Most of Leopard's problems are traced back to bad 3rd party software that uses undocumented hooks.
Every Ubuntu user I know (~6 people) has had issues with the Gutsy upgrade; more than half of them "resolved" the issue by wiping the machine. Given that Ubuntu's development process is far more "open" and there was no "third party" software involved (none were using third party binary drivers), what's the excuse?
I've seen CUPS break so badly that it constantly "stops" all the printers. Monitor resolutions and scan rates that were completely wrong and required hand-editing Xorg's config file, when the old config had worked just fine. One machine had an ethernet port completely disappear- and it was the one the ethernet cable was plugged into! Most were machines in use by programmer types, who didn't go mucking about save what was available via the GUI, because they don't know linux well enough. I can't blame the user in these cases.
Even with the previous release, when I upgraded a very simple server, there were problems with device-mapper pegging the machine until I spent half an hour screwing around with it, and finally found a post and bug in the ubuntu bugtracker. Of course, the bug had been known for months, and do you think anyone bothered to release a fix? Nope!
I'm sure someone could come up with a feature list that would make an XServe look competitive to a straw man "equivalent" box. Let's see, dual dual-core woodcrest plus *3* drive bays? Most 1U servers only have 2 bays, so that'll narrow things down
Not really, considering there are many 4-bay choices out there; ALL of the 1U NAS boxes are 4-bay. The Xserve used to have 4 bays, and they cut it back to put in cooling ducts. The third drive is largely useless considering that all most people want to do in that market segment is have a pair of mirrored drives.
Have you actually spec'd out rackmount gear in an enterprise setting? Everyone's got "dual core woodcrests". In fact, you have far more processor options from other vendors; HP offers almost a dozen processor choices for 1U machines, and that's just on the intel side; they offer another slew of choices from AMD. Apple only offers intel, dual-core xeon processors, while everyone else offers everything from celeron/Pentium-D to quad-core, dual socket solutions. I'm sure that's coming- in 6 months or so, Apple will have a big press event where they offer Xserves with a pair of quad-core processors. "Whoop de doo, welcome to January 2007", everyone else will say.
Dual power supplies, dual network, two [PCI/PCI-e/PCI-x] slots, dual processor slot, etc- it's all available from IBM, HP, Dell, etc. All the other vendors also offer superior iLOM features Apple doesn't, like virtual floppy/CDROM images and full virtual KVM emulation; the best Apple offers is a serial port and software-based remote control.
Most of them also support hardware raid- again, something the Xserve doesn't- unless you pony up another thousand bucks. Everywhere else, $1k will buy you a raid controller with gobs of on-card memory, battery-backup capabilities, and a lot more than three ports.
This probably presages a thoroughly overhauled Xserve product with greater capability for acting as a virtual machine server, too.
Huh? The current Xserve supports 3 SATA drives, 32GB of memory in 8 slots, and redundant power. Oh yeah, and 4 processor cores. Far as I know, all recent Xeon processors support intel virtualization features.
Regardless- I don't think you'll see Apple kowtowing to the virtualization fetish. Beyond the usual desktop virtualization needs, I don't think Apple's target audience for the Xserve needs this capability.
Let's all take a step back and realize that the current base Xserve is THREE GRAND and pretty damn bare-bones for that price-point; that does include OS X server unlimited, but yeeeeeesh- that's still almost $2k. I'm the first to argue that Apple's hardware isn't as overpriced as everyone claims, but this is one notable exception. It doesn't even include basic hardware RAID capabilities- you have to buy a (inserts pinky into mouth) ONE THOUSAND DOLLAR proprietary raid card to do hardware raid! Jeeeeesus christ, even the cheapest 1U boxen support BASIC raid, typically, or it's a $100-200 option...
API. While privacy advocates have been concerned about Google for the past several years, most of us are just beginning to comprehend Facebook's growing impact on who, when, what and how we connect with friends
Especially since we just learned that Facebook considers it a "perk" to allow their employees to surf people's profiles, read their email (which they're pushing HARD to get people to use as a sort of bastardized webmail) and see their "private" photos and such.
Put it this way: my Hackintosh in it's original incarnation had a 2.6ghz Celeron, 1GB of RAM, 160GB of Hard Drive space, a DVD Burner, and a Geforce 7300LE. Now, this was kind of a toss up between a bare-bones Mac Mini at the time. The mini had it in processor speed, but the $599 machine had less ram, less hard drive space (and a slower hard drive), and a slower video card. That and it wasn't really upgradeable.
And a BMW M5 probably costs more than a 20 passenger minibus. What's your point?
The mini is a TINY system. That's why it costs more than a standard, large Dell or HP. Go pick any major manufacturer, and spec out their smallest "SFF" PC. Now put it next to the mini, and laugh at how much smaller and quieter it is. And no 802.11n or bluetooth in that price tag, generally. The mini can be had/comes with both inside (no dongles necessary.)
Now go online and try and build a mini-itx box similarly configured. Not such a drastic price difference anymore, eh?
One big reason your system is a better value is because your "Hackentosh" is running an operating system you did not buy a license for.
Pete Worden, a Lunar Lander Challenge judge - and director of NASA's Ames Research Center, told SPACE.com that the engine blew up, with the rocket's engine chamber tossing out pieces onto the pad.
"It's over for them for this X Prize Cup," Worden said. But he added: "I do think they are getting there...it's a robust design.
Uh... I don't think anything was "lifted". In zero G, there is no up and down, AFAIK.
Since you're being a nitpick: they're not in "zero g", they're in orbit. There is a difference. One means there are no (or, in practical terms, very little) gravitational forces acting on you; the other means you're hurtling through space fast enough that you counteract gravitational forces trying to pull you down to the planet.
Minister of Education Fioroni brought in to trial two of Google Italy's managers and then proposed and strongly sustained his idea of censoring the Internet to protect the young. Now Ricardo Levi, the prime minister's right hand, is finally successful in promulgating his law on internet censorship
We're talking about a country that changes governments like people change underwear. Relax and wait a week.
If you build it, NASA will not only come, it'll give you $2 million dollars for you troubles. The space agency today said it will offer $2 million in prizes if competing teams can successfully build a lunar lander at the Northrop Grumman Lunar Lander Challenge at Holloman Air Force Base, in Alamogordo, N.M. Oct. 27 and 28th.
And will they get extra credit for video/photo/3D telemetry? How about spectrography gear and other testing equipment? Because they've got all that. On the way to the moon or already there.
I'm so tired of my tax dollars being wasted on international dick-waving contests like this. I wish NASA et al would just whip out the rulers- it'd be cheaper. Then again, it wouldn't feed the defense contractors, now would it?
Multi-layer cells and mechanical trackers and even mirrors mitigate these problems in existing system
Having recently spoken with someone who worked for a solar energy contractor that did large-scale commercial installations, tracking systems double the cost of the system.
Before you scream "zOMG I googled it and they only cost $x dollars", remember that any time you transition from a static system to a dynamic one, complexity and cost go up. It's not just the cost of the tracker device; it's the cost of additional installation work, maybe they need to be calibrated/adjusted individually, maybe they need repair or maintenance, maybe they require a different power hookup method which costs more or takes longer to setup, etc.
I don't suppose you would have noticed that they're shipping from Israel? In which case, $40.00 isn't too bad.
Who said they're shipping from Israel? They probably have a US distributor; they'd be stupid not to, as otherwise it means the customer pays duties and deals with customs headaches.
Even if they are shipping from Israel, it still means the thing costs $320+. Besides, even in Israel, they're charging $20. That's highway robbery.
Do yourselves a favor and get a VIA-based mini-itx board for that kind of money.
Seems you can get a VB7001G (1.5Ghz) for about $130; add in $30 for 512MB of ram (2x the fitPC), and however much you feel like spending on a compactflash card, USB memory key, or smaller laptop drive. Say, $50 for a 60GB drive (more than the fitPC's 40). $40 for a picoPSU; $30 for a AC adapter. Buy a crap case for $30 if you don't have one you can use already. Install a gigabit NIC for under $20 (dunno if there are any cheap dual-interface gigabit NICs.) That's under $310, and quite a bit more bang for the buck. It probably won't be 5w, but it'll be well under 20w given that board seems to use about 10w.
If you want to go even cheaper, intel is fighting back against via, like with the D201GLY. It's $70, 1.3ghz celeron, DDR2 ram...
It makes one wonder how many of these companies eschewed open-source solutions, in favor of expensive "supported" software.
Hopefully enough of these examples will eventually reach the tipping point where PHBs will finally begin to wonder what exactly they're getting for their money.
You don't understand why companies prefer commercial solutions.
If I buy services from a company and they fail to deliver, I have choices. Like suing them (example: breach of contract) and recovering damages. I can't do that if I install open-source software, unless I hire a firm to take care of the implementation, and *they* fuck up on what they promised they'd deliver.
Just because you don't understand how something works, doesn't mean it's broken.
I think you need to step back, and remember that IT works for the company. The company does not work for the IT department.
You need to stop lecturing people with more IT experience than you, kid. We don't just sit around waiting for your beck and call; we've got shit to do, too. Our time costs the company just like your time does. Furthermore, time spent fixing your laptop that could have been avoided with giving you a desktop, means we couldn't spend time on other things that could have helped the company make money. We Big Boys refer to that sort of thing as 'opportunity cost'.
Second: everything about laptops are more expensive for employers. They're far more easily damaged (and the resulting lost employee productivity waiting for it to be fixed, time devoted to it by IT staff, etc.), less upgradeable (which means it isn't as useful as a capital expenditure), much more easily stolen (and the resulting loss of trade secrets, lost employee productivity, etc.)
A good chunk of the people who claim they need a laptop to 'do work at home' don't do a drop of work at home...it's just so they get a nice 'home computer' out of their employer.
It's true that you can't expand a RAID-Z set (I think, anyway -- if you replace all of the drives, one at a time, does that work?), but you can add another RAID-Z set, and expand the pool.
That's the big thing in ZFS, combining all of the resources into a pool, rather than treating disks (or groups of disks) as part of a volume. The other part of this was making filesystems nearly as light-weight as directories.
I understand all this, and don't care.
If I have a 10x500GB drive RAID-Z pool in a twelve drive chassis and I need to add roughly another 1TB of storage, I'm fucked unless I go and buy 1TB drives, and I just wasted ~$400 (or whatever they cost these days) because I had to add them as a mirror to the pool instead of restriping the array.
Same problem if I want to upgrade those drives to 750's. Not a problem with 100% of the raid arrays and raid controllers on the market. Impossible with ZFS and RAIDZ.
If RAIDZ supported re-striping and odd disk sizes (ie, yank a 500, insert a 750 and bam, get the extra space), it would be hands down perfect.
Notebooks break, they get lost, and they are replaced frequently, so the cheaper, the better.
No. The more reason to drop the laptop fetish. Laptops absolutely have their appropriate uses- but desktops work just fine for a huge percentage of people. Their components are cheaper, more easily replaced, and usually superior in performance. Nevermind that forcing you to sit in front of the computer, as opposed to being available to you in bed, on your couch, on your porch, etc- means you're more prone to wasting more time on the internet.
Yet...very few people I know will even consider a desktop. It drives me insane in business settings- I can do all manner of repairs and data recovery very, very easily on a desktop. Laptops are a total mixed bag ranging from "the company will have a tech here tomorrow morning" to "ARRRRG its going to take an hour to get the damn thing apart."
Thinkpads and Dells are the best, from my experience; HP sells a lot of consumer-ish crap. Apple gets a failing grade in almost every regard; iBooks, Powerbooks, and Macbook Pros are MISERABLE to disassemble for hard drive replacement. iBooks require damn near COMPLETE disassembly to get to the drive. The only plus is firewire target disk mode, but that is near useless in case of hardware failure.
Resource forks are far better than the idiotic "everything is a folder" model.
Want to upload that Keynote project to your friendly CMS via a web browser? Can't, because it's not a file, it's a #@$!ing FOLDER. You have to zip it first. Words cannot accurately describe how tiresome this becomes.
It also makes data recovery (should the file get accidentally deleted) nearly impossible- the files inside the folder are not named uniquely or in any identifiable manner.
ZFS isn't nearly all it is cracked up to be- among other things, you can't expand RAID-Z...absolutely moronic. I'm not even sure you can expand a simple mirrored pool. Users have been repeatedly asking for growing abilities, and the developer reaction was "just create a larger pool and move it over". That's hilariously stupid advice given that you usually don't have that kind of storage hanging around- not even in enterprise environments.
There's simply no comprehension amongst the ZFS developers that virtually EVERY raid card on the market supports such an operation. Even more shocking was when one developer said (paraphrasing) "gosh, how would one even go about doing that sort of thing?"
Don't get me wrong- checksumming and automatic disk scrubbing are features long overdue, but ZFS is not magic bullet.
Ff you were at the same gathering as I was (and it sounds like you were) the laptop thing was a bit of a glitch owing to the fact that that presentation was setup with burned DVD's that played find on a DVD player but didn't navigate correctly on the laptop they used.
They were movie files of some sort on the laptop. It wasn't the same "gathering", but it is good proof that the man apparently still can't get his shit together.
I don't think that skeptics worship Randi any more than Open-source enthusiasts "worship" Linus Torvalds.
The two other speakers at the presentation were so busy verbally sucking Randi's dick while whipping out rulers, they couldn't notice the audience was bored to tears. We clearly were not at the same presentation. This was outright "this is how great Randi is, this is how great I am for being this close to Randi." I really wish they had just let the man speak for himself.
I've personally attended a "skeptics" meeting where he was giving a talk. One thing that struck me was how much hero-worshiping was going on. Some guy spent a good 20-25 minutes telling us about how well he knew James Randi, how close he was to James Randi, how he could pick up the phone and call James Randi, yadda yadda.
This was after waiting about 30 minutes for them to start- they had to get a laptop working with various videos of James Randi on (mostly asian) TV. Each video, of course, did not play properly, or they played the wrong ones (ie, the same thing over and over.)
His work is worthwhile and he's decent showman, but he's also grossly over-sold and over-hyped. The devotion (if not downright worshiping) is hilarious, given that it is being done by a bunch of people who call themselves "skeptics."
The problem is between the computer and Microsoft via the tubes.
Bullshit; it's not just Microsoft. A long-standing complaint of mine about Apple Mail is that it does not show the true URL in an HTML email via tooltips or any other method. The only way the user has to see the URL is to copy it, then manually paste it into the address bar in Safari or Firefox. So, "Click here to login to your account" is impossible to verify without extra work.
It'd take one engineer about 10 minutes to code in such a display, and they can't be bothered. But OH BOY, in Leopard, I'll get fancy pre-formatted emails to use for sending vacation photos!
Similarly, for all the fuss about how secure and better Ubuntu is, you have to recompile Netatalk with custom options (and the instructions provided don't work) to enable secure login. Why? Because of OpenSSL/GPL licensing issues that have existed for several years. Has anyone bothered to rewrite the hundred or so lines of Netatalk code to use GnuTLS instead? Nope!
PS: For those of you about to tell me "use samba", Netatalk handily outperforms samba and supports full MacOS filenames.
Slashdot Mirror
Most of Leopard's problems are traced back to bad 3rd party software that uses undocumented hooks.
Every Ubuntu user I know (~6 people) has had issues with the Gutsy upgrade; more than half of them "resolved" the issue by wiping the machine. Given that Ubuntu's development process is far more "open" and there was no "third party" software involved (none were using third party binary drivers), what's the excuse?
I've seen CUPS break so badly that it constantly "stops" all the printers. Monitor resolutions and scan rates that were completely wrong and required hand-editing Xorg's config file, when the old config had worked just fine. One machine had an ethernet port completely disappear- and it was the one the ethernet cable was plugged into! Most were machines in use by programmer types, who didn't go mucking about save what was available via the GUI, because they don't know linux well enough. I can't blame the user in these cases.
Even with the previous release, when I upgraded a very simple server, there were problems with device-mapper pegging the machine until I spent half an hour screwing around with it, and finally found a post and bug in the ubuntu bugtracker. Of course, the bug had been known for months, and do you think anyone bothered to release a fix? Nope!
I'm sure someone could come up with a feature list that would make an XServe look competitive to a straw man "equivalent" box. Let's see, dual dual-core woodcrest plus *3* drive bays? Most 1U servers only have 2 bays, so that'll narrow things down
Not really, considering there are many 4-bay choices out there; ALL of the 1U NAS boxes are 4-bay. The Xserve used to have 4 bays, and they cut it back to put in cooling ducts. The third drive is largely useless considering that all most people want to do in that market segment is have a pair of mirrored drives.
Have you actually spec'd out rackmount gear in an enterprise setting? Everyone's got "dual core woodcrests". In fact, you have far more processor options from other vendors; HP offers almost a dozen processor choices for 1U machines, and that's just on the intel side; they offer another slew of choices from AMD. Apple only offers intel, dual-core xeon processors, while everyone else offers everything from celeron/Pentium-D to quad-core, dual socket solutions. I'm sure that's coming- in 6 months or so, Apple will have a big press event where they offer Xserves with a pair of quad-core processors. "Whoop de doo, welcome to January 2007", everyone else will say.
Dual power supplies, dual network, two [PCI/PCI-e/PCI-x] slots, dual processor slot, etc- it's all available from IBM, HP, Dell, etc. All the other vendors also offer superior iLOM features Apple doesn't, like virtual floppy/CDROM images and full virtual KVM emulation; the best Apple offers is a serial port and software-based remote control.
Most of them also support hardware raid- again, something the Xserve doesn't- unless you pony up another thousand bucks. Everywhere else, $1k will buy you a raid controller with gobs of on-card memory, battery-backup capabilities, and a lot more than three ports.
This probably presages a thoroughly overhauled Xserve product with greater capability for acting as a virtual machine server, too.
Huh? The current Xserve supports 3 SATA drives, 32GB of memory in 8 slots, and redundant power. Oh yeah, and 4 processor cores. Far as I know, all recent Xeon processors support intel virtualization features.
Regardless- I don't think you'll see Apple kowtowing to the virtualization fetish. Beyond the usual desktop virtualization needs, I don't think Apple's target audience for the Xserve needs this capability.
Let's all take a step back and realize that the current base Xserve is THREE GRAND and pretty damn bare-bones for that price-point; that does include OS X server unlimited, but yeeeeeesh- that's still almost $2k. I'm the first to argue that Apple's hardware isn't as overpriced as everyone claims, but this is one notable exception. It doesn't even include basic hardware RAID capabilities- you have to buy a (inserts pinky into mouth) ONE THOUSAND DOLLAR proprietary raid card to do hardware raid! Jeeeeesus christ, even the cheapest 1U boxen support BASIC raid, typically, or it's a $100-200 option...
API. While privacy advocates have been concerned about Google for the past several years, most of us are just beginning to comprehend Facebook's growing impact on who, when, what and how we connect with friends
Especially since we just learned that Facebook considers it a "perk" to allow their employees to surf people's profiles, read their email (which they're pushing HARD to get people to use as a sort of bastardized webmail) and see their "private" photos and such.
Oh yeah, and get your password, log in to your account, and upload explicit photos.
Put it this way: my Hackintosh in it's original incarnation had a 2.6ghz Celeron, 1GB of RAM, 160GB of Hard Drive space, a DVD Burner, and a Geforce 7300LE. Now, this was kind of a toss up between a bare-bones Mac Mini at the time. The mini had it in processor speed, but the $599 machine had less ram, less hard drive space (and a slower hard drive), and a slower video card. That and it wasn't really upgradeable.
And a BMW M5 probably costs more than a 20 passenger minibus. What's your point?
The mini is a TINY system. That's why it costs more than a standard, large Dell or HP. Go pick any major manufacturer, and spec out their smallest "SFF" PC. Now put it next to the mini, and laugh at how much smaller and quieter it is. And no 802.11n or bluetooth in that price tag, generally. The mini can be had/comes with both inside (no dongles necessary.)
Now go online and try and build a mini-itx box similarly configured. Not such a drastic price difference anymore, eh?
One big reason your system is a better value is because your "Hackentosh" is running an operating system you did not buy a license for.
Joke------> * !
/ !
/ !
/ !
/ !
!
!
You------>X # <---- Team Armadillo
Pete Worden, a Lunar Lander Challenge judge - and director of NASA's Ames Research Center, told SPACE.com that the engine blew up, with the rocket's engine chamber tossing out pieces onto the pad. "It's over for them for this X Prize Cup," Worden said. But he added: "I do think they are getting there...it's a robust design.
That's one exciting definition of robust :-)
Zero G and free fall are equivalent from the point of view of the object,
It doesn't matter. They're still two different things...zero g means NO gravity.
Uh... I don't think anything was "lifted". In zero G, there is no up and down, AFAIK.
Since you're being a nitpick: they're not in "zero g", they're in orbit. There is a difference. One means there are no (or, in practical terms, very little) gravitational forces acting on you; the other means you're hurtling through space fast enough that you counteract gravitational forces trying to pull you down to the planet.
Congratulations on pointing out an obvious joke.
Minister of Education Fioroni brought in to trial two of Google Italy's managers and then proposed and strongly sustained his idea of censoring the Internet to protect the young. Now Ricardo Levi, the prime minister's right hand, is finally successful in promulgating his law on internet censorship
We're talking about a country that changes governments like people change underwear. Relax and wait a week.
If you build it, NASA will not only come, it'll give you $2 million dollars for you troubles. The space agency today said it will offer $2 million in prizes if competing teams can successfully build a lunar lander at the Northrop Grumman Lunar Lander Challenge at Holloman Air Force Base, in Alamogordo, N.M. Oct. 27 and 28th.
Will they let the chinese show up? Or maybe the Japanese?
And will they get extra credit for video/photo/3D telemetry? How about spectrography gear and other testing equipment? Because they've got all that. On the way to the moon or already there.
I'm so tired of my tax dollars being wasted on international dick-waving contests like this. I wish NASA et al would just whip out the rulers- it'd be cheaper. Then again, it wouldn't feed the defense contractors, now would it?
Multi-layer cells and mechanical trackers and even mirrors mitigate these problems in existing system
Having recently spoken with someone who worked for a solar energy contractor that did large-scale commercial installations, tracking systems double the cost of the system.
Before you scream "zOMG I googled it and they only cost $x dollars", remember that any time you transition from a static system to a dynamic one, complexity and cost go up. It's not just the cost of the tracker device; it's the cost of additional installation work, maybe they need to be calibrated/adjusted individually, maybe they need repair or maintenance, maybe they require a different power hookup method which costs more or takes longer to setup, etc.
Kinda matters when you're installing 500 panels.
I don't suppose you would have noticed that they're shipping from Israel? In which case, $40.00 isn't too bad.
Who said they're shipping from Israel? They probably have a US distributor; they'd be stupid not to, as otherwise it means the customer pays duties and deals with customs headaches.
Even if they are shipping from Israel, it still means the thing costs $320+. Besides, even in Israel, they're charging $20. That's highway robbery.
Another lovely company that tricks you with outrageous shipping costs to artificially drop the "price" of the computer. Also, check out the super friendly support and warranty policies.
Do yourselves a favor and get a VIA-based mini-itx board for that kind of money.
Seems you can get a VB7001G (1.5Ghz) for about $130; add in $30 for 512MB of ram (2x the fitPC), and however much you feel like spending on a compactflash card, USB memory key, or smaller laptop drive. Say, $50 for a 60GB drive (more than the fitPC's 40). $40 for a picoPSU; $30 for a AC adapter. Buy a crap case for $30 if you don't have one you can use already. Install a gigabit NIC for under $20 (dunno if there are any cheap dual-interface gigabit NICs.) That's under $310, and quite a bit more bang for the buck. It probably won't be 5w, but it'll be well under 20w given that board seems to use about 10w.
If you want to go even cheaper, intel is fighting back against via, like with the D201GLY. It's $70, 1.3ghz celeron, DDR2 ram...
It makes one wonder how many of these companies eschewed open-source solutions, in favor of expensive "supported" software. Hopefully enough of these examples will eventually reach the tipping point where PHBs will finally begin to wonder what exactly they're getting for their money.
You don't understand why companies prefer commercial solutions.
If I buy services from a company and they fail to deliver, I have choices. Like suing them (example: breach of contract) and recovering damages. I can't do that if I install open-source software, unless I hire a firm to take care of the implementation, and *they* fuck up on what they promised they'd deliver.
Just because you don't understand how something works, doesn't mean it's broken.
I think you need to step back, and remember that IT works for the company. The company does not work for the IT department.
You need to stop lecturing people with more IT experience than you, kid. We don't just sit around waiting for your beck and call; we've got shit to do, too. Our time costs the company just like your time does. Furthermore, time spent fixing your laptop that could have been avoided with giving you a desktop, means we couldn't spend time on other things that could have helped the company make money. We Big Boys refer to that sort of thing as 'opportunity cost'.
Second: everything about laptops are more expensive for employers. They're far more easily damaged (and the resulting lost employee productivity waiting for it to be fixed, time devoted to it by IT staff, etc.), less upgradeable (which means it isn't as useful as a capital expenditure), much more easily stolen (and the resulting loss of trade secrets, lost employee productivity, etc.)
A good chunk of the people who claim they need a laptop to 'do work at home' don't do a drop of work at home...it's just so they get a nice 'home computer' out of their employer.
It's true that you can't expand a RAID-Z set (I think, anyway -- if you replace all of the drives, one at a time, does that work?), but you can add another RAID-Z set, and expand the pool. That's the big thing in ZFS, combining all of the resources into a pool, rather than treating disks (or groups of disks) as part of a volume. The other part of this was making filesystems nearly as light-weight as directories.
I understand all this, and don't care.
If I have a 10x500GB drive RAID-Z pool in a twelve drive chassis and I need to add roughly another 1TB of storage, I'm fucked unless I go and buy 1TB drives, and I just wasted ~$400 (or whatever they cost these days) because I had to add them as a mirror to the pool instead of restriping the array.
Same problem if I want to upgrade those drives to 750's. Not a problem with 100% of the raid arrays and raid controllers on the market. Impossible with ZFS and RAIDZ.
If RAIDZ supported re-striping and odd disk sizes (ie, yank a 500, insert a 750 and bam, get the extra space), it would be hands down perfect.
Notebooks break, they get lost, and they are replaced frequently, so the cheaper, the better.
No. The more reason to drop the laptop fetish. Laptops absolutely have their appropriate uses- but desktops work just fine for a huge percentage of people. Their components are cheaper, more easily replaced, and usually superior in performance. Nevermind that forcing you to sit in front of the computer, as opposed to being available to you in bed, on your couch, on your porch, etc- means you're more prone to wasting more time on the internet.
Yet...very few people I know will even consider a desktop. It drives me insane in business settings- I can do all manner of repairs and data recovery very, very easily on a desktop. Laptops are a total mixed bag ranging from "the company will have a tech here tomorrow morning" to "ARRRRG its going to take an hour to get the damn thing apart."
Thinkpads and Dells are the best, from my experience; HP sells a lot of consumer-ish crap. Apple gets a failing grade in almost every regard; iBooks, Powerbooks, and Macbook Pros are MISERABLE to disassemble for hard drive replacement. iBooks require damn near COMPLETE disassembly to get to the drive. The only plus is firewire target disk mode, but that is near useless in case of hardware failure.
Want to upload that Keynote project to your friendly CMS via a web browser? Can't, because it's not a file, it's a #@$!ing FOLDER. You have to zip it first. Words cannot accurately describe how tiresome this becomes.
It also makes data recovery (should the file get accidentally deleted) nearly impossible- the files inside the folder are not named uniquely or in any identifiable manner.
ZFS isn't nearly all it is cracked up to be- among other things, you can't expand RAID-Z...absolutely moronic. I'm not even sure you can expand a simple mirrored pool. Users have been repeatedly asking for growing abilities, and the developer reaction was "just create a larger pool and move it over". That's hilariously stupid advice given that you usually don't have that kind of storage hanging around- not even in enterprise environments.
There's simply no comprehension amongst the ZFS developers that virtually EVERY raid card on the market supports such an operation. Even more shocking was when one developer said (paraphrasing) "gosh, how would one even go about doing that sort of thing?"
Don't get me wrong- checksumming and automatic disk scrubbing are features long overdue, but ZFS is not magic bullet.
Men can be drafted into military service. Women can't. I agree the insurance thing sucks, but the draft has the potential to suck a whole lot more.
Ff you were at the same gathering as I was (and it sounds like you were) the laptop thing was a bit of a glitch owing to the fact that that presentation was setup with burned DVD's that played find on a DVD player but didn't navigate correctly on the laptop they used.
They were movie files of some sort on the laptop. It wasn't the same "gathering", but it is good proof that the man apparently still can't get his shit together.
I don't think that skeptics worship Randi any more than Open-source enthusiasts "worship" Linus Torvalds.
The two other speakers at the presentation were so busy verbally sucking Randi's dick while whipping out rulers, they couldn't notice the audience was bored to tears. We clearly were not at the same presentation. This was outright "this is how great Randi is, this is how great I am for being this close to Randi." I really wish they had just let the man speak for himself.
I've personally attended a "skeptics" meeting where he was giving a talk. One thing that struck me was how much hero-worshiping was going on. Some guy spent a good 20-25 minutes telling us about how well he knew James Randi, how close he was to James Randi, how he could pick up the phone and call James Randi, yadda yadda.
This was after waiting about 30 minutes for them to start- they had to get a laptop working with various videos of James Randi on (mostly asian) TV. Each video, of course, did not play properly, or they played the wrong ones (ie, the same thing over and over.)
His work is worthwhile and he's decent showman, but he's also grossly over-sold and over-hyped. The devotion (if not downright worshiping) is hilarious, given that it is being done by a bunch of people who call themselves "skeptics."
If only these idiots were running more of my life.
Given California is easily the biggest nannystate in the union, if you live there- you've already got your wish.
The problem is between the computer and Microsoft via the tubes.
Bullshit; it's not just Microsoft. A long-standing complaint of mine about Apple Mail is that it does not show the true URL in an HTML email via tooltips or any other method. The only way the user has to see the URL is to copy it, then manually paste it into the address bar in Safari or Firefox. So, "Click here to login to your account" is impossible to verify without extra work.
It'd take one engineer about 10 minutes to code in such a display, and they can't be bothered. But OH BOY, in Leopard, I'll get fancy pre-formatted emails to use for sending vacation photos!
Similarly, for all the fuss about how secure and better Ubuntu is, you have to recompile Netatalk with custom options (and the instructions provided don't work) to enable secure login. Why? Because of OpenSSL/GPL licensing issues that have existed for several years. Has anyone bothered to rewrite the hundred or so lines of Netatalk code to use GnuTLS instead? Nope!
PS: For those of you about to tell me "use samba", Netatalk handily outperforms samba and supports full MacOS filenames.