>If Oracle or whoever and code their databases for SMP, then why can't OS vendors or game manufacturers?
Databases can only do certain operations in parallel, and some of those require additional optimizations like separate drives to operate at maxmimum efficiency. An OLTP database with 10000 users is an inherently parallel problem domain.
OS vendors do code for SMP. Not all of them code for large numbers of processors, though, or they may not have succeeded at the high end. Sometimes there are performance trade-offs between uniprocessor and SMP performance, or between wall-clock execution time and machine throughput. If you're designing a big honkin' database server, you might have to make a 20ms query take 40ms due to DB and OS locking overhead, but if you get 3x the overall system throughput, you won. That trade-off might not make sense on a uniprocessor commodity desktop, so you find that apps and OSs intended for desktop hardware don't really know what to do with an 8-way machine.
Anyway, I'm not trying to say that it's futile, I'm just trying to point out that it's non-trivial, and will take some time. It's pretty clear that scaling sideways is going to be necessary more and more (more threads instead of just jacking up the clock speed), so this is a challenge that software developers are just going to have to learn to deal with.
>Im a software developer and REALLY hate the movement towards dual-cores.
Tough. Chip makers are up against a technology barrier right now, and clock speed increases in the CPU don't make RAM or disk or interconnect faster anyway. How about just putting a 4MB cache on-die? That wouldn't require a massive clock speed increase but it would speed things up. I'm not an EE but I'm just pointing out that there are many, many things that have been left in the dust by Moore's law that could catch up and make quite a difference. Does your computer have 4+GB of DDR memory? ATA-133 drives with 8MB cache? PCI-X? A 64-bit CPU and an OS that knows how to use it fully? In what other ways are CPUs waiting on everything else, that could be improved to make things run faster overall?
Learn to parallelize your code where possible. Optimize your existing code. Software optimizations yield stunning improvements compared to incremental clock speed bumps anyway, and (unlike hardware) affect every installation of your app.
>Its a bad move IMO on AMDs and Intels part
OK genius, what's the alternative? No improvements in processors for years, until somebody makes a breakthrough that enables 4+ GHz processors? What happens when they hit the next roadblock?
Hardware has been so far ahead of software for so long that we've become accustomed to solving bloat with "just buy a new computer". It wouldn't kill us to spend a little time profiling code. The economics have been (in many cases) such that it just made more sense to throw money at new hardware. If that no longer makes sense, throw money at software optimizations for a little while. It doesn't exclusively mean that we have to force every algorithm to operate in parallel. It could be as simple as releasing fat binaries of apps that are compiled to target recent CPUs (no more shipping 386-optimized code to every customer), or *gasp* writing more efficient code in the first place.
There's only one problem with your list: all of these things are COMPLETELY unreasonable to ask every single user to do. This is what vendors are for.
A better approach comes from the vendor: 1) Install nonessential software with a disabled configuration. If they (or an app they use) need it, they can enable it then.
2) Enable automatic software updates. If you have a vendor trust issue with this, turn it off on your computer and those that you have influence over. If the computer isn't connected to the internet, nag the user to get an update (vendor-provided near-free update CD + a reasonable S&H fee, or via LAN, etc.) once every few months.
3) Add a host firewall, and make the OS work OK with the host firewall enabled. Provide simple high-level checkboxes for people who want to allow incoming connections to high-level services, and warn them if they do.
4) Add scripts (which get updated with the auto software update) that know what traffic is noise and what traffic looks like a successful attack or exploit. Send interesting packets upstream to the vendor for analysis. If you have a problem with this, turn it off for computers you control. A computer with always-on internet access is constantly sprayed with random attack packets; there's no point in wasting the user's time. Hmm, a connection attempt to port 25 on a laptop that has no SMTP server running. There's no point in wasting the user's time with this information. At most, increment a counter and leave it at that.
Actually, Windows XP SP2 has made some of these improvements. The problem is that the underlying services are also insecure, and the services are so entangled with one another that you can't really function on a LAN without opening yourself up.
> So...you're basically saying thin client is a great idea that has been hampered by poor implementations.
That's not at all what the parent post said.
It's an idea that sounds good until you think about the fact that you're intentionally buying something that's less functional. Then, people who aren't looking for an appliance (cash register, kiosk, etc.) start thinking about all of the things that a personal computer does well, and they buy it. See also: the "Mac OS X and Linux don't have any applications so I must buy Windows" rationale.
I've said it before 'n I'll say it again: the people who are pitching thin clients and network computers are the people who make their money off of large-scale server hardware and software. Just as Intel encouraged modem manufacturers to push processing loads onto the CPU (winmodems) to sell CPUs, NCs and thin clients are largely an invention of Sun and Oracle (and their ilk) who want to encroach on Wintel's revenues.
Oh yes, life will be so very much cheaper when you get rid of all of those silly little x86 servers and desktop PCs, and buy a bunch of proprietary, non-commodity-priced clients that will use gobs of network bandwidth to run apps on proprietary, non-commodity-priced servers that we just happen to sell, using a proprietary, non-commodity-priced app virtualization scheme (and perhaps even a proprietary remote GUI technology!).
Google might try a more modern, open approach to this: some kind of DHTML app suite, with a super cheap subscription, not based on proprietary hardware or (client-side) software. That's the only speculation I've heard about this that is in the least bit plausible, but there are still issues.
1) Yahoo has been in a position to do this for years. Why haven't they? They have a bunch of light-HTML apps, and a bunch of money. They could build heavy-HTML apps too. I don't believe that the idea just hasn't occurred to them yet.
2) Who the hell would use it? People like PCs largely because they're personal. Nobody wants to push their precious secret nekkid vacation pictures and financial data out onto some random server out there in virus land. Try telling folks that not only do they have to get a PC, but also install a new web browser, and that buys them the privilege of paying even more to run shitty apps that aren't Microsoft Office and that are on the far end of a slow, laggy pipe. WHERE DO I SIGN??!!?
Scott McNealy's strategy of "tons of people will buy this, because by doing so they will make us rich" just doesn't work.
BTW, I do think that thin clients make great sense for specific, limited applications. They just aren't the PC-killer for typical office and home users that NC and thin client pitchmen wish they were.
Given the law of averages, the fact that there is no such thing as a "law of averages" means that it's likely that eventually one will spontaneously coalesce into existence.
At this time, Marilyn vos Savant will marry Monty Hall and they will rule the universe.
Remember that G3, G4, and G5 are just Apple marketing terms. They can take a G4 with a bigger cache and call it a G5 if they want, or take a G5 and yank out a bunch of the superscalar bits to make it use less power.
256MB of the correct memory costs $48 with free shipping from Monarch Computer. So, we're talking about a difference in price ($27) that's about as big as the sales tax on the Mac mini.
(Because it's Slashdot, some dickhead is going to loudly point out that the Mac mini appears to have a single DIMM socket, so it'd really be a 512MB DIMM replacing a 256MB DIMM, so the price would theoretically be slightly less if Apple offered a 0MB option. Now I've said it so you can apply your brilliant insight elsewhere.)
If you want a big computer that you can open up and service yourself, get one. If you want a teeny tiny computer that has everything crammed in as tightly as it can be, while still using industry standard parts, don't be shocked when you find out that some skill is required to install the stuff.
Or, you can just wait for a web site to publish the directions, and do it yourself, and take your chances with breaking something. If you do it properly, they'll never know. If not, well, maybe you'll wish you hadn't been so cheap.
If all you want to do is look at your hardware and run benchmarks (a.k.a. 3D FPS games that bore anyone over 12 years old), that's great.
Watch the keynote and look at the software. Apple is not trying to make a Windows PC that looks nice. They're making software too, and leading the pack in many categories.
Most people actually want to accomplish tasks with their computers, and a few percentage points in hardware price/performance don't make nearly as much difference as better usability (which is measured in units of time to accomplish a specific task).
But, if you just like to spend your whole day messing around with your computer instead of getting anything done, by all means, avoid Macs.
Or, if you believe what Steve Jobs said in the keynote, they found that people use shuffle anyway, so they just focused on that, and took away the feature of random access. You can turn shuffle off and just use single-song fast-forward and rewind, but with 120 songs that might get pretty annoying.
I had a Rio 500 128MB flash player before I bought an iPod, and with only about 2 albums' worth of space on it, just FF and Rew controls were fine. There was a wheel, and a whole complicated UI, with EQ and bookmarks and such, but really, with about 20 or 30 songs (2 albums) tops, there isn't a need. It's better to just leave it in your pocket and hit the FF button a lot.
I think I'd miss the EQ, and selecting albums. But that's part of why I won't sell my 20GB iPod to buy an iPod shuffle.:)
I suspect that most people, who are song oriented and don't play albums in their entirety like I do, will embrace this and love the whole "what am I going to be hearing today" idea. After all, if you don't exhaust the 120 songs before your next iTunes reloading session, it's effectively shuffling from your entire collection since you won't hear a song repeated.
Plus that $99 thing is handy. Amazon and Best Buy and similar retailers will almost certainly undercut that by a few bucks.
Hellooooo.... ISPs sell bandwidth. DSL costs more than dialup. A T3 costs more than DSL. Don't take my word for it - check with your ISP. It's actually true.
You might as well say that a gas station owner would do well to ban trucks and SUVs becuase they're fuel hogs.
Yeah, the short filenames thing is really annoying. We've had long filenames on OS X for how many years now? How hard is it to just change the code to allow me to save a file with a long filename from Office? Apparently, it's much too hard.
Word X for Mac with a blank new document needs 38MB RAM. Not "200MB".
It does do that annoying "use some CPU for nothing in particular" thing, though. That's pretty ridiculous. (Oh wait we're on/., the correct localized spelling is "rediculous". My bad.)
>If the primary objective was to maintain an OS monoploy, and if it believed Netscape was going to >release a competitive OS (a doubtful proposition) the most direct course of action would have been to buy Netscape.
No. Netscape was worth a lot back then. That might have been "direct", but it wasn't the most profitable approach. That would have been a very foolish and expensive solution.
Instead, Microsoft bought (or licensed rights to? I don't remember) a much less successful browser and built it into a Netscape competitor on the client side, but how they really preserved their monopoly was by killing Netscape's actual revenue stream by bundling IIS for free with NT 4. There was a significant market for commercial web servers back then; Netscape Enterprise server cost about a thousand dollars. SuiteSpot cost several thousand dollars. Lots of companies actually bought it and used it for a while; in many ways it was far more advanced than Apache (which was pretty new at the time), blew away CERN and NCSA httpd, and worked quite well running on NT 4.0.
Microsoft knew that Netscape was around the corner from making the OS irrelevant via HTML, JavaScript, and Java. Netscape was explicitly trying to do this, calling it "crossware", meaning cross platform software that didn't depend on a specific OS, but was built on internet and web standard technologies (standards Netscape invented or championed). Microsoft saw the threat and headed it off. This has been well documented in the DOJ and Sun cases and Microsoft was found guilty in both.
Microsoft absolutely does not have the ability to drop Windows quickly. To suggest that they do is totally absurd.
>then they started calling me and I quickly realized very few people should have a computer.
Maybe you should change your number, or start charging for your time.
>Palm has a really smart way to deal with that by >limiting any interaction with the OS and making the App king.
DOS worked that way too. People actually want multitasking, it turns out.
By the way, you still have to interact with a host OS in order to use a PDA (installing new apps, backing up data, etc.) except in the most limited case where someone is totally happy with the default apps and doesn't want to back up their data, ever. Every app I've ever seen for the Palm requires you to download it on a PC (or Mac or Linux box or...), unzip it, and install it via Palm Desktop. So, they just didn't bother duplicating that on the handheld, which is smart, but doesn't mean that the user never has to touch an OS to use a Palm.
>Plus having everything running all the time makes everything faster.
Actually, it makes things slower, unless you have a huge amount of physical memory, which Palm PDAs don't. Besides, since when does PalmOS run all installed applications simultaneously? I've written a couple of PalmOS apps, and they actually have to launch, and you can't run more than one at a time.
>The smart money is on going BACK to mainframe type applications and computing.
Not really. Sun Microsystems, clueless and sinking fast Oracle, not looking to healthy And Marimba (an early Java-apps-as-a-service company) was coughing up blood and just got bought out.
The quality of their individual products doesn't matter if their strategy sucks. (I have a lot of respect for Sun gear, Solaris, and Oracle, but that doesn't mean they're doing well in the market, or that they have a plan for how they're going to make stuff that people want for a price they're willing to pay.)
Apparently the smart money is in short-selling companies that preach the Network Computer gospel. Any year now, corporations will buy $800 boxes that do less than a PC, in order to save money. Oh don't mind that hugely expensive server and software stack that you have to run on the server, don't mind the training and limited apps. We promise, it's cheaper to have everybody sitting on their hands if the network is down, and laptops are just a fad anyway.
While bandwidth is getting cheaper, so are PCs, and NCs have essentially zero volume, so they aren't getting any cheaper at all. Meanwhile, user expectations are keeping up with PC capabilities. Maybe the NC hype will come true in the form of PCs configured as thin clients running apps on Windows servers, or Linux servers (dare we dream) but not in large numbers for a long time. Hard disk price/performance from a speed and storage POV still blows away bandwidth, and until that changes in many orders of magnitude in the favor of network being cheaper and faster than a local disk, the NC idea will remain a fantasy of the folks trying to sell you the back-end stuff at premium prices.
You can wave your hands about bandwidth being essentially free, but you'll have to back that up by installing a T3 to your house and telling me that you don't notice the monthly bill. Hard disk storage is moving towards "essentially free" (that is, it's getting cheaper) quite quickly too. What matters is not the separate rate of change of each, but the ratio. Which one is getting cheaper faster? How long will it take for them to cross over and for one to become drastically cheaper than the other, or are they diverging?
>Java (etc) have been invented so what's the wait? The wait is for that 25MB JRE download and install that has to happen on every computer before that Java app/applet can run on a computer.
The funny thing is, it's not enough to make a set of products that would make you filthy rich *if
>Could I get a $500 used Mac with a CRT monitor? Sure, but who wants that big bulky thing around? >Instead, give me something I can use with a USB KVM switch, and then I can explore it on my own pace.
Plug the KB, mouse, and monitor into the Mac and use Windows via Remote Desktop client or X11. I did that for over a year adminning a Win2K Server dev box and it worked quite well. There's very little that you *have* to do at the console of a Windows box, and if you have the 2 machines on 100 meg ethernet together it's really responsive.
Or, just buy an old iBook G4 or Titanium PB G4 and use the built-in kb and display for a while.:)
>by letting go, they created the market we have today, even though they didn't benefit from it. TFA says IBM lost market dominance as a result.
No, it says that "in belatedly attempting to recover it, fatally broke with established industry standards", meaning PS/2 and MCA. IBM thought it could close the market again but obviously was wrong about that.
Who knows where IBM would be today if they had just stuck with the standard PC business instead of trying to trap the market with the PS/2?
And, of IBM and Apple, who's making money selling computers, and who's getting out of the business?
Microsoft didn't trash Netscape because of a presence or absence of standards. HTML and HTTP were already there, and Netscape did create some things like HTML extensions and SSL and cookies and JavaScript that were theirs, but they were open about it.
But, it was on the server side (Microsoft IIS vs. Netscape Enterprise Server) that Microsoft killed Netscape, and it was due to dumping funded by monoply profits that Netscape didn't have. Perhaps Netscape could have implemented NHTTP and locked users into that, but I doubt that would have succeeded.
If Microsoft wants to kill Google they should just give away MSN Search ad placements for free for the next five years, and allow customers to license MSN Search for their intranets and portals for free for the next five years. That (as far as I know) is where Google gets its money from, and that would likely kill them (as well as probably landing Microsoft in court again).
They aren't preparing it. They're more than two orders of magnitude, and well over five years away. This is another typical/. "someday we might make a product you want" article. ("Animated holographic XXX playing cards running Linux on Crusoe? Save your allowance kiddies!!!")
From TFA: "researchers say they expect to one day build cartridges that can store as much as 100T bytes of data."
One day?
"in order to store more than the 1T byte of data that IBM is planning for its next-generation products"
"he said cartridges that can store a terabyte of data will hit the market within 18 months."
So, the next generation (only 18 months away) will feature 2.5x more storage. Only 100x more to go. We're practically there!!
"His group of ten researchers hopes to shrink that size down to about 0.5 micron, or 500 nanometers, within the next five years."
Hmm, 2.5x in 18 months, and then a little over twice that long after that they'll have the right core technology in place to get to 100TB.
Steve Jobs hoped IBM would hit 3GHz in one year. Look how well IBM performed on that one.
'"This will carry us all the way to the 100T byte regime," he said.'
So, in five years they hope to have the core technology in place that they'll then have to refine in order to get to 100TB per tape. How long will that take?
"Narayan was reluctant to predict when IBM might bring its first 100T byte tape devices to market"
Slashdot Mirror
>If Oracle or whoever and code their databases for SMP, then why can't OS vendors or game manufacturers?
Databases can only do certain operations in parallel, and some of those require additional optimizations like separate drives to operate at maxmimum efficiency. An OLTP database with 10000 users is an inherently parallel problem domain.
OS vendors do code for SMP. Not all of them code for large numbers of processors, though, or they may not have succeeded at the high end. Sometimes there are performance trade-offs between uniprocessor and SMP performance, or between wall-clock execution time and machine throughput. If you're designing a big honkin' database server, you might have to make a 20ms query take 40ms due to DB and OS locking overhead, but if you get 3x the overall system throughput, you won. That trade-off might not make sense on a uniprocessor commodity desktop, so you find that apps and OSs intended for desktop hardware don't really know what to do with an 8-way machine.
Anyway, I'm not trying to say that it's futile, I'm just trying to point out that it's non-trivial, and will take some time. It's pretty clear that scaling sideways is going to be necessary more and more (more threads instead of just jacking up the clock speed), so this is a challenge that software developers are just going to have to learn to deal with.
>Im a software developer and REALLY hate the movement towards dual-cores.
Tough. Chip makers are up against a technology barrier right now, and clock speed increases in the CPU don't make RAM or disk or interconnect faster anyway. How about just putting a 4MB cache on-die? That wouldn't require a massive clock speed increase but it would speed things up. I'm not an EE but I'm just pointing out that there are many, many things that have been left in the dust by Moore's law that could catch up and make quite a difference. Does your computer have 4+GB of DDR memory? ATA-133 drives with 8MB cache? PCI-X? A 64-bit CPU and an OS that knows how to use it fully? In what other ways are CPUs waiting on everything else, that could be improved to make things run faster overall?
Learn to parallelize your code where possible. Optimize your existing code. Software optimizations yield stunning improvements compared to incremental clock speed bumps anyway, and (unlike hardware) affect every installation of your app.
>Its a bad move IMO on AMDs and Intels part
OK genius, what's the alternative? No improvements in processors for years, until somebody makes a breakthrough that enables 4+ GHz processors? What happens when they hit the next roadblock?
Hardware has been so far ahead of software for so long that we've become accustomed to solving bloat with "just buy a new computer". It wouldn't kill us to spend a little time profiling code. The economics have been (in many cases) such that it just made more sense to throw money at new hardware. If that no longer makes sense, throw money at software optimizations for a little while. It doesn't exclusively mean that we have to force every algorithm to operate in parallel. It could be as simple as releasing fat binaries of apps that are compiled to target recent CPUs (no more shipping 386-optimized code to every customer), or *gasp* writing more efficient code in the first place.
There's only one problem with your list: all of these things are COMPLETELY unreasonable to ask every single user to do. This is what vendors are for.
A better approach comes from the vendor:
1) Install nonessential software with a disabled configuration. If they (or an app they use) need it, they can enable it then.
2) Enable automatic software updates. If you have a vendor trust issue with this, turn it off on your computer and those that you have influence over. If the computer isn't connected to the internet, nag the user to get an update (vendor-provided near-free update CD + a reasonable S&H fee, or via LAN, etc.) once every few months.
3) Add a host firewall, and make the OS work OK with the host firewall enabled. Provide simple high-level checkboxes for people who want to allow incoming connections to high-level services, and warn them if they do.
4) Add scripts (which get updated with the auto software update) that know what traffic is noise and what traffic looks like a successful attack or exploit. Send interesting packets upstream to the vendor for analysis. If you have a problem with this, turn it off for computers you control. A computer with always-on internet access is constantly sprayed with random attack packets; there's no point in wasting the user's time. Hmm, a connection attempt to port 25 on a laptop that has no SMTP server running. There's no point in wasting the user's time with this information. At most, increment a counter and leave it at that.
Actually, Windows XP SP2 has made some of these improvements. The problem is that the underlying services are also insecure, and the services are so entangled with one another that you can't really function on a LAN without opening yourself up.
> So...you're basically saying thin client is a great idea that has been hampered by poor implementations.
That's not at all what the parent post said.
It's an idea that sounds good until you think about the fact that you're intentionally buying something that's less functional. Then, people who aren't looking for an appliance (cash register, kiosk, etc.) start thinking about all of the things that a personal computer does well, and they buy it. See also: the "Mac OS X and Linux don't have any applications so I must buy Windows" rationale.
I've said it before 'n I'll say it again: the people who are pitching thin clients and network computers are the people who make their money off of large-scale server hardware and software. Just as Intel encouraged modem manufacturers to push processing loads onto the CPU (winmodems) to sell CPUs, NCs and thin clients are largely an invention of Sun and Oracle (and their ilk) who want to encroach on Wintel's revenues.
Oh yes, life will be so very much cheaper when you get rid of all of those silly little x86 servers and desktop PCs, and buy a bunch of proprietary, non-commodity-priced clients that will use gobs of network bandwidth to run apps on proprietary, non-commodity-priced servers that we just happen to sell, using a proprietary, non-commodity-priced app virtualization scheme (and perhaps even a proprietary remote GUI technology!).
Google might try a more modern, open approach to this: some kind of DHTML app suite, with a super cheap subscription, not based on proprietary hardware or (client-side) software. That's the only speculation I've heard about this that is in the least bit plausible, but there are still issues.
1) Yahoo has been in a position to do this for years. Why haven't they? They have a bunch of light-HTML apps, and a bunch of money. They could build heavy-HTML apps too. I don't believe that the idea just hasn't occurred to them yet.
2) Who the hell would use it? People like PCs largely because they're personal. Nobody wants to push their precious secret nekkid vacation pictures and financial data out onto some random server out there in virus land. Try telling folks that not only do they have to get a PC, but also install a new web browser, and that buys them the privilege of paying even more to run shitty apps that aren't Microsoft Office and that are on the far end of a slow, laggy pipe. WHERE DO I SIGN??!!?
Scott McNealy's strategy of "tons of people will buy this, because by doing so they will make us rich" just doesn't work.
BTW, I do think that thin clients make great sense for specific, limited applications. They just aren't the PC-killer for typical office and home users that NC and thin client pitchmen wish they were.
Given the law of averages, the fact that there is no such thing as a "law of averages" means that it's likely that eventually one will spontaneously coalesce into existence.
At this time, Marilyn vos Savant will marry Monty Hall and they will rule the universe.
Remember that G3, G4, and G5 are just Apple marketing terms. They can take a G4 with a bigger cache and call it a G5 if they want, or take a G5 and yank out a bunch of the superscalar bits to make it use less power.
You are high.
See also: BeOS.
Would it be a better computer if they removed iMovie and iDVD from iLife on systems that didn't include a DVD writer from the factory?
Well, that sucks. Still, outlying data points such as yours are just that.
Did you try to use the Keychain instead?
What did you eventually do? Roll back and give up for now, or did you find a solution?
256MB of the correct memory costs $48 with free shipping from Monarch Computer. So, we're talking about a difference in price ($27) that's about as big as the sales tax on the Mac mini.
(Because it's Slashdot, some dickhead is going to loudly point out that the Mac mini appears to have a single DIMM socket, so it'd really be a 512MB DIMM replacing a 256MB DIMM, so the price would theoretically be slightly less if Apple offered a 0MB option. Now I've said it so you can apply your brilliant insight elsewhere.)
If you want a big computer that you can open up and service yourself, get one. If you want a teeny tiny computer that has everything crammed in as tightly as it can be, while still using industry standard parts, don't be shocked when you find out that some skill is required to install the stuff.
Or, you can just wait for a web site to publish the directions, and do it yourself, and take your chances with breaking something. If you do it properly, they'll never know. If not, well, maybe you'll wish you hadn't been so cheap.
>but 256MB ram is nothing
Too bad you can't upgrade it to 1GB.
Oh wait.
If all you want to do is look at your hardware and run benchmarks (a.k.a. 3D FPS games that bore anyone over 12 years old), that's great.
Watch the keynote and look at the software. Apple is not trying to make a Windows PC that looks nice. They're making software too, and leading the pack in many categories.
Most people actually want to accomplish tasks with their computers, and a few percentage points in hardware price/performance don't make nearly as much difference as better usability (which is measured in units of time to accomplish a specific task).
But, if you just like to spend your whole day messing around with your computer instead of getting anything done, by all means, avoid Macs.
Or, if you believe what Steve Jobs said in the keynote, they found that people use shuffle anyway, so they just focused on that, and took away the feature of random access. You can turn shuffle off and just use single-song fast-forward and rewind, but with 120 songs that might get pretty annoying.
:)
I had a Rio 500 128MB flash player before I bought an iPod, and with only about 2 albums' worth of space on it, just FF and Rew controls were fine. There was a wheel, and a whole complicated UI, with EQ and bookmarks and such, but really, with about 20 or 30 songs (2 albums) tops, there isn't a need. It's better to just leave it in your pocket and hit the FF button a lot.
I think I'd miss the EQ, and selecting albums. But that's part of why I won't sell my 20GB iPod to buy an iPod shuffle.
I suspect that most people, who are song oriented and don't play albums in their entirety like I do, will embrace this and love the whole "what am I going to be hearing today" idea. After all, if you don't exhaust the 120 songs before your next iTunes reloading session, it's effectively shuffling from your entire collection since you won't hear a song repeated.
Plus that $99 thing is handy. Amazon and Best Buy and similar retailers will almost certainly undercut that by a few bucks.
>(TM)
Show-off.
Argh. It's the old "why make decisions when you can just add yet another preference pane" cop-out. See also: Mozilla.
>also eliminating their bandwidth hogs.
Hellooooo.... ISPs sell bandwidth. DSL costs more than dialup. A T3 costs more than DSL. Don't take my word for it - check with your ISP. It's actually true.
You might as well say that a gas station owner would do well to ban trucks and SUVs becuase they're fuel hogs.
The fourth amendment doesn't grant individuals the right to commit crimes anonymously.
Yeah, the short filenames thing is really annoying. We've had long filenames on OS X for how many years now? How hard is it to just change the code to allow me to save a file with a long filename from Office? Apparently, it's much too hard.
Word X for Mac with a blank new document needs 38MB RAM. Not "200MB".
/., the correct localized spelling is "rediculous". My bad.)
It does do that annoying "use some CPU for nothing in particular" thing, though. That's pretty ridiculous. (Oh wait we're on
>If the primary objective was to maintain an OS monoploy, and if it believed Netscape was going to
>release a competitive OS (a doubtful proposition) the most direct course of action would have been to buy Netscape.
No. Netscape was worth a lot back then. That might have been "direct", but it wasn't the most profitable approach. That would have been a very foolish and expensive solution.
Instead, Microsoft bought (or licensed rights to? I don't remember) a much less successful browser and built it into a Netscape competitor on the client side, but how they really preserved their monopoly was by killing Netscape's actual revenue stream by bundling IIS for free with NT 4. There was a significant market for commercial web servers back then; Netscape Enterprise server cost about a thousand dollars. SuiteSpot cost several thousand dollars. Lots of companies actually bought it and used it for a while; in many ways it was far more advanced than Apache (which was pretty new at the time), blew away CERN and NCSA httpd, and worked quite well running on NT 4.0.
Microsoft knew that Netscape was around the corner from making the OS irrelevant via HTML, JavaScript, and Java. Netscape was explicitly trying to do this, calling it "crossware", meaning cross platform software that didn't depend on a specific OS, but was built on internet and web standard technologies (standards Netscape invented or championed). Microsoft saw the threat and headed it off. This has been well documented in the DOJ and Sun cases and Microsoft was found guilty in both.
Microsoft absolutely does not have the ability to drop Windows quickly. To suggest that they do is totally absurd.
>then they started calling me and I quickly realized very few people should have a computer.
Maybe you should change your number, or start charging for your time.
>Palm has a really smart way to deal with that by
>limiting any interaction with the OS and making the App king.
DOS worked that way too. People actually want multitasking, it turns out.
By the way, you still have to interact with a host OS in order to use a PDA (installing new apps, backing up data, etc.) except in the most limited case where someone is totally happy with the default apps and doesn't want to back up their data, ever. Every app I've ever seen for the Palm requires you to download it on a PC (or Mac or Linux box or...), unzip it, and install it via Palm Desktop. So, they just didn't bother duplicating that on the handheld, which is smart, but doesn't mean that the user never has to touch an OS to use a Palm.
>Plus having everything running all the time makes everything faster.
Actually, it makes things slower, unless you have a huge amount of physical memory, which Palm PDAs don't. Besides, since when does PalmOS run all installed applications simultaneously? I've written a couple of PalmOS apps, and they actually have to launch, and you can't run more than one at a time.
>The smart money is on going BACK to mainframe type applications and computing.
Not really.
Sun Microsystems, clueless and sinking fast
Oracle, not looking to healthy
And Marimba (an early Java-apps-as-a-service company) was coughing up blood and just got bought out.
The quality of their individual products doesn't matter if their strategy sucks. (I have a lot of respect for Sun gear, Solaris, and Oracle, but that doesn't mean they're doing well in the market, or that they have a plan for how they're going to make stuff that people want for a price they're willing to pay.)
Apparently the smart money is in short-selling companies that preach the Network Computer gospel. Any year now, corporations will buy $800 boxes that do less than a PC, in order to save money. Oh don't mind that hugely expensive server and software stack that you have to run on the server, don't mind the training and limited apps. We promise, it's cheaper to have everybody sitting on their hands if the network is down, and laptops are just a fad anyway.
While bandwidth is getting cheaper, so are PCs, and NCs have essentially zero volume, so they aren't getting any cheaper at all. Meanwhile, user expectations are keeping up with PC capabilities. Maybe the NC hype will come true in the form of PCs configured as thin clients running apps on Windows servers, or Linux servers (dare we dream) but not in large numbers for a long time. Hard disk price/performance from a speed and storage POV still blows away bandwidth, and until that changes in many orders of magnitude in the favor of network being cheaper and faster than a local disk, the NC idea will remain a fantasy of the folks trying to sell you the back-end stuff at premium prices.
You can wave your hands about bandwidth being essentially free, but you'll have to back that up by installing a T3 to your house and telling me that you don't notice the monthly bill. Hard disk storage is moving towards "essentially free" (that is, it's getting cheaper) quite quickly too. What matters is not the separate rate of change of each, but the ratio. Which one is getting cheaper faster? How long will it take for them to cross over and for one to become drastically cheaper than the other, or are they diverging?
>Java (etc) have been invented so what's the wait?
The wait is for that 25MB JRE download and install that has to happen on every computer before that Java app/applet can run on a computer.
The funny thing is, it's not enough to make a set of products that would make you filthy rich *if
And then the only-slightly-defective unit is taped shut and/or re-shrinkwrapped and sold as new, again and again, until somebody finally keeps it.
Oh wait, maybe that's only at Fry's...
>Could I get a $500 used Mac with a CRT monitor? Sure, but who wants that big bulky thing around?
:)
>Instead, give me something I can use with a USB KVM switch, and then I can explore it on my own pace.
Plug the KB, mouse, and monitor into the Mac and use Windows via Remote Desktop client or X11. I did that for over a year adminning a Win2K Server dev box and it worked quite well. There's very little that you *have* to do at the console of a Windows box, and if you have the 2 machines on 100 meg ethernet together it's really responsive.
Or, just buy an old iBook G4 or Titanium PB G4 and use the built-in kb and display for a while.
>by letting go, they created the market we have today, even though they didn't benefit from it. TFA says IBM lost market dominance as a result.
No, it says that "in belatedly attempting to recover it, fatally broke with established industry standards", meaning PS/2 and MCA. IBM thought it could close the market again but obviously was wrong about that.
Who knows where IBM would be today if they had just stuck with the standard PC business instead of trying to trap the market with the PS/2?
And, of IBM and Apple, who's making money selling computers, and who's getting out of the business?
Microsoft didn't trash Netscape because of a presence or absence of standards. HTML and HTTP were already there, and Netscape did create some things like HTML extensions and SSL and cookies and JavaScript that were theirs, but they were open about it.
But, it was on the server side (Microsoft IIS vs. Netscape Enterprise Server) that Microsoft killed Netscape, and it was due to dumping funded by monoply profits that Netscape didn't have. Perhaps Netscape could have implemented NHTTP and locked users into that, but I doubt that would have succeeded.
If Microsoft wants to kill Google they should just give away MSN Search ad placements for free for the next five years, and allow customers to license MSN Search for their intranets and portals for free for the next five years. That (as far as I know) is where Google gets its money from, and that would likely kill them (as well as probably landing Microsoft in court again).
They aren't preparing it. They're more than two orders of magnitude, and well over five years away. This is another typical /. "someday we might make a product you want" article. ("Animated holographic XXX playing cards running Linux on Crusoe? Save your allowance kiddies!!!")
From TFA: "researchers say they expect to one day build cartridges that can store as much as 100T bytes of data."
One day?
"in order to store more than the 1T byte of data that IBM is planning for its next-generation products"
"he said cartridges that can store a terabyte of data will hit the market within 18 months."
So, the next generation (only 18 months away) will feature 2.5x more storage. Only 100x more to go. We're practically there!!
"His group of ten researchers hopes to shrink that size down to about 0.5 micron, or 500 nanometers, within the next five years."
Hmm, 2.5x in 18 months, and then a little over twice that long after that they'll have the right core technology in place to get to 100TB.
Steve Jobs hoped IBM would hit 3GHz in one year. Look how well IBM performed on that one.
'"This will carry us all the way to the 100T byte regime," he said.'
So, in five years they hope to have the core technology in place that they'll then have to refine in order to get to 100TB per tape. How long will that take?
"Narayan was reluctant to predict when IBM might bring its first 100T byte tape devices to market"
So, even longer than five years.