge step forward for our environmental policy', president Bush stated, and added 'It's another example of how business takes care of environmental concerns far more effectively than government regulations can do'."
This was actually the firsat I've heard of that happening...
Actually, Region 1 DVDs aren't illegal to sell in themselves. There is a law about 'gray imports' or parallell imports, however, where a local company that's the official reseller of foreign companies' goods (like movie distributors) have an exclusive right to import those goods. I think that's what they mean by 'illegal'.
However, this law only applies for importing for commercial purposes (like reselling), so nothing can stop me from ordering DVDs from USA or anywhere for myself, as long as I don't do it with the intent of reselling them and making a buck.
Also, that law is fairly controversial - granting monopoly rights, screwing over consumers and all that, and there are signs that it might be overturned. This is partially for practical reasons (some medicines sell for less than a third of the Swedish price in southern Europe, and hospitals aren't allowed to shop there due to this law) and partially for consumer benefit.
At least here in sweden, one selling point has become that the player is region-free, i.e. that it it ignores the region coding, or can be set to whatever region you please. At first players needed to be modded, but today all but a few name brand players usually come region-free out of the box.
With consumers becoming used to the idea of buying copyright-avoiding technology, and manufacturers seeing there is a very large market for it, I'd expect tv-sets and VCRs that ignore this as well.
Remember, most manufacturers are not content providers, and has little incentive not to do this, especially with competitors taking market share with their products.
I was commenting on the fact that sites have not tried _individual_ subscriptions, by and large, not the harebrained idea of making ISP:s pay for users' content.
Well, that's not quite true. They haven't given up as much as largely been afraid to ask in the first place. There is a lot of 'conventional wisdom' that states that nobody will pay for content (excepting those sites that _do_ charge for content and makes a living from it).
A few mainstream sites have begun trying a subscription model (Salon is an example, pr0n is another). From the site operators perspective is that they probably will lose a good deal of their readers; the question is if they lose too many, making the model a losing proposition or not.
From the POV of us surfers, the question becomes wether we are prepared to pay for the offered content. That answer is probably heavily dependent on what the content is and how much it would cost. If it's information readily available at fifty other locations, we would probably not bother, while if it's a site with unique material - and that we have become dependent on - many might well be prepared to pay. That given, for newssites (like dn or nytimes), I'd expect to get the full text from the printed version and no banner ads.
This also ties in to the question of payment on the net in general, of course. It doesn't matter if the enformation is perceived as very cheap, if the hassle to actually pay for it is too great.
With the demise of some sites recently, and a subscription model being tried for others, I've started to really think about what sites I'd be prepared to pay for. In the end, it turns out that there's about four regular sites that'd get my money (given a reasonable fee), while another dozen would be dropped.
Hate to rain on your parade, but competition will _not_ solve any problem... If a problem will take a minimum of theoretical and practical advances, there is a practical limit as to how quickly it will be solved (an if it requires a lot of advances, it will take a lot of time no matter how much money you throw at it). If a problem has no solution, no solution will be found.
For that five to eight year gap, there is plenty to do still to enable us to cram evermore Divx files and mp3:s on our systems.
Just because the drives stop becoming bigger, it won't stop them from becoming cheaper (up to a point). Instead of larger drives, we can always turn to adding more of them into a system (which is sort of what the article implies).
Another way to enlarge the storage would be to bring back the 5 1/4 inch drives - more area = more storage. Add an ultrafast 2.5 inch drive as a cache, and you may get performance almost as good as current drives.
Of course, this applies mostly to individual systems, where there is room to grow in various ways. For large serverfarms, already scrambling for space and power, this problem is rather more serious. Probably, virtualized storage is the only way forward at that point; there are of course other advantages to that technology besides space, so it may not affect organisations that badly as many of them would be moving towards this technology anyway.
Slashdot Mirror
I know this is a trite argument, but it is appropriate (and no, I'm not trying to flame anybody, or even single you out or anything):
If you really think another way is the way to go, please start doing it. You don't even need to be a programmer yourself; write up a paper detailing the failings of the current efforts and propose a better way. Disseminate this text, and persuade other coders to join in and implement it. Even if you do not succeed in getting your project started, your work will not be wasted as your analysis will be helpful in guiding the current projects.
The people working on Mono, DotGNU and Portable.NET are all doing it because they believe their project is the right way to go about it. Any productive feedback - in the form of a design document or a competing project - is very helpful for all involved. A random 'I don't like this', on the other hand, is likely to be ignored.
/Janne
He isn't asking about the moral issues, he wants to know the legal aspects - these are not always congruent, you know.
I vaguely seem to remember that where I live (sweden) you are not free to redistribute or publish stuff that's gotten into your hands by mistake if the stuff is clearly sent to you by mistake or is obviously confidential. We've had some incidents where hospitals or social services have faxed journals and other files to private citizens by mistake, and I think that was the result of those incidents. Note that you are not required to destroy the documents, or alert anybody that the information's got astray, you just aren't allowed to spread it around.
/Janne
Yes, sometimes you do need JavaScript (my bank is the same). I've found, however, that very few of my regular or semi-regular sites need Javascript, and those few I can easily add to the exception list for Mozilla.
Most sites that require a login will work quite happily if you log in as usual and allow them to set a couple of cookies. Then you can turn everything off, and the site will continue to work fien everytime you go there, as long as you make sure to keep the cookies it set that first time.
/Janne
I agree with the above, and would suggest the following to supplement these:
Tanenbaum, "Structured Computer Organisation" - a very well written, accessible volume explaining how computers are actually organized.
Rich, Knight, "Artificial Intelligence" - a good introductory text to this field.
Glassner, et. al., "Graphics Gems" - a series of volumes that (if you can afford them) collects everything you need graphics-wise in an encyclopedic, rather than narrative, format.
Boolos, Jeffrey, "Computabgility and Logic" - Another introductory text that is great reading for anybody interested in the field of computability (and you even get to prove Gödel's theorem!).
There are others that are true classics (the Cinderella book comes to mind) but as you are asking for titles for a general-interst library they might be a bit too obscure.
/Janne
Oh no; I could not aspire to the lofty heights of that sentence! It is magnificient in its enormity; sort of a linguistic Matterhorn.
Just to clarify: I'm kidding - it's no worse (and probably a bit better) than any other blurbs you tend to see on sites like this.
/Janne
Your point about the oppressed Gnome being the secret reason the Linux community does not want to throw its weight behind a Microsoft developer standard is... a strange one.
.NET or not that I'm reacting to, it's the _very_ harsh language used, aimed at .NET, mono, Ximian, Miguel, and Gnome (it's been even worse at LinuxToday). Everybody involved has been painted as craven traitors, and the language used has at times been just shy of character assassination. I've been trying to understand where this reaction stems from, and it seems it is for some people an opportunity to attack a project they see as an "enemy".
You misunderstood, perhaps. It's not the adoption of
/Janne
The OpenPrivacy project would like the announce the creation and initial release (0.0.1) of Reptile . Reptile is an Open Source/Free Software, Peer-to-Peer, content syndication engine (think RSS/OCS), which is driven by Java/XML and is privacy protection/Reputation enabled. Reptile nodes can publish to each other (everything is driven by XML based subscriptions) and provides a decentralized authentication model based on public/private key crypto (and Reputation).
To paraphrase a classic line:
"Your Honour, that sentence should be taken out and shot."
/Janne
A group of open-source developers have started a project to implement a set of programs and protocols that have been submitted to a standards comittee. Like so many other open-source projects, others are free to use it - or to help developing it - or not as they choose. This is nothing strange; this has happened many times before. So why this sudden outpouring of emotions, and even, in a few cases, hate?
.NET is not a danger; it might be beneficial to Linux; it's not connected to Gnome other than being another Gnome application; even if you hate .NET, don't connect it with GNome; it won't hurt KDE or other desktops; you don't have to use it if you don't want.
Well, is it becuse the protocols are under another entities control? No, it can't be that; after all, Java Run Time environments, and GNU:s gcc Java compiler, have not received this reaction, even though Java is not only under the control of a corporation, but isn't even submitted as an open standard.
Is it because the controlling entity happens to be Microsoft? No, or the SAMBA developers would have been boiled alive a long time ago.
Are people perhaps afraid that this will giveMicrosoft a foothold to subvert the open-source movement? Maybe, but not likely; the existence of Samba, for instance, have only resulted in more Linux servers being deployed than would have been the case otherwise. It is likely that, by allowing other services to run on Linux, we will increase its prevalence further - and this time on the desktop.
Is it because it happens to be developed by some of the people that also work on Gnome? Well, it does seem that way, unfortunately. We have not seen the same backlash against FSF for their DotGNU, after all. What seems to be happening here is that some people have taken an ABG (Anything But Gnome) approach to their lives, and will take any opportunity to help their desktop of choice "win". Of course, the same immature personalities exist among Gnome users and among any even remotely divisive application/framework/whatever - just look at how long and bitter the Emacs/Vi flamefests were before a realisation crept in that they can peacefully coexist.
This would not have blown up like this, of course, had not Petreley written a very inflammatory, factually erroneous piece in Inforworld last week. For those of us who read him every week (I've stopped since that last piece), he has, ever since he started to use KDE, used every opportunity to attck Gnome for any reason he has been able to come up with. He is, in fact, a good example of the ABG crowd that seems incapable of lifting his eyes a bit and realize that the choice of desktop isn't really important (after all, you can run either, both or none without problems).
To summarize: no,
/Janne
The moment that we become dependent on software with restrictive licensing is the moment that this whole party goes down the toilet.
.NET specification is submitted as a standard. Mono is GPL. The people implementing it _are_ a part of the GNU/Linux community. What more do you want?
You mean, like depending on Java, as some people suggest as an alternative?
The
If the worst-case scenario happens, and MS finds a way to block implementation of mono, then what? Well, we are back at square one, without having lost anything. And as Miguel said, the only way would be to patent specifica ways to implement it, which still leaves developers free rein to implement it in a patent-free manner.
/Janne
Add *that* to the fact that all most no open-source developer would even consider using Gnome...
.NET has to Gnome is that it's initially being developed on Gnome as a platform - there is nothing inherently Gnome:ish to it.
/. seems to be against peoples religion or something...
OK, I was going to stay out of this flamefest, but I just can't.
To begin with, the only connection
Second, statements like the one above are rude, inflammatory, and - in this case - patently wrong. I suggest you do a comparison (pool data from Freshmeat and the Gnome and KDE sites), and you will find more applications for Gnome than KDE, and _much_ more apps written using GTK than Qt. I know that checking of facts before posting here at
/Janne
Unfortunately, the next step is probably to force ISP:s to filter out peer-to-peer traffic. No way to connect = kill the P2P networks... And even if you're not directly affected (not living in the US, say), you still get a far smaller selection of stuff on the networks.
The logical next step for P2P networks is then to tunnel stuff through other services. Gnutella/HTML, anybody?
/Janne
Let's say I write a paper (which I'm supposed to do anyway, because it's part of my job). I send it in to a journal - without getting paid anything by the journal - to get the result out, and to get another line in my CV. They farm that paper out to a reviewer. That reviewer is sometimes paid, but more often than not, they only get _their_ datapoint in their resume. If it's accepted, I edit the paper and resend it. After a few iterations it is accepted in its final form. Then I often have to format the paper according to the standards of the journal (sometimes the journal does this step) and sign over all rights to that text to the publisher. Eventually it gets printed.
After all this, where most of the work is done unpaid and outside of the publisher, they can still charge enough for the journal that my department cannot afford to actually get a copy... And today, prices are getting high enough that not even the university library will take in the more expensive journals without a massive interest among employees.
/Janne
Only problem is that the FAQ dates from sometimes last year, and others have indicated that this development has stalled, so there's really no knowing when (or even if) it will really happen.
/Janne
I've seen the Delphi environment under Windows, and it's fairly sweet. Object Pascal is a much nicer language than C++ for development and prototyping. And don't forget that you can code GTK+ stuff as well if you want to.
Now, if only they'd ship the C++Builder stuff as well so Linux will get the NeverWinter Nights toolset...
/Janne
Nothing to be upset about. The very reason that Yellowstone is such a biologically and geologically interesting place is because of these periodic eruptions. If the activity wasn't there, neither would Yellowstone.
It's the same way with earthquakes, really. Sometimes people wonder aloud what possessed people to live in places with frequent earthquakes and/or volcanoes, and the answer is that the geological activity - or, rather the result of that activity - is the very reason it's a good place to live. You get a varied seafloor with lots of marine life, natural deep harbours, very fertile volcanic soil and varied terrain where it's easy to find secluded, protected sites on which to build.
/Janne
If it's not in the database, it doesn't exist, and will thus not need to be disposed of.
"MS announced today that the next version of SQL Server will have a tenfold increase in its lauded 'virtual disposal' capability as compared to previous versions. 'This is a huge step forward for our environmental policy', president Bush stated, and added 'It's another example of how business takes care of environmental concerns far more effectively than government regulations can do'."
Sorry.
/Janne
It's not rotten, it's fermented, just like yoghurt or sour cream. The smell is a bit off-putting, but the taste is great!
/Janne
Hey, lutfisk is great with potatoes and a mustard sauce! Even better is surströmming, though foreigners tend to resist that dish for some reason :-)
/Janne
Oh, yeah, IANAL, btw.
/Janne
This was actually the firsat I've heard of that happening...
Actually, Region 1 DVDs aren't illegal to sell in themselves. There is a law about 'gray imports' or parallell imports, however, where a local company that's the official reseller of foreign companies' goods (like movie distributors) have an exclusive right to import those goods. I think that's what they mean by 'illegal'.
However, this law only applies for importing for commercial purposes (like reselling), so nothing can stop me from ordering DVDs from USA or anywhere for myself, as long as I don't do it with the intent of reselling them and making a buck.
Also, that law is fairly controversial - granting monopoly rights, screwing over consumers and all that, and there are signs that it might be overturned. This is partially for practical reasons (some medicines sell for less than a third of the Swedish price in southern Europe, and hospitals aren't allowed to shop there due to this law) and partially for consumer benefit.
/Janne
At least here in sweden, one selling point has become that the player is region-free, i.e. that it it ignores the region coding, or can be set to whatever region you please. At first players needed to be modded, but today all but a few name brand players usually come region-free out of the box.
With consumers becoming used to the idea of buying copyright-avoiding technology, and manufacturers seeing there is a very large market for it, I'd expect tv-sets and VCRs that ignore this as well.
Remember, most manufacturers are not content providers, and has little incentive not to do this, especially with competitors taking market share with their products.
/Janne
I was commenting on the fact that sites have not tried _individual_ subscriptions, by and large, not the harebrained idea of making ISP:s pay for users' content.
Well, that's not quite true. They haven't given up as much as largely been afraid to ask in the first place. There is a lot of 'conventional wisdom' that states that nobody will pay for content (excepting those sites that _do_ charge for content and makes a living from it).
A few mainstream sites have begun trying a subscription model (Salon is an example, pr0n is another). From the site operators perspective is that they probably will lose a good deal of their readers; the question is if they lose too many, making the model a losing proposition or not.
From the POV of us surfers, the question becomes wether we are prepared to pay for the offered content. That answer is probably heavily dependent on what the content is and how much it would cost. If it's information readily available at fifty other locations, we would probably not bother, while if it's a site with unique material - and that we have become dependent on - many might well be prepared to pay. That given, for newssites (like dn or nytimes), I'd expect to get the full text from the printed version and no banner ads.
This also ties in to the question of payment on the net in general, of course. It doesn't matter if the enformation is perceived as very cheap, if the hassle to actually pay for it is too great.
With the demise of some sites recently, and a subscription model being tried for others, I've started to really think about what sites I'd be prepared to pay for. In the end, it turns out that there's about four regular sites that'd get my money (given a reasonable fee), while another dozen would be dropped.
/Janne
Hate to rain on your parade, but competition will _not_ solve any problem... If a problem will take a minimum of theoretical and practical advances, there is a practical limit as to how quickly it will be solved (an if it requires a lot of advances, it will take a lot of time no matter how much money you throw at it). If a problem has no solution, no solution will be found.
/Janne
For that five to eight year gap, there is plenty to do still to enable us to cram evermore Divx files and mp3:s on our systems.
Just because the drives stop becoming bigger, it won't stop them from becoming cheaper (up to a point). Instead of larger drives, we can always turn to adding more of them into a system (which is sort of what the article implies).
Another way to enlarge the storage would be to bring back the 5 1/4 inch drives - more area = more storage. Add an ultrafast 2.5 inch drive as a cache, and you may get performance almost as good as current drives.
Of course, this applies mostly to individual systems, where there is room to grow in various ways. For large serverfarms, already scrambling for space and power, this problem is rather more serious. Probably, virtualized storage is the only way forward at that point; there are of course other advantages to that technology besides space, so it may not affect organisations that badly as many of them would be moving towards this technology anyway.
/Janne