I was hoping you were being fecetious, but by your statement by the end it doesn't sound like it. This attitute about "profiling" pisses me off.
Profiling, racial or otherwise, is a valid security technique. When applied correctly, it allows greater flexibility and accuracy than security policies alone can provide.
What was needed, instead of a blind knee-jerk reaction to profiling because of poor application, was a concerted effort to train police officers, airport security, and other security professionals how to do it properly. This would have actually increased our security while keeping airport hassles to a minimum.
Basically, poor profiling is little more than stereotyping, which might be good for an individual trying to avoid dangerous situations, but is practically useless for a security professional. Pulling someone over for no reason other than that he is a young black man in an old car is just stereotyping, and is useless. However, if it's a young black man driving eratically with smoke billowing out his windows you can probably guess what he is doing, and you'd be a terrible cop if you didn't pull them over to check it out. Note here that there are dozens of other scenarios where a person might be driving a little erratically, and it would warrant being pulled over, and they don't involve being young and black.
Also, with good profiling, at airport security there would be no reason to give someone who looks like they might be Arab any more than a slightly more careful look than a white mom with three kids in tow. However, if that Arab-looking guy looks a little nervous, it would be a very good idea to be extra thorough with his bag and screening, and maybe alert the guys in the rest of airport security of a suspicious character, so they can watch for signs of a threat. This does not mean pull him off to a side room for a strip search! Not unless you find suspicious materials in his bag, or if he exhibits more suspicious behavior. The fact is, the vast majority of recent terrorist attacks have been perpetrated by Arabic people, so giving people who look Arabic a slightly more careful look with a much higher sensitivity to suspicious behavior is just plain smart and doesn't hurt anybody but the occasional suspicious character.
What we have now are rules that pretty much eliminate the use of common sense, and a clever person could sneak between some of these idiotic rules because "Profiling is Bad".
the real problem with the molten iron core theory, happens to be, as any good metallurgist
Actually, he isn't a very good metallurgist if he doesn't also tell you that rotating currents of a conducting but non-magnetic substance will produce a magnetic field.
You're saying a molten iron core is implausible because it becomes non-magnetic at those temperatures in defense of a theory that says the conductive nature of the Ocean combined with ocean currents create magnetic fields.
I've got a bit of news for you, molten iron is still conductive and when it spins it creates a neat magnetic field.
The theory of molten metal creating a magnetic field was proven with molten lead in a large sphere - the scientists had a sphere that was designed to hold and contain molten lead (I think, much lower melting point than iron though whatever they used). They read no magnetic field before spinning, then spun the sphere and suddenly there was a neat magnetic field similar to the earth's. I don't remember the documentary, but I wathed it on the Discovery Channel, it was cool.
This doesn't "prove" that a molten iron core is causing the magnetic field, but it proves the concept. I could see oceanic currents influencing the field, but I think a molten core is much more likely to produce the substantial field we have. I.e. I could see oceanic currents being responsible for moving the magnetic poles, but not creating them in the first place.
Contrast that with Mars, which has no magnetic field and also has a solid, cool core (it also has no oceans though).
I believe they were talking about volume, though they didn't specify.
Champlain is 30% larger than this lake by volume, but this lake is about 80% the size of Champlain by volume. Using the second figure, with a bit of hyperbole you can say "roughly equivalent". They like to do that kind of crap when describing stuff on Mars and other earth-ish sized solar and planetary satellites.
Not the most accurate description, but it gives a rough idea of volume at least.
I'm pretty sure they could do that now and they'd be un-sue-able.
There is no such thing as "unsueable". You can sue anyone you want for anything you want. Most things if you tried to sue for you would lose - you need a pretty decent case in line with the law to actually win - and there is quite a lot that would be tossed immediately because it is rediculous, but even then it must be thrown out by a judge who hears at least the beginning of your case. You sued, but you lost at the very beginning.
If you meant the others could sue and Microsoft would win, you are probably right, but the further MS pushes it, the further a court case would drag out, and soon the rest of the industry will be in a position to drag a court case out. That would still generally work in MS's favor, but they have fallen on hard times like everybody else and will not want to waste millions of dollars on something that is rather frivolous. A few wording changes could actually make a legitimate case for IE8, it's a pretty solid browser with some nice features.
Man, I can't tell you how much I love that one little feature, the first time I saw it I just thought "Brilliant!".
One of my biggest pet-peeves is lost browser view-window space. I'm there to browse the web, not look at an interface - no matter how pretty that interface may be. I want it to look nice and have all the bells and whistles, but I -hate- the wasted space of FF and IE. FF is actually a little worse than IE8 for that (IE drops the tab bar if you only have one tab), but not much. I don't do browser-bar add-ons because they waste so much space.
It's one of my "things", and that plus the speed of Chrome are the two reasons I really like it.
I love how the article within the article comparing a high end (but not top end) HDD and a solid SSD makes a point of mentioning that SSD read/write metrics are generally stated in sequential operations, but fails to mention that random seeks have virtually zero performance penalty on an SSD. They don't bother to mention random read/write speeds because they are almost identical to sequential! Now, try a fragmented hard drive - there is a MASSIVE penalty for non-sequential reads on a hard drive. SSDs are actually fragmented in an orderly fashion by design. The nature of the technology means this significantly boosts performance. It also means data is much more carefully laid out on the SSD than hard disks.
It's also funny that they did a 1gb file copy of videos and photos - almost certainly dumped onto the hard drive all at once making it a folder full of large blocks of sequential data - optimum conditions for a hard disk but not very realistic, especially in a folder that would have files added and removed over a period of months.
I think it's odd that they gave the HDD optimum conditions, watched it lose in every single case - in most cases significantly, the 1gb sequential write is the only place it didn't lose by much in - and then decared it superior. I mean, seriously? Yeah, it's a lot cheaper, but it's also very inferior. It takes a $300 hard disk to match the performance of a... $300 SSD, OMG!
Their "expert" said most people won't notice quicker application load times - that's bullshit. Microsoft and Apple both have invested a lot of money in OS technologies that make the OS seem like it acts more quickly - loading parts of the desktop before everything is ready, caching commonly used applications, caching in the FIRST place, etc. Hard drive companies have -also- spent a ton of money making the drives feel like they are faster, things like "bursting" up to 64mb of cache to help mitigate the severe weakness a hard disk has in random reads and writes, pushing 7200 RPM electric motors into a 2.5" drive, etc.
A not-insignificant portion of modern advances in both hard drive and operating system technology have been about getting things to look like they are moving faster, regardless of whether or not they actually are moving much faster, and the author has the audacity to say people won't notice when things really ARE faster?
Go ahead and make the argument that the speed isn't worth the price tag right now, that's certainly a valid discussion. But don't close your eyes, plug your ears, and say "lalalala" to the fact that speed is really, really important to the average computer user. It may not be worth as much money to them as tech heads, but they sure do notice it, and they sure do love it.
I mean seriously, why the hell do you think Vista got such a poor reception? The only other major issue besides sluggishness at launch was drivers, and between them they prompted Microsoft to completely re-vamp and re-brand it almost immediately.
Plus, good 80gb drives can be had for about $180. Not enough for some people, but plenty for most. Anything that requires lots of hard drive space can go on a separate high capacity drive.
The upgrade is not all that expensive, and if it weren't for the fact that I have to drop $1500 on a new boiler in the next day or two, I'd be upgrading my laptop and home computer next week.
Another fun little tidbit - ALL SSD drives are currently in 2.5" format, they simply don't need the room for that many chips - you'd have to do terrabyte plus drives to fill that much space, and that would be hella expensive.
Try getting 15k rpm hard drive in a 2.5" form factor - you won't.
Oddly enough, the American Flag has no "freedom" or "liberty" stripes. You'll be happy to know it does have "Rhode Island" and "Massachusettes" stripes though, along with some others.
Some companies do this, they'll spend $500k on the system, claim it works, turn it over and razzle dazzle the project manager who signs it, then pocket the remaining $35.5 million. Of course, they don't pocket the $35.5 million after they are done, they pocket it up front and all throughout the project, so that by the end they've spent their $40 million, and then some most likely.
Then when it doesn't work the company just points to the paper the PM signed and says "You said it worked just fine, we're not supporting it any further."
I've got a quarter that says that is how the $28 million failure went.
Starting out with something sane is the way to go, but generally speaking large organizations aren't initially large organizations. They are small organizations that grow, sometimes splitting off into sub-organizations, sometimes merging with other organizations (creating some nice confusing heterogeny for payroll there, in particular), expanding to new regions, changing structure, etc.
A complete re-organization generally isn't possible without bankruptcy of some kind first.
This is why projects like this end up costing $40 million after failing with $28 million.
The fact is, you don't know shit about the problem, but you assume you have it all worked out, so you throw out a number and just say go. Then, when you start to realize with it will take to comply with city, local, state, and federal tax laws, as well as privacy laws, laws like S/O, not to mention INTERNAL company payroll needs. It's not too bad if it is a small organization operating in one little area, but as soon as you start crossing boarders of any kind, shit gets fucked up. Laws and regulations you've never even heard of almost certainly apply.
And you have to program it to comply with -all- of it. One little mistake could cost the organization millions.
There is a reason large organizations have teams of accountants/programmers, tax lawyers, accountant/lawyers to deal with this shit. It's not easy.
The punishment for stealing music worth less than $250 retail is a class 1 misdemeanor, 6 months in jail and/or a $2500 fine.
This is copyright infringement, and it is a whole different beast than stealing.
To test this, try stealing music you've already purchased - it's impossible. But you can sure infringe on the copyright on music you have already purchased!
I didn't realize the "adult" thing to do was to close your eyes and put everything on credit.
I thought bad debt was how we got into this mess, but apparently we just weren't spending enough!!
I think from now on I'll start balancing my budget by getting more credit cards. I didn't realise spending less than my income was the immature way to manage finances.
Your insight has helped me out a ton, mortgage re-finance, here I come!
The federal government has been shoving money down the throat of the States via the Bailout. They have to do something with it, this is a great opportunity for some cities that are too expensive to manage since the population/economic base has dwindled.
The alternative is doing things like massive studies that duplicate work, or do idiotic work that researchers know cannot produce any viable results. That second is happening where I am, a chemist friend of mine told me of a study funded by the bailout being done locally to track the diet of a particular type of bird over a long term period using fat samples. The trouble is, for this particular bird the diet and metabolism is such that the fat can only be used to look at about the last week of the bird's diet. That means to do any kind of long term study, these birds most be tested every week for YEARS. Researchers know this, and they know of much more efficient methods of getting the data, but they've got to spend the money somehow right?
That's a primo 80gb drive. The spinning disk equivalent is here. That's right, 147gb for $955.00. There is an HP drive at the same size for $320, but they only list the Ultra320 spec, not actual transfer rates, so I'm suspicious of it. In any case, U320 Raid controllers are $200+ as well. Double the capacity and sequential speed (NO spinning disk touches a SSD for random seeks, 2-3ms vs 0.1ms), 3-4x the price. Raid0 with two top of the line SSDs would be cheaper and less likely to fail than one of these drives, and would have the same capacity and speed. SSDs are also silent, these drives are the opposite of silent. You actually need hearing protection if you will be working near them on a day-to-day basis.
To match the performance of that 80gb drive with consumer level gear, you would need 3 decent 80gb SataII in a Raid0 configuration. Granted, that gets you a lot more capacity, but in most cases you will also need a $100 raid controller also. At $80-$100 a pop (cheapest 10krpm drive I could find, cost per gig drops as you go higher but price always rises) You are looking at between $340 and $400. You also have 4x the failure risk with no fault tolerance. A single drive goes and all the data is gone. Add to the fact that the mean time between failures is about 3-5 years on these drives, versus about 130+ years for the SSD, and your advantages start to disappear.
To take care of the reliability issues, you can certainly go with a Raid10 or Raid5 configuration - Raid5 won't cost you anything more but only provides a little fault tolerance, Raid10 will provide great fault tolerance and great speed, but also requires 6 disks minimum (don't quote me, but it's at least 4), so you're now looking at almost double the price, and probably a beefier RAID card, but maybe not.
In other words, it depends on what you use it for. For capacity, nothing beats slow spinning disc. For speed, it's cheaper to RAID up 2-3 SSD drives in a 0 config (very high MTBF of SSD justifies this, it would be suicide with conventional) than anything else you can do.
Theoretically speaking (in practical terms it is much less, but still very high) a Flash based hard drive will last 10,000 full writes of the entire hard drive. In practice, because we usually don't write a full hard drive, empty it, and then write it again it will be less than that, however wear leveling technology looks very sound and would only drop the number of full hard drive writes (in terms of data written) to 4,000+, and I'm being conservative in my estimation.
How many people write 320TB of data in 5 years? That's how much you would need to write before an 80gb SSD fails based on my 4,000 writes number (which is conservative). A more conservative number of 1,000 writes still requires 80TB of data in 5 years to cause the flash drive to fail. The theoretical limit is at least 800tb of data before it fails. For a 250gb drive, the numbers from most conservative to theoretical are: 250TB, 1PB (1 million gb), and 2.5PB.
Physical disks have a tiny but constant chance to fail in the act of spinning up. They more they are used the faster they wear and the sooner they fail. A conventional 250gb hard drive writing 250tb of data over its lifetime will one of the drives pulling the average down, not up. If you can write that much, your disk will fail very quickly.
Flash failure is usage related, not time related. A flash drive written once and then stored in a safe place would last at least 100 years and still function like new, with 9,999 writes left to go. Also note that reads do not wear flash drives, but even sitting unused but powered up a conventional drive wears down, and reads wear almost as much as writes do. Catastrophic failure of a conventional drive often means lost data, even with expensive data recovery services. Not so for SSD.
So what you have is a drive with unlimited reads (even after failure - the data simply cannot be overwritten if flash fails) and a significant, predictable number of writes vs a drive that wears constantly whenever it is powered and has the potential to fail at any time - though the usually last a while.
The -only- disadvantage an SSD has vs a conventional drive is capacity. That's it. SSDs trump hard disks in every other catagory.
Just my personal opinion, but I think for wear-leveling purposes the OS should do no more than flag for TRIM (the delete flag would be fine for this) until a de-frag is run. In this way, the data can be re-organized and both speed and wear-leveling can be more consistantly maintained with factory specs.
Also don't for get that full conventional hard-drives suffer significant speed degradation as well. The tools are already built into the OS to fix this now, but they didn't used to be. We're at the same place with SSDs. And even with performance degraded because of a full disk, a good SSD is still a lot faster than the fastest consumer magnetic disk.
Hibernate dumps ram to disk and shuts the machine off.
Suspend (or sleep) simply supplies minimal power to everything - the most agressive basically kills everything except a very small amount of power to keep the contents of RAM from disappearing. Less agressive sleep states can leave your NIC available and hard disk writable, but they consume a lot more power than an agressive sleep state.
From what some have said about SSDs, hibernate should come up only a few seconds slower than suspend. Pretty cool, might make me believe in hibernate after all.
However, if you only write to it once (i.e. long-term storage) flash storage will never degrade.
Well, we're talking long enough for the metal itself to deteriorate via weathering and such, at which point you'll be dead and gone!
Every time a spinning disk is plugged in, there is a small chance it will fail just in the act of spinning up. It's tiny sure, but it's there for spinning disks and not for SSDs.
Honestly, the flash drive is capable of so many writes (it's in the thousands, and the drives write every bit on the disk before it will go back and re-write) that they likely outlast conventional drives in all but the most write-heavy applications.
Plus, when a flash drive fails, the failure does not prevent reads, only writes. Recovery would be as simple as plugging it in and pulling your data off. With conventional disks, a physical failure -often- results in unrecoverable data. The reason data recovery services cost thousands of dollars is because it is extremely difficult, if not impossible, to recover the data in many cases. Trust me, I know a number of people who have gotten backup religion only after spending thousands of dollars to fail to recover a critical piece of data (one was a single file needed for a massive audit - all was recovered except that file). The ability to read even after a failure makes SSDs -significantly- safer for data storage.
Lastly, the speed boost is HUGE. Go check the read/write numbers for SSDs vs the top end consumer drives. The worst SSDs on the market only do a little better than the best conventional drives, the benchmark I saw put various SSDs against a top of the line WD Raptor - 12k rpm I think. Due to a poor controler design, the lower end drives only performed a little better. The high end drives made the Raptor look like a complete joke. Even with the performance flaw another slashdot article mentioned (drives aren't as fast after they have been filled up, it is fixable on most drives) the top SSDs were still -significantly- faster than the Raptor. I think the prices were even similar for the same capacity, but I could be remembering funny on that one. You definitely don't get 12krpm out of a 1tb hard drive though, they are still in the 300gb range for the consumer market if I remember right.
I second this advice. Do what you love. If you love what you do, you will be passionate about it and continually strive to improve. This is HUGE for employers, good ones that can see past next quarter's earnings anyway. What you love to do needs to be useful to an employer, but the bottom line is if you love it you will be good at it. If you are good at it you will progress in your career.
It sounds like you're smart but have a very general degree. This is not a problem, but you should start getting some professional certifications - Microsoft, Cisco, what have you. This will do two things for you: it will help you decide what you like to do and what you don't, as they all require a good deal of study to pass, and they look really good on a resume. What certifications you put down will help focus the employer's attention on what you are capable of, as well as provide them with a baseline.
In truth, you should have started in helpdesk while you were still in school. Helpdesk positions are not the best place for breaking into the rest of the market any more (they used to be great for that, not so much now), but given enough time something might turn up.
I would stick with your current job, but work on those certs, and be constantly searching for an entry-level sysadmin or unix admin (very rare, but huge potential if you can get it!) position. Or, once you have some cableing/cisco (sorry, they've got the most widely transferable certs even if you don't like them) certifications you should be able to break into the entry level physical network market.
Valuable things to take away from your helpdesk position are customer relations, troubleshooting skills, and general industry familiarity.
Oh yeah, and make sure your writing and communications skills are up to par. Chances are non-techs will need to read and understand something you write at some point, and sucky writing can be a career limiter.
Slashdot Mirror
I was hoping you were being fecetious, but by your statement by the end it doesn't sound like it. This attitute about "profiling" pisses me off.
Profiling, racial or otherwise, is a valid security technique. When applied correctly, it allows greater flexibility and accuracy than security policies alone can provide.
What was needed, instead of a blind knee-jerk reaction to profiling because of poor application, was a concerted effort to train police officers, airport security, and other security professionals how to do it properly. This would have actually increased our security while keeping airport hassles to a minimum.
Basically, poor profiling is little more than stereotyping, which might be good for an individual trying to avoid dangerous situations, but is practically useless for a security professional. Pulling someone over for no reason other than that he is a young black man in an old car is just stereotyping, and is useless. However, if it's a young black man driving eratically with smoke billowing out his windows you can probably guess what he is doing, and you'd be a terrible cop if you didn't pull them over to check it out. Note here that there are dozens of other scenarios where a person might be driving a little erratically, and it would warrant being pulled over, and they don't involve being young and black.
Also, with good profiling, at airport security there would be no reason to give someone who looks like they might be Arab any more than a slightly more careful look than a white mom with three kids in tow. However, if that Arab-looking guy looks a little nervous, it would be a very good idea to be extra thorough with his bag and screening, and maybe alert the guys in the rest of airport security of a suspicious character, so they can watch for signs of a threat. This does not mean pull him off to a side room for a strip search! Not unless you find suspicious materials in his bag, or if he exhibits more suspicious behavior. The fact is, the vast majority of recent terrorist attacks have been perpetrated by Arabic people, so giving people who look Arabic a slightly more careful look with a much higher sensitivity to suspicious behavior is just plain smart and doesn't hurt anybody but the occasional suspicious character.
What we have now are rules that pretty much eliminate the use of common sense, and a clever person could sneak between some of these idiotic rules because "Profiling is Bad".
the real problem with the molten iron core theory, happens to be, as any good metallurgist
Actually, he isn't a very good metallurgist if he doesn't also tell you that rotating currents of a conducting but non-magnetic substance will produce a magnetic field.
You're saying a molten iron core is implausible because it becomes non-magnetic at those temperatures in defense of a theory that says the conductive nature of the Ocean combined with ocean currents create magnetic fields.
I've got a bit of news for you, molten iron is still conductive and when it spins it creates a neat magnetic field.
The theory of molten metal creating a magnetic field was proven with molten lead in a large sphere - the scientists had a sphere that was designed to hold and contain molten lead (I think, much lower melting point than iron though whatever they used). They read no magnetic field before spinning, then spun the sphere and suddenly there was a neat magnetic field similar to the earth's. I don't remember the documentary, but I wathed it on the Discovery Channel, it was cool.
This doesn't "prove" that a molten iron core is causing the magnetic field, but it proves the concept. I could see oceanic currents influencing the field, but I think a molten core is much more likely to produce the substantial field we have. I.e. I could see oceanic currents being responsible for moving the magnetic poles, but not creating them in the first place.
Contrast that with Mars, which has no magnetic field and also has a solid, cool core (it also has no oceans though).
Which means the fact that the pole is moving further north is difinative proof that global warming is a hoax!! :P
I believe they were talking about volume, though they didn't specify.
Champlain is 30% larger than this lake by volume, but this lake is about 80% the size of Champlain by volume. Using the second figure, with a bit of hyperbole you can say "roughly equivalent". They like to do that kind of crap when describing stuff on Mars and other earth-ish sized solar and planetary satellites.
Not the most accurate description, but it gives a rough idea of volume at least.
I'm pretty sure they could do that now and they'd be un-sue-able.
There is no such thing as "unsueable". You can sue anyone you want for anything you want. Most things if you tried to sue for you would lose - you need a pretty decent case in line with the law to actually win - and there is quite a lot that would be tossed immediately because it is rediculous, but even then it must be thrown out by a judge who hears at least the beginning of your case. You sued, but you lost at the very beginning.
If you meant the others could sue and Microsoft would win, you are probably right, but the further MS pushes it, the further a court case would drag out, and soon the rest of the industry will be in a position to drag a court case out. That would still generally work in MS's favor, but they have fallen on hard times like everybody else and will not want to waste millions of dollars on something that is rather frivolous. A few wording changes could actually make a legitimate case for IE8, it's a pretty solid browser with some nice features.
Man, I can't tell you how much I love that one little feature, the first time I saw it I just thought "Brilliant!".
One of my biggest pet-peeves is lost browser view-window space. I'm there to browse the web, not look at an interface - no matter how pretty that interface may be. I want it to look nice and have all the bells and whistles, but I -hate- the wasted space of FF and IE. FF is actually a little worse than IE8 for that (IE drops the tab bar if you only have one tab), but not much. I don't do browser-bar add-ons because they waste so much space.
It's one of my "things", and that plus the speed of Chrome are the two reasons I really like it.
I love how the article within the article comparing a high end (but not top end) HDD and a solid SSD makes a point of mentioning that SSD read/write metrics are generally stated in sequential operations, but fails to mention that random seeks have virtually zero performance penalty on an SSD. They don't bother to mention random read/write speeds because they are almost identical to sequential! Now, try a fragmented hard drive - there is a MASSIVE penalty for non-sequential reads on a hard drive. SSDs are actually fragmented in an orderly fashion by design. The nature of the technology means this significantly boosts performance. It also means data is much more carefully laid out on the SSD than hard disks.
It's also funny that they did a 1gb file copy of videos and photos - almost certainly dumped onto the hard drive all at once making it a folder full of large blocks of sequential data - optimum conditions for a hard disk but not very realistic, especially in a folder that would have files added and removed over a period of months.
I think it's odd that they gave the HDD optimum conditions, watched it lose in every single case - in most cases significantly, the 1gb sequential write is the only place it didn't lose by much in - and then decared it superior. I mean, seriously? Yeah, it's a lot cheaper, but it's also very inferior. It takes a $300 hard disk to match the performance of a... $300 SSD, OMG!
Their "expert" said most people won't notice quicker application load times - that's bullshit. Microsoft and Apple both have invested a lot of money in OS technologies that make the OS seem like it acts more quickly - loading parts of the desktop before everything is ready, caching commonly used applications, caching in the FIRST place, etc. Hard drive companies have -also- spent a ton of money making the drives feel like they are faster, things like "bursting" up to 64mb of cache to help mitigate the severe weakness a hard disk has in random reads and writes, pushing 7200 RPM electric motors into a 2.5" drive, etc.
A not-insignificant portion of modern advances in both hard drive and operating system technology have been about getting things to look like they are moving faster, regardless of whether or not they actually are moving much faster, and the author has the audacity to say people won't notice when things really ARE faster?
Go ahead and make the argument that the speed isn't worth the price tag right now, that's certainly a valid discussion. But don't close your eyes, plug your ears, and say "lalalala" to the fact that speed is really, really important to the average computer user. It may not be worth as much money to them as tech heads, but they sure do notice it, and they sure do love it.
I mean seriously, why the hell do you think Vista got such a poor reception? The only other major issue besides sluggishness at launch was drivers, and between them they prompted Microsoft to completely re-vamp and re-brand it almost immediately.
Plus, good 80gb drives can be had for about $180. Not enough for some people, but plenty for most. Anything that requires lots of hard drive space can go on a separate high capacity drive.
The upgrade is not all that expensive, and if it weren't for the fact that I have to drop $1500 on a new boiler in the next day or two, I'd be upgrading my laptop and home computer next week.
Another fun little tidbit - ALL SSD drives are currently in 2.5" format, they simply don't need the room for that many chips - you'd have to do terrabyte plus drives to fill that much space, and that would be hella expensive.
Try getting 15k rpm hard drive in a 2.5" form factor - you won't.
Oddly enough, the American Flag has no "freedom" or "liberty" stripes. You'll be happy to know it does have "Rhode Island" and "Massachusettes" stripes though, along with some others.
Sorry, couldn't help myself, good post. :)
Well at least he didn't call you incestuous, you insensitive clod!!
Some companies do this, they'll spend $500k on the system, claim it works, turn it over and razzle dazzle the project manager who signs it, then pocket the remaining $35.5 million. Of course, they don't pocket the $35.5 million after they are done, they pocket it up front and all throughout the project, so that by the end they've spent their $40 million, and then some most likely.
Then when it doesn't work the company just points to the paper the PM signed and says "You said it worked just fine, we're not supporting it any further."
I've got a quarter that says that is how the $28 million failure went.
But Kenya's got lions!!
And tigers too.
Starting out with something sane is the way to go, but generally speaking large organizations aren't initially large organizations. They are small organizations that grow, sometimes splitting off into sub-organizations, sometimes merging with other organizations (creating some nice confusing heterogeny for payroll there, in particular), expanding to new regions, changing structure, etc.
A complete re-organization generally isn't possible without bankruptcy of some kind first.
This is why projects like this end up costing $40 million after failing with $28 million.
The fact is, you don't know shit about the problem, but you assume you have it all worked out, so you throw out a number and just say go. Then, when you start to realize with it will take to comply with city, local, state, and federal tax laws, as well as privacy laws, laws like S/O, not to mention INTERNAL company payroll needs. It's not too bad if it is a small organization operating in one little area, but as soon as you start crossing boarders of any kind, shit gets fucked up. Laws and regulations you've never even heard of almost certainly apply.
And you have to program it to comply with -all- of it. One little mistake could cost the organization millions.
There is a reason large organizations have teams of accountants/programmers, tax lawyers, accountant/lawyers to deal with this shit. It's not easy.
See my sig, I can't say it better than that.
It isn't stealing music.
The punishment for stealing music worth less than $250 retail is a class 1 misdemeanor, 6 months in jail and/or a $2500 fine.
This is copyright infringement, and it is a whole different beast than stealing.
To test this, try stealing music you've already purchased - it's impossible. But you can sure infringe on the copyright on music you have already purchased!
Class 1 misdemeanor, 6 months in jail and/or $2500 fine.
She would have been better off shoplifting.
Ahh, thanks for that.
I didn't realize the "adult" thing to do was to close your eyes and put everything on credit.
I thought bad debt was how we got into this mess, but apparently we just weren't spending enough!!
I think from now on I'll start balancing my budget by getting more credit cards. I didn't realise spending less than my income was the immature way to manage finances.
Your insight has helped me out a ton, mortgage re-finance, here I come!
The federal government has been shoving money down the throat of the States via the Bailout. They have to do something with it, this is a great opportunity for some cities that are too expensive to manage since the population/economic base has dwindled.
The alternative is doing things like massive studies that duplicate work, or do idiotic work that researchers know cannot produce any viable results. That second is happening where I am, a chemist friend of mine told me of a study funded by the bailout being done locally to track the diet of a particular type of bird over a long term period using fat samples. The trouble is, for this particular bird the diet and metabolism is such that the fat can only be used to look at about the last week of the bird's diet. That means to do any kind of long term study, these birds most be tested every week for YEARS. Researchers know this, and they know of much more efficient methods of getting the data, but they've got to spend the money somehow right?
That's a primo 80gb drive. The spinning disk equivalent is here. That's right, 147gb for $955.00. There is an HP drive at the same size for $320, but they only list the Ultra320 spec, not actual transfer rates, so I'm suspicious of it. In any case, U320 Raid controllers are $200+ as well. Double the capacity and sequential speed (NO spinning disk touches a SSD for random seeks, 2-3ms vs 0.1ms), 3-4x the price. Raid0 with two top of the line SSDs would be cheaper and less likely to fail than one of these drives, and would have the same capacity and speed. SSDs are also silent, these drives are the opposite of silent. You actually need hearing protection if you will be working near them on a day-to-day basis.
To match the performance of that 80gb drive with consumer level gear, you would need 3 decent 80gb SataII in a Raid0 configuration. Granted, that gets you a lot more capacity, but in most cases you will also need a $100 raid controller also. At $80-$100 a pop (cheapest 10krpm drive I could find, cost per gig drops as you go higher but price always rises) You are looking at between $340 and $400. You also have 4x the failure risk with no fault tolerance. A single drive goes and all the data is gone. Add to the fact that the mean time between failures is about 3-5 years on these drives, versus about 130+ years for the SSD, and your advantages start to disappear.
To take care of the reliability issues, you can certainly go with a Raid10 or Raid5 configuration - Raid5 won't cost you anything more but only provides a little fault tolerance, Raid10 will provide great fault tolerance and great speed, but also requires 6 disks minimum (don't quote me, but it's at least 4), so you're now looking at almost double the price, and probably a beefier RAID card, but maybe not.
In other words, it depends on what you use it for. For capacity, nothing beats slow spinning disc. For speed, it's cheaper to RAID up 2-3 SSD drives in a 0 config (very high MTBF of SSD justifies this, it would be suicide with conventional) than anything else you can do.
Just an update to my previous long winded post, wish I had seen this earlier:
Mean Time Between Fail on an Intel X25-M 80gb SSD is *drumroll* 1.2 million hours, or 136 years and change.
Just a touch more than 5 years.
Also has a max shock resistance of 1,000G, beat that with a spinning disk!
Theoretically speaking (in practical terms it is much less, but still very high) a Flash based hard drive will last 10,000 full writes of the entire hard drive. In practice, because we usually don't write a full hard drive, empty it, and then write it again it will be less than that, however wear leveling technology looks very sound and would only drop the number of full hard drive writes (in terms of data written) to 4,000+, and I'm being conservative in my estimation.
How many people write 320TB of data in 5 years? That's how much you would need to write before an 80gb SSD fails based on my 4,000 writes number (which is conservative). A more conservative number of 1,000 writes still requires 80TB of data in 5 years to cause the flash drive to fail. The theoretical limit is at least 800tb of data before it fails. For a 250gb drive, the numbers from most conservative to theoretical are: 250TB, 1PB (1 million gb), and 2.5PB.
Physical disks have a tiny but constant chance to fail in the act of spinning up. They more they are used the faster they wear and the sooner they fail. A conventional 250gb hard drive writing 250tb of data over its lifetime will one of the drives pulling the average down, not up. If you can write that much, your disk will fail very quickly.
Flash failure is usage related, not time related. A flash drive written once and then stored in a safe place would last at least 100 years and still function like new, with 9,999 writes left to go. Also note that reads do not wear flash drives, but even sitting unused but powered up a conventional drive wears down, and reads wear almost as much as writes do. Catastrophic failure of a conventional drive often means lost data, even with expensive data recovery services. Not so for SSD.
So what you have is a drive with unlimited reads (even after failure - the data simply cannot be overwritten if flash fails) and a significant, predictable number of writes vs a drive that wears constantly whenever it is powered and has the potential to fail at any time - though the usually last a while.
The -only- disadvantage an SSD has vs a conventional drive is capacity. That's it. SSDs trump hard disks in every other catagory.
Just my personal opinion, but I think for wear-leveling purposes the OS should do no more than flag for TRIM (the delete flag would be fine for this) until a de-frag is run. In this way, the data can be re-organized and both speed and wear-leveling can be more consistantly maintained with factory specs.
Also don't for get that full conventional hard-drives suffer significant speed degradation as well. The tools are already built into the OS to fix this now, but they didn't used to be. We're at the same place with SSDs. And even with performance degraded because of a full disk, a good SSD is still a lot faster than the fastest consumer magnetic disk.
Just my 2 cents.
Hibernate dumps ram to disk and shuts the machine off.
Suspend (or sleep) simply supplies minimal power to everything - the most agressive basically kills everything except a very small amount of power to keep the contents of RAM from disappearing. Less agressive sleep states can leave your NIC available and hard disk writable, but they consume a lot more power than an agressive sleep state.
From what some have said about SSDs, hibernate should come up only a few seconds slower than suspend. Pretty cool, might make me believe in hibernate after all.
However, if you only write to it once (i.e. long-term storage) flash storage will never degrade.
Well, we're talking long enough for the metal itself to deteriorate via weathering and such, at which point you'll be dead and gone!
Every time a spinning disk is plugged in, there is a small chance it will fail just in the act of spinning up. It's tiny sure, but it's there for spinning disks and not for SSDs.
Honestly, the flash drive is capable of so many writes (it's in the thousands, and the drives write every bit on the disk before it will go back and re-write) that they likely outlast conventional drives in all but the most write-heavy applications.
Plus, when a flash drive fails, the failure does not prevent reads, only writes. Recovery would be as simple as plugging it in and pulling your data off. With conventional disks, a physical failure -often- results in unrecoverable data. The reason data recovery services cost thousands of dollars is because it is extremely difficult, if not impossible, to recover the data in many cases. Trust me, I know a number of people who have gotten backup religion only after spending thousands of dollars to fail to recover a critical piece of data (one was a single file needed for a massive audit - all was recovered except that file). The ability to read even after a failure makes SSDs -significantly- safer for data storage.
Lastly, the speed boost is HUGE. Go check the read/write numbers for SSDs vs the top end consumer drives. The worst SSDs on the market only do a little better than the best conventional drives, the benchmark I saw put various SSDs against a top of the line WD Raptor - 12k rpm I think. Due to a poor controler design, the lower end drives only performed a little better. The high end drives made the Raptor look like a complete joke. Even with the performance flaw another slashdot article mentioned (drives aren't as fast after they have been filled up, it is fixable on most drives) the top SSDs were still -significantly- faster than the Raptor. I think the prices were even similar for the same capacity, but I could be remembering funny on that one. You definitely don't get 12krpm out of a 1tb hard drive though, they are still in the 300gb range for the consumer market if I remember right.
I second this advice. Do what you love. If you love what you do, you will be passionate about it and continually strive to improve. This is HUGE for employers, good ones that can see past next quarter's earnings anyway. What you love to do needs to be useful to an employer, but the bottom line is if you love it you will be good at it. If you are good at it you will progress in your career.
It sounds like you're smart but have a very general degree. This is not a problem, but you should start getting some professional certifications - Microsoft, Cisco, what have you. This will do two things for you: it will help you decide what you like to do and what you don't, as they all require a good deal of study to pass, and they look really good on a resume. What certifications you put down will help focus the employer's attention on what you are capable of, as well as provide them with a baseline.
In truth, you should have started in helpdesk while you were still in school. Helpdesk positions are not the best place for breaking into the rest of the market any more (they used to be great for that, not so much now), but given enough time something might turn up.
I would stick with your current job, but work on those certs, and be constantly searching for an entry-level sysadmin or unix admin (very rare, but huge potential if you can get it!) position. Or, once you have some cableing/cisco (sorry, they've got the most widely transferable certs even if you don't like them) certifications you should be able to break into the entry level physical network market.
Valuable things to take away from your helpdesk position are customer relations, troubleshooting skills, and general industry familiarity.
Oh yeah, and make sure your writing and communications skills are up to par. Chances are non-techs will need to read and understand something you write at some point, and sucky writing can be a career limiter.