A lot people like to brag that their servers have three years or more of uptime. That's not hardcore, that's a server that hasn't been patched for three years!
Truly hardcore servers are those that can be patched weekly with little or no service downtime.
Amen, and why I just love ZFS (or any filesystem that supports instant snapshots). I use mirroring to cover drive failures, and I use weekly snapshots as backups. Once every three months, I offline to external disk. Minimum cost, more than reasonable protection.
The short answer is that there is no good reason why this should be the case
It doesn't turn off; it sends a break, which puts you at the ok prompt. The machine isn't running at the ok prompt; it's waiting for you to type something (like "go" to resume operation or "boot -s" to reboot to single-user mode). Some people equate this with "turning the server off" which is incorrect. Why is it useful? In case your terminal couldn't (or wouldn't) send a break, and you really REALLY needed to send a break because the server was going haywire or something, and didn't have physical access to the server. Turn terminal off, wait a bit, turn on, hit ENTER, see ok prompt, fix the machine.
What's irritating about that story is that the guy telling it is the real idiot: If you have a Sun server that reacts that way (and only a handful ever did), you put the keyswitch in the LOCKED position so that it DOES NOT HAPPEN WHEN YOU DO THAT. Or, have the decency to use a version of Solaris made after 1999 that supports LOGGING for the UFS filesystem so that you don't spend hours cleaning up the filesystem after an outage.
The author mistakenly thinks that the TPB's infrastructure must not cost a lot simply because the website is spartan. The Pirate Bay is a tracker, and the author should look into what a bittorrent tracker does, and then multiply that by the millions of people that visit TPB each hour.
That's a very easy thing to say when you are single and living alone. Get married, have children, have your wife go back to work to make ends meet... then get back to us on how eating pre-processed food or eating at different times makes the family "lazy".
4) Pop disk into PC drive, and either use a custom utility, or just use the FORMAT command specifying that only tracks out to 17 be formatted (FORMAT A:/T:17/N:9). This allows BOTH sides of the disk to be formatted up to track 17, giving you about 180K to play with.
Right, except that, if you actually read the post, you'd know that what I found was that every even track != 0 was C64 and every odd track *and* the entire second side was IBM.
I mean, come on, I posted a FAT dump as a screenshot. So no, it wasn't truly that easy (even though our definition of "easy" is a lot different than most people's). It required a little more planning, and manual patching of *both* filesystems.
Ostensibly, you would be burying things that you would want to preserve, like objects. If you're burying electronic media, you're missing the point that electronic media can be preserved much more easily by copying it to new media.
What will survive best in your dirt grave? Probably tape, followed by flash media, floppy diskettes, and finally hard drives. But the entire point is that you don't bury that stuff; you archive it by making backups (above ground).
Time capsules are a novelty. Put something in it worthy of being in a capsule, like a popular culture object (ie. toy or ipod or something).
Why would you even consider running a benchmark program you don't have source code for and cannot compile yourself?
Consistency. You mentioned using md5 hashes to tell different binaries apart, but Futuremark has to publish a single program that provides a meaningful benchmark score across all platforms, for whatever the current definition of meaningful is.
This issue is blown way out of proportion. They chose a bad way to determine CPU capabilities (CPUID instead of testing for capabilities) in 2005 and now people are surprised that three years later the brand-new CPU from neither Intel nor AMD that didn't exist at the time the benchmark was written isn't performing at full capacity? And people are surprised?
The problem with MPEG encoding and decoding is that the data itself is not well suited to multi-threaded analysis.
Spoken like someone who has never written an encoder. There are dozens of opportunities to thread, even if it's just as simple as each stage breaking a frame up into N bands, one for each core. DCT transforms, motion vector synthesis, etc. are not dependent on the entire frame being encoded as a whole (although they do need to be assembled correctly once all threads finish).
I'm surprised nobody has mentioned that there are specialized scanners for this. Google "book scanners" for a starting point. The cheaper ones have glass right up to the edge so that you can scan to the spine and always keep the spine at a 45-degree angle. The more expensive ones are cameras above the surface, and the software that comes with them compensates for the page curvature.
Yes, with both rubber-dome and buckling spring keyswitches. So far, no issues, but I only do it with keyboards I don't care about. Whenever my model Ms need service, 95% of the time I open them up and clean them properly (carefully wipe down the impact-side of the board, airblast the gunk out, etc.)
Has the possibility occured to you that you don't know how to type?
Amen, brother. When I read the post above that said "I've had my ergonomic keyboard for 12 years and it still works great, only the n and m keys are slightly faded", my initial reaction was, "Then you're clearly a crappy typist." If you've had a keyboard that gets "heavy use" for 12 years, you are either lying or you type at a glacial WPM.
I remember reading responses by kernel devs saying they would not put ZFS into the kernel, regardless of license. IIRC, it was because it violated in so many spectacular ways the concept of layering.
Yes, which is how it is able to do the amazing things that it does. Some of the stuff ZFS does -- and only ZFS does -- is because the storage management and filesystem are merged.
The people who bash ZFS haven't used it, haven't researched it, or both.
We're talking about PCs, so not AmigaOS, but rather Geoworks. It was years ahead of its time and even more ahead of Windows, but Microsoft used its monopoly to muscle it out of competition.
Hold onto the tapes. I'm assuming this is DV, yes? If DV, simply hold onto the tapes. They're less than a buck, hold 13G (1 hour of DV), and your video is already on them. You'll be able to read them for at least 10 years (they'll last for 20 but DVcams might not be around then).
The best computing chair I ever sat in was an ergonomic chair I saw at some conference in 1993 where a split keyboard was placed at the end of both armrests (ie. one half was on the left armrest, and the other half on the right). It was like being One With The Machine -- you just sit down and type.
Unfortunately, I've been unable to locate the name or manufacturer of this chair ever since. Does anyone know of anything like this?
I will tell you from recent experience that typing on one of these old beasts will slow you down immensly.
You must be a terrible typist then. I type at 90wpm and the genius of the model M is not only the audible "click" feedback, but the tactile feedback you get from the buckling spring mechanism "popping" back at you. You have two pieces of feedback that tell you, through sound and touch, that your keypress has been registered.
This not only speeds me up, but reduces pain. For a challenge at a previous job, I offered to use a regular keyboard (rubber-dome variety) for a week. At the end of the week, my hands hurt from pressing too hard (because without the unconscious feedback I got from buckling-spring mechanisms, I wasn't as sure that my keypresses were registering).
The funny slogan of Model M enthusiasts is "type hard or go home" but the ironic thing is, you type less hard with a model M.
The reason rarity != value is because, if nobody knows about it, nobody wants it. I own a fairly nice copy of Wibarm, and I believe I'm the only one left in the USA to own it. But since nobody has heard about it, and it's not part of some Infocom/Sierra/Lucasarts legacy, nobody would offer me more than $20 for it.
Condition is obviously important. Incomplete items are worth nearly nothing, and even if it's complete it should be in decent condition (ie. the box isn't crushed). If it's in mint condition (still shrinkwrapped), you are holding gold.
One exception to this is diskettes: For reasons I don't quite agree with, most collectors feel that the condition of the diskette media is not nearly as important as the other materials, mainly because most of the software has been cracked and available. I disagree, because without working originals, you can never be sure if the cracked versions are complete (and in my experience easily 15% of them are not).
The ebay market for collectible software started to dry up around 2005, but for a very long time it was a hotbed of collectible software buying and selling. You can still find some reasonable bargains (ie. an average of $20-$30 a title) but most of the time it still costs $200 for a Kilrathi Saga, or $1600 for an original Infocom Starcross Saucer.
Slashdot Mirror
A lot people like to brag that their servers have three years or more of uptime. That's not hardcore, that's a server that hasn't been patched for three years!
Truly hardcore servers are those that can be patched weekly with little or no service downtime.
Amen, and why I just love ZFS (or any filesystem that supports instant snapshots). I use mirroring to cover drive failures, and I use weekly snapshots as backups. Once every three months, I offline to external disk. Minimum cost, more than reasonable protection.
"for the next two years we...helped with spelling...opps on are part"
I can't begin to imagine what kind of spelling advice you were giving as "opps on are part".
The short answer is that there is no good reason why this should be the case
It doesn't turn off; it sends a break, which puts you at the ok prompt. The machine isn't running at the ok prompt; it's waiting for you to type something (like "go" to resume operation or "boot -s" to reboot to single-user mode). Some people equate this with "turning the server off" which is incorrect. Why is it useful? In case your terminal couldn't (or wouldn't) send a break, and you really REALLY needed to send a break because the server was going haywire or something, and didn't have physical access to the server. Turn terminal off, wait a bit, turn on, hit ENTER, see ok prompt, fix the machine.
What's irritating about that story is that the guy telling it is the real idiot: If you have a Sun server that reacts that way (and only a handful ever did), you put the keyswitch in the LOCKED position so that it DOES NOT HAPPEN WHEN YOU DO THAT. Or, have the decency to use a version of Solaris made after 1999 that supports LOGGING for the UFS filesystem so that you don't spend hours cleaning up the filesystem after an outage.
...which is a better product anyway.
The author mistakenly thinks that the TPB's infrastructure must not cost a lot simply because the website is spartan. The Pirate Bay is a tracker, and the author should look into what a bittorrent tracker does, and then multiply that by the millions of people that visit TPB each hour.
This shouldn't be surprising, or a joke -- C was intended from the very start to be a portable assembler. And it has been, wonderfully.
That's a very easy thing to say when you are single and living alone. Get married, have children, have your wife go back to work to make ends meet... then get back to us on how eating pre-processed food or eating at different times makes the family "lazy".
4) Pop disk into PC drive, and either use a custom utility, or just use the FORMAT command specifying that only tracks out to 17 be formatted (FORMAT A: /T:17 /N:9). This allows BOTH sides of the disk to be formatted up to track 17, giving you about 180K to play with.
Right, except that, if you actually read the post, you'd know that what I found was that every even track != 0 was C64 and every odd track *and* the entire second side was IBM.
I mean, come on, I posted a FAT dump as a screenshot. So no, it wasn't truly that easy (even though our definition of "easy" is a lot different than most people's). It required a little more planning, and manual patching of *both* filesystems.
Ingenious, but really not that hard at all...
Which is precisely the part that blew my mind :-)
"Show me an action that sold for more than local store retail."
Two words: Neo Geo.
Ostensibly, you would be burying things that you would want to preserve, like objects. If you're burying electronic media, you're missing the point that electronic media can be preserved much more easily by copying it to new media.
What will survive best in your dirt grave? Probably tape, followed by flash media, floppy diskettes, and finally hard drives. But the entire point is that you don't bury that stuff; you archive it by making backups (above ground).
Time capsules are a novelty. Put something in it worthy of being in a capsule, like a popular culture object (ie. toy or ipod or something).
Why would you even consider running a benchmark program you don't have source code for and cannot compile yourself?
Consistency. You mentioned using md5 hashes to tell different binaries apart, but Futuremark has to publish a single program that provides a meaningful benchmark score across all platforms, for whatever the current definition of meaningful is.
This issue is blown way out of proportion. They chose a bad way to determine CPU capabilities (CPUID instead of testing for capabilities) in 2005 and now people are surprised that three years later the brand-new CPU from neither Intel nor AMD that didn't exist at the time the benchmark was written isn't performing at full capacity? And people are surprised?
The problem with MPEG encoding and decoding is that the data itself is not well suited to multi-threaded analysis.
Spoken like someone who has never written an encoder. There are dozens of opportunities to thread, even if it's just as simple as each stage breaking a frame up into N bands, one for each core. DCT transforms, motion vector synthesis, etc. are not dependent on the entire frame being encoded as a whole (although they do need to be assembled correctly once all threads finish).
Whoops -- meant to write "90-degree angle" instead of 45...
I'm surprised nobody has mentioned that there are specialized scanners for this. Google "book scanners" for a starting point. The cheaper ones have glass right up to the edge so that you can scan to the spine and always keep the spine at a 45-degree angle. The more expensive ones are cameras above the surface, and the software that comes with them compensates for the page curvature.
Yes, with both rubber-dome and buckling spring keyswitches. So far, no issues, but I only do it with keyboards I don't care about. Whenever my model Ms need service, 95% of the time I open them up and clean them properly (carefully wipe down the impact-side of the board, airblast the gunk out, etc.)
Has the possibility occured to you that you don't know how to type?
Amen, brother. When I read the post above that said "I've had my ergonomic keyboard for 12 years and it still works great, only the n and m keys are slightly faded", my initial reaction was, "Then you're clearly a crappy typist." If you've had a keyboard that gets "heavy use" for 12 years, you are either lying or you type at a glacial WPM.
I remember reading responses by kernel devs saying they would not put ZFS into the kernel, regardless of license. IIRC, it was because it violated in so many spectacular ways the concept of layering.
Yes, which is how it is able to do the amazing things that it does. Some of the stuff ZFS does -- and only ZFS does -- is because the storage management and filesystem are merged.
The people who bash ZFS haven't used it, haven't researched it, or both.
We're talking about PCs, so not AmigaOS, but rather Geoworks. It was years ahead of its time and even more ahead of Windows, but Microsoft used its monopoly to muscle it out of competition.
Hold onto the tapes. I'm assuming this is DV, yes? If DV, simply hold onto the tapes. They're less than a buck, hold 13G (1 hour of DV), and your video is already on them. You'll be able to read them for at least 10 years (they'll last for 20 but DVcams might not be around then).
Nagel face prints. Nothing screams 1980's computer fascination like Nagel prints.
The best computing chair I ever sat in was an ergonomic chair I saw at some conference in 1993 where a split keyboard was placed at the end of both armrests (ie. one half was on the left armrest, and the other half on the right). It was like being One With The Machine -- you just sit down and type.
Unfortunately, I've been unable to locate the name or manufacturer of this chair ever since. Does anyone know of anything like this?
I will tell you from recent experience that typing on one of these old beasts will slow you down immensly.
You must be a terrible typist then. I type at 90wpm and the genius of the model M is not only the audible "click" feedback, but the tactile feedback you get from the buckling spring mechanism "popping" back at you. You have two pieces of feedback that tell you, through sound and touch, that your keypress has been registered.
This not only speeds me up, but reduces pain. For a challenge at a previous job, I offered to use a regular keyboard (rubber-dome variety) for a week. At the end of the week, my hands hurt from pressing too hard (because without the unconscious feedback I got from buckling-spring mechanisms, I wasn't as sure that my keypresses were registering).
The funny slogan of Model M enthusiasts is "type hard or go home" but the ironic thing is, you type less hard with a model M.
As someone who runs a software collector's mailing list and a co-author of a collectible software grading scale, I think I'm qualified to report: It depends. The collectible value of software is pretty much the same as any other collectible:
The reason rarity != value is because, if nobody knows about it, nobody wants it. I own a fairly nice copy of Wibarm, and I believe I'm the only one left in the USA to own it. But since nobody has heard about it, and it's not part of some Infocom/Sierra/Lucasarts legacy, nobody would offer me more than $20 for it.
Condition is obviously important. Incomplete items are worth nearly nothing, and even if it's complete it should be in decent condition (ie. the box isn't crushed). If it's in mint condition (still shrinkwrapped), you are holding gold.
One exception to this is diskettes: For reasons I don't quite agree with, most collectors feel that the condition of the diskette media is not nearly as important as the other materials, mainly because most of the software has been cracked and available. I disagree, because without working originals, you can never be sure if the cracked versions are complete (and in my experience easily 15% of them are not).
The ebay market for collectible software started to dry up around 2005, but for a very long time it was a hotbed of collectible software buying and selling. You can still find some reasonable bargains (ie. an average of $20-$30 a title) but most of the time it still costs $200 for a Kilrathi Saga, or $1600 for an original Infocom Starcross Saucer.