It's not an assumption, it's part of the definition of MTBF! Look, I don't expect anyone to just accept the fact that I know what I'm talking about because it's my job and I've had years of statistical training, so check around.
But use a little common sense. If a hard drive has an MTBF of something in excess of a hundred years, do you really expect that the typical hard drive will last longer than you will before it experiences a single failure? Honestly...do you think that the average hard drive will last over 100 years of continuous operation?
I'm sorry, but that's not what MTBF means. "Mean Time Between Failure" is a very specific statistical measure of the failure rate of a piece of equipment, not just hard drives.
I'm sure that there's a measure for the rate of bit-level defects on a drive, but it isn't MTBF.
That's pretty much the same arguement that I heard at the beginning of the migration from 5.25 to 3.5. There'll be early adopters, just like with every technology and, at some point, 2.5" drives will become the standard.
MTBF is not a measure of how long the drive will last, it's a measure of how often you should expect a failure, given a program of maintenance and drive replacement at the end of its useful life.
The end of a drive's useful life isn't when it fails, by the way. You'll probably have to pull the manufacturer's toenails to find out what their recommendation is (although maybe the warranty period might be a clue).
So if you follow the maintenance and replacement schedule, you will have a drive failure only once very hours. If you've got a slew of drives in your server room, you should expect a failure of any single drive every/n hours, where n is the number of drives.
So, yes, the numbers are accurate. Got a single drive? Maintain and replace it proplerly and you'll only be unpleasantly surprised once every 160 years.
That's not exactly correct, either. MTBF is the amount of time that you can expect before the failure of a single drive, assuming that it is properly maintained and replaced when its useful service life has been reached.
MTBF assumes that you are not running the device until it just wears out. The figure assumes that you recognize that the device has a finite lifetime and that you are prepared to replace it before it reaches that lifetime.
So, if you follow that guideline, you should expect a single device failure every 1.4 million hours for a single device. From that point, your explanation is correct - what is not correct is assuming that MTBF has anything to do with an indication of when it is necessary to replace a device...MTBF assumes that you are replacing the devices before they wear out.
You've got that wrong. Under the 1909 Act, all you had to do was publish with notice to have a copyright. Most works were never registered with the copyright office until their renewal term.
Not so. That was only the first step. You also had to provide two copies of the work to the Copyright Office. Then you received a certificate entitleing you to a copyright on the work.
In fact, it's entirely possible that the copyright was voided if Guthrie didn't send his copies in - the 1909 law provided for that.
I never understood why this logic has not been applied (legally at least) to Happy Birthday which was music originally published in the 1890's (as "Good Morning to All/You"), had the alternative lyrics published or alluded to a number of times from 1909 onwards, but wasn't copyrighted until 1935 by one of the original authors (who didn't even write the changed lyrics!) and is still under copyright until something like 2030, allowing the current holders to sue the Girl Guides and force restaurants to come up with those silly hand clapping songs.
The reason is because copyright laws have changed. Back in the day, you didn't automatically get a copyright by virtue of publishing - you had to apply for a copyright. The date of publication didn't have any bearing on the date of copyright.
I'm sure that in the case of Ludlow, they applied for the copyright when they purchased the rights from Guthrie (or whomever they bought them from), so they must figure that the clock started ticking at that point.
For those of you who make fun of the Soviet system wen you probably wheren't even born then, this is a lesson: Soviets took care of their people well and their medicine was top.
You're kidding, of course. Although the USSR's health care system was universal, the quality was utterly abyssmal for the average citizen.
I was unfortunate enough to see first-hand the state of Soviet-era medical facilities and the quality of care in the mid 1980's. Many third-world countries had much better medical care than that of the "typical" Soviet hospital that we toured. And, given that this was a state-sponsored tour (as was everything that we saw), I suspect that it was something better than typical.
Actually, it's a function of the PABX system that they are using - it's not that they have an 800 number (although plenty of places with 800 numbers have PABX's), it's that they are using a PABX that ignores the flag that suppresses the caller ID info. Or maybe it can read the ANI info.
There's a reason they tell you not to look at eclipses, and that's because there is sunlight that can damage your eyes can't be "seen" and as such you don't look away...
Ah yes, here it is - "Damage to the eyes comes predominantly from invisible infrared wavelengths, so the fact that you feel no discomfort while gazing at a partial eclipse does not guarantee that your eyes are safe"
Holy cow, if the IR from those floods is strong enough to damage your eyes, you'll definitely feel it somewhere else. IR causes burns to the cornea. You're thinking of UV, which can damage the retina because it is relatively high energy radiation.
The reason that an eclipse causes corneal burning is because the sun pumps out a tremendous amount of IR...it's a huge fusion furnace, after all.
It wasn't "the House of Representatives", it was Representative Duncan Hunter, a San Diego Republican who makes Rush Limbaugh look moderate...and that's coming from a registered Republican!
Work closely with your college's alumni association and with the CS department's industry liason. Both of them are excellent resources for job placement assistance.
Don't forget the power of social networking, either. I was lucky enough to get my "dream job" before I graduated (BSEE) because a friend of a friend was a manager at the company. In fact, that may be the best way to get the job, regardless of your experience.
Thirdly, consider joining the Computer Society of the IEEE and attend the functions, email with the members and even consider volunteering for some of the tasks that come with the Society. The membership dues are significantly reduced for recent graduates. Also, the IEEE's GOLD (Graduates of the Last Decade) organization can help as well.
And don't hesitate to apply for a job, even if you don't think that you meet the experience criteria. Even though many of the resumes are screened by HR and you may not get past them, many are not. Something in your resume may stick out and get you an interview - and that's what you're really after.
Given that your reasoning is both simplistic and polemic and that it appears that, in the face of reason you still cling to your fallacies, I'd say that I'd be wasting my time attempting to explain the rather obvious conclusions of the creators of our republican system. You're clearly not aware of the history and reasoning behind their decision (or you choose to ignore it) or you wouldn't be clinging to the worn out arguments that just don't have any basis in fact.
Since my time is worth something, I'll spend it elsewhere, comforted by the hope that perhaps in time you'll take my advice and actually learn something about the history of the founding of our system of government instead of simply repeating nonsense that was discredited years ago.
Don't buy into the PC thought of the present. The issue of slave ownership was certainly present in the day, but there were other, more principled ideas behind both the electoral college and the Senate. Some significant issues were those of taxation and commerce and what effect states with large populations would have on those with small - and vice versa. The fact that slavery existed and was considered does not invalidate the concept. That's a red herring.
I don't claim that one adult = one vote is undue influence and neither did the creators of our system of government. They were looking at a broad picture of influence. In fact, the system that governs us (in the US) has a great many features that tend to override the one person, one vote rule. A specific example is the referendum process. A referendum, approved by the majority of the people can be declared invalid if it infringes upon the rights of a minority. Admittedly, that is somewhat tangental to the idea of the electoral college and the Senate, but the underlying principle is sound.
And, finally, the electoral college does not create a situation that allows two small states to be more influential than a large one. Neither does the Senate. The electoral college somewhat dilutes the power of large states while it somewhat increases the power of small states. And in the Senate, all states are equal. Taken as a whole, though, Congress somewhat dilutes the power of large states and somewhat increases the power of small states.
Anyway, like I said, take a civics class. Or read some scholarly books on the subject. These ideas were really quite well discussed and thought out. They didn't just pop out of nowhere.
For all of those who can't see the underlying fairness of the electoral college, I would suggest taking a civics course from your local community college. Also, you may want to ponder on just why the US Senate is organized the way that it is.
Both were set up with particular aims in mind - chief among them that states with large populations would not be able to use those large populations to unduly influence elections and legislation to the deteriment of sparsely populated states. The electoral college also dovetails quite nicely with the concept of a democratic republic.
I think that if you filled the box with Apoxie, a little piece of velcro would easily hold the lid closed. It would certainly solve the mechanical problems! And you're right, heat is not a problem with 9V. Even with two 9V batteries, heat isn't an issue.
Filling the tin is a great idea - maybe I'll give it a try and see what happens!
It's the tins. The electronics are plenty durable because the circuit board is very small. But the typical Panasonic volume control and the micro switch will probably break because tin doesn't provide them with enough support in a catastrophic flying event. Your choices are pretty limited for 10K stereo pots that will fit in a small tin (and not sound like fingernails scratching on a chalkboard) and the leads that were really designed for plugging into a PCB are kind of fragile. And if the lid pops open during its mid-air acrobatics and the battery starts flailing around, something will almost certainly break. That, and the tins are surprisingly non-durable when they aren't full of mints;-)
I agree that the Hammond case isn't nearly as stylish as the Altoids or Penguin case. I guess my deal is that since I sell them, I'd just as soon not have one come back because the case wasn't strong enough to stand up. On the other hand, if somebody REALLY wants it in an Altoids tin, I'll build it that way for them...just so they know that they have to be pretty careful. I guess that's really the key to using a mint tin as a case. I always tried to be careful, but then I'd forget what was on the other end of my headphones and -zip- off it would go.
Also, you really should power the CMoys with two 9V batteries. I know that you can do it with one, but that only gives you about one volt of margin before the opamp starts to clip and if the rails become unbalanced (something that the resistor divider power supply is prone to do), then the margin can suddenly drop to nothing. The opamps that work best (OPA2132, 2134, etc.) need +/-3V just to operate. That leaves only 1.5V peak for the audio signal...OK if your cans are sensitive, but if you've got something like my Grado SR-125's, you'll be in trouble. Anyway, making a short story long, the other reason that I like the Hammond case is because it will let another battery fit in or a slightly larger PCB with a TI TLE2426 rail splitter to keep the voltage rails in balance.
Well, as a blatent plug for me, you can buy any one of those players, a low cost set of quality headphones and one of these. The sound quality is almost always a product of the low budget headphones combined with low voltage (which causes clipping which causes distortion). I use a Neuros player which supports OGG, lasts for between 10 and 12 hours per charge and sounds pretty good with that CMoy amp and a ~$50 pair of headphones.
$200 for the Neuros $50 for the 'phones (Grado SR-60) $50 or $60 for the amp
Just be aware that if you put it in a mint tin, you'll probably break it within a couple of weeks. Those tins just aren't beefy enough. Even on your desk. I sent my Cmoy amp zipping across the floor when I turned around one day and my headphone cord yanked it. I build all of mine in a nice extruded case from Hammond.
Yeah, I'll second that. I've had one since last summer and I love it. The 2.13 firmware works fine for me and I guess I'll be upgrading to USB 2.0 now.
Slashdot Mirror
It's not an assumption, it's part of the definition of MTBF! Look, I don't expect anyone to just accept the fact that I know what I'm talking about because it's my job and I've had years of statistical training, so check around.
But use a little common sense. If a hard drive has an MTBF of something in excess of a hundred years, do you really expect that the typical hard drive will last longer than you will before it experiences a single failure? Honestly...do you think that the average hard drive will last over 100 years of continuous operation?
=h=
I'm sorry, but that's not what MTBF means. "Mean Time Between Failure" is a very specific statistical measure of the failure rate of a piece of equipment, not just hard drives.
I'm sure that there's a measure for the rate of bit-level defects on a drive, but it isn't MTBF.
=h=
That's pretty much the same arguement that I heard at the beginning of the migration from 5.25 to 3.5. There'll be early adopters, just like with every technology and, at some point, 2.5" drives will become the standard.
=h=
MTBF is not a measure of how long the drive will last, it's a measure of how often you should expect a failure, given a program of maintenance and drive replacement at the end of its useful life.
/n hours, where n is the number of drives.
The end of a drive's useful life isn't when it fails, by the way. You'll probably have to pull the manufacturer's toenails to find out what their recommendation is (although maybe the warranty period might be a clue).
So if you follow the maintenance and replacement schedule, you will have a drive failure only once very hours. If you've got a slew of drives in your server room, you should expect a failure of any single drive every
So, yes, the numbers are accurate. Got a single drive? Maintain and replace it proplerly and you'll only be unpleasantly surprised once every 160 years.
=h=
That's not exactly correct, either. MTBF is the amount of time that you can expect before the failure of a single drive, assuming that it is properly maintained and replaced when its useful service life has been reached.
MTBF assumes that you are not running the device until it just wears out. The figure assumes that you recognize that the device has a finite lifetime and that you are prepared to replace it before it reaches that lifetime.
So, if you follow that guideline, you should expect a single device failure every 1.4 million hours for a single device. From that point, your explanation is correct - what is not correct is assuming that MTBF has anything to do with an indication of when it is necessary to replace a device...MTBF assumes that you are replacing the devices before they wear out.
=h=
...the same sources are reporting the tragic death of BSD.
The sum of his income was $659,000. He paid $90,575 in federal income tax. That's 13.7%. Pretty damn close to 13%.
You've got that wrong. Under the 1909 Act, all you had to do was publish with notice to have a copyright. Most works were never registered with the copyright office until their renewal term.
Not so. That was only the first step. You also had to provide two copies of the work to the Copyright Office. Then you received a certificate entitleing you to a copyright on the work.
In fact, it's entirely possible that the copyright was voided if Guthrie didn't send his copies in - the 1909 law provided for that.
Like I said, it'll take a judge to sort this out.
=h=
I never understood why this logic has not been applied (legally at least) to Happy Birthday which was music originally published in the 1890's (as "Good Morning to All/You"), had the alternative lyrics published or alluded to a number of times from 1909 onwards, but wasn't copyrighted until 1935 by one of the original authors (who didn't even write the changed lyrics!) and is still under copyright until something like 2030, allowing the current holders to sue the Girl Guides and force restaurants to come up with those silly hand clapping songs.
The reason is because copyright laws have changed. Back in the day, you didn't automatically get a copyright by virtue of publishing - you had to apply for a copyright. The date of publication didn't have any bearing on the date of copyright.
I'm sure that in the case of Ludlow, they applied for the copyright when they purchased the rights from Guthrie (or whomever they bought them from), so they must figure that the clock started ticking at that point.
This is definitely crap for a judge to sort out.
=h=
For those of you who make fun of the Soviet system wen you probably wheren't even born then, this is a lesson: Soviets took care of their people well and their medicine was top.
You're kidding, of course. Although the USSR's health care system was universal, the quality was utterly abyssmal for the average citizen.
I was unfortunate enough to see first-hand the state of Soviet-era medical facilities and the quality of care in the mid 1980's. Many third-world countries had much better medical care than that of the "typical" Soviet hospital that we toured. And, given that this was a state-sponsored tour (as was everything that we saw), I suspect that it was something better than typical.
-h-
Actually, it's a function of the PABX system that they are using - it's not that they have an 800 number (although plenty of places with 800 numbers have PABX's), it's that they are using a PABX that ignores the flag that suppresses the caller ID info. Or maybe it can read the ANI info.
-h-
Ah yes, here it is - "Damage to the eyes comes predominantly from invisible infrared wavelengths, so the fact that you feel no discomfort while gazing at a partial eclipse does not guarantee that your eyes are safe"
Holy cow, if the IR from those floods is strong enough to damage your eyes, you'll definitely feel it somewhere else. IR causes burns to the cornea. You're thinking of UV, which can damage the retina because it is relatively high energy radiation.
The reason that an eclipse causes corneal burning is because the sun pumps out a tremendous amount of IR...it's a huge fusion furnace, after all.
-h-
It wasn't "the House of Representatives", it was Representative Duncan Hunter, a San Diego Republican who makes Rush Limbaugh look moderate...and that's coming from a registered Republican!
The amendment will never leave the House.
-h-
Work closely with your college's alumni association and with the CS department's industry liason. Both of them are excellent resources for job placement assistance.
Don't forget the power of social networking, either. I was lucky enough to get my "dream job" before I graduated (BSEE) because a friend of a friend was a manager at the company. In fact, that may be the best way to get the job, regardless of your experience.
Thirdly, consider joining the Computer Society of the IEEE and attend the functions, email with the members and even consider volunteering for some of the tasks that come with the Society. The membership dues are significantly reduced for recent graduates. Also, the IEEE's GOLD (Graduates of the Last Decade) organization can help as well.
And don't hesitate to apply for a job, even if you don't think that you meet the experience criteria. Even though many of the resumes are screened by HR and you may not get past them, many are not. Something in your resume may stick out and get you an interview - and that's what you're really after.
Good Luck!
-h-
Common sense of the common man tends to be uncommonly nonsensical.
Given that your reasoning is both simplistic and polemic and that it appears that, in the face of reason you still cling to your fallacies, I'd say that I'd be wasting my time attempting to explain the rather obvious conclusions of the creators of our republican system. You're clearly not aware of the history and reasoning behind their decision (or you choose to ignore it) or you wouldn't be clinging to the worn out arguments that just don't have any basis in fact.
Since my time is worth something, I'll spend it elsewhere, comforted by the hope that perhaps in time you'll take my advice and actually learn something about the history of the founding of our system of government instead of simply repeating nonsense that was discredited years ago.
-h-
Don't buy into the PC thought of the present. The issue of slave ownership was certainly present in the day, but there were other, more principled ideas behind both the electoral college and the Senate. Some significant issues were those of taxation and commerce and what effect states with large populations would have on those with small - and vice versa. The fact that slavery existed and was considered does not invalidate the concept. That's a red herring.
I don't claim that one adult = one vote is undue influence and neither did the creators of our system of government. They were looking at a broad picture of influence. In fact, the system that governs us (in the US) has a great many features that tend to override the one person, one vote rule. A specific example is the referendum process. A referendum, approved by the majority of the people can be declared invalid if it infringes upon the rights of a minority. Admittedly, that is somewhat tangental to the idea of the electoral college and the Senate, but the underlying principle is sound.
And, finally, the electoral college does not create a situation that allows two small states to be more influential than a large one. Neither does the Senate. The electoral college somewhat dilutes the power of large states while it somewhat increases the power of small states. And in the Senate, all states are equal. Taken as a whole, though, Congress somewhat dilutes the power of large states and somewhat increases the power of small states.
Anyway, like I said, take a civics class. Or read some scholarly books on the subject. These ideas were really quite well discussed and thought out. They didn't just pop out of nowhere.
-h-
For all of those who can't see the underlying fairness of the electoral college, I would suggest taking a civics course from your local community college. Also, you may want to ponder on just why the US Senate is organized the way that it is.
Both were set up with particular aims in mind - chief among them that states with large populations would not be able to use those large populations to unduly influence elections and legislation to the deteriment of sparsely populated states. The electoral college also dovetails quite nicely with the concept of a democratic republic.
-h-
Except Gore in the last election (Tennesee). What were we talking about again?
As the parent said: historically candidates very rarely lose their home state.
I'd say that your example fits the category of "very rarely".
-h-
I think that if you filled the box with Apoxie, a little piece of velcro would easily hold the lid closed. It would certainly solve the mechanical problems! And you're right, heat is not a problem with 9V. Even with two 9V batteries, heat isn't an issue.
Filling the tin is a great idea - maybe I'll give it a try and see what happens!
-h-
It's the tins. The electronics are plenty durable because the circuit board is very small. But the typical Panasonic volume control and the micro switch will probably break because tin doesn't provide them with enough support in a catastrophic flying event. Your choices are pretty limited for 10K stereo pots that will fit in a small tin (and not sound like fingernails scratching on a chalkboard) and the leads that were really designed for plugging into a PCB are kind of fragile. And if the lid pops open during its mid-air acrobatics and the battery starts flailing around, something will almost certainly break. That, and the tins are surprisingly non-durable when they aren't full of mints ;-)
;-)
I agree that the Hammond case isn't nearly as stylish as the Altoids or Penguin case. I guess my deal is that since I sell them, I'd just as soon not have one come back because the case wasn't strong enough to stand up. On the other hand, if somebody REALLY wants it in an Altoids tin, I'll build it that way for them...just so they know that they have to be pretty careful. I guess that's really the key to using a mint tin as a case. I always tried to be careful, but then I'd forget what was on the other end of my headphones and -zip- off it would go.
Also, you really should power the CMoys with two 9V batteries. I know that you can do it with one, but that only gives you about one volt of margin before the opamp starts to clip and if the rails become unbalanced (something that the resistor divider power supply is prone to do), then the margin can suddenly drop to nothing. The opamps that work best (OPA2132, 2134, etc.) need +/-3V just to operate. That leaves only 1.5V peak for the audio signal...OK if your cans are sensitive, but if you've got something like my Grado SR-125's, you'll be in trouble. Anyway, making a short story long, the other reason that I like the Hammond case is because it will let another battery fit in or a slightly larger PCB with a TI TLE2426 rail splitter to keep the voltage rails in balance.
That's my 2 cents on CMoy design
-h-
Well, as a blatent plug for me, you can buy any one of those players, a low cost set of quality headphones and one of these. The sound quality is almost always a product of the low budget headphones combined with low voltage (which causes clipping which causes distortion). I use a Neuros player which supports OGG, lasts for between 10 and 12 hours per charge and sounds pretty good with that CMoy amp and a ~$50 pair of headphones.
$200 for the Neuros
$50 for the 'phones (Grado SR-60)
$50 or $60 for the amp
-h-
Just be aware that if you put it in a mint tin, you'll probably break it within a couple of weeks. Those tins just aren't beefy enough. Even on your desk. I sent my Cmoy amp zipping across the floor when I turned around one day and my headphone cord yanked it. I build all of mine in a nice extruded case from Hammond.
-h-
Yeah, I'll second that. I've had one since last summer and I love it. The 2.13 firmware works fine for me and I guess I'll be upgrading to USB 2.0 now.
.oggs.
Best of all, it plays all my
-h-
Decibels are logarithmic. 6 watts gets you about 117dB. But even 110dB is loud! And the article said that at the couch, 1 watt delivered 100dB. Loud!