Yeah, Slowlaris does suck, but it has its good points, and I still think it's on top of the rest in the commercial market.
Now AIX I have a bone to pick with. I've spent a lot of time with AIX in the 4.x versions, on a large number of RS6ks (SP/2 frames actually). I think AIX is a horrible implementation of unix, totally departing from the ideals of unix. I could go on for days on end (and I have, causing some of my co-workers to die of boredom), but to state a few simpler ones:
Commandline text tools have arbitrary limits. AIX's built-in text tools (grep, sed, awk, etc) all have arbitrary limitations, like "no lines longer than 2048 chars". This flies in the face of the unix philosophy. Normally I wouldn't have noticed this as I don't regularly deal with lines that long. However, I regularly encountered such lines in the cluttered command output of AIX's configuration tool "smit", which is a whole other rant on it's own.
System config data that should be in human (and unix tool) parseable config files in etc, is instead in a binary object oriented database. Further, the tools to manipulate the database are poorly documented if at all. Forget grepping or vi-ing your config. It's even worse specifically on AIX on SP/2s, all the SP/2 stuff is similarly in binary object databases.
Their standard linker is really nonstandard and jacked, making configure and make scripts have many headaches.
I have no opinion on MacOS X, except that I don't except it to be much of a datacenter contender.
I'm gonna go out on a limb here and be very opinionated and biased - but hey, it is an opinion question.
I think Solaris will be the only commercial unix to survive more than a few short years from now (here I define survive as being useful, current and having a decent market share and attracting new users - technically Netware still survives today, but it hardly meets my definition, it's just slowly dying off as legacy users switch away).
That being said, I don't think Solaris has much of a future either if they don't change their ways soon. They've already been trounced for web and application servers by thin cheap commodity linux stuff. They're only real foothold at the moment is large databases (think E10K-E15K class machines, 20TB databases, etc), and highly available databases at that.
They're currently in the process of losing this to Oracle9's RAC linux clusters, which blow Sun away in terms of bang for the buck, and can scale just as well in overall bang and reliability.
Now - all of the above is from the perspective of someone who believes only in the technical data and is willing to be on the slightly bleeding edge. When you factor in typical corporate environments/attitudes and whatnot, the picture slows down and pushes off into the future a bit further for the mainstream unix consumers.
None the less, I think my assessment here will prove to be accurate over time, with all commercial unices eventually falling to Linux, with Sun being the longest and strongest holdout.
Of course, in this long term sense, I really mean "the idea of Linux" when I say Linux. Linux could be supplanted by some other GPL (or GPL-ish) kernel down the line that re-uses a lot of the drivers and OS componentry from Linux and not really change my point.
And for one final caveat, since I can't really see the future, I don't know for sure that Sun won't manage to correct their currently tragic course and get back in shape and survive. If they were smart, they'd stop trying to marginalize linux as a thin edge device, and start contributing large amounts of their man-hours to perfecting linux on the sparc64 platform as a way of protecting their hardware and support businesses against the demise of Solaris.
Yeah, but in many branches of science, our theories have settled down to some degree, and change as a course of slow refinements (Newtonian physics still works great for most human-sized macro events, for instace - everything afterward merely refines it for small/huge and special cases).
IMHO, theories of the universe are still in their infancy, and every new theory seems to be at complete odds with previous theories, we're still in very much of a guessing game stage without much solidity.
In a few years a different theory will emerge, ad infinitum. Astrophysics is a black art at this stage. It's important that we pursue it, but we all know that the "current theories" are just working models to help us along, they're almost always disproved and new models are born. It will be along time before we settle into a model that we can believe with some degree of certainty.
I agree, let them choose their own course. But I don't think he's trying to force hsi kids into being math/sci geeks, he just wants them to enjoy the subject and be decent at it. US public schools do a fairly poor job of educating kids on math/sci, especially in the early grades, which makes a big difference down the road. It's important that he as a parent take over where the school leaves off to ensure that his children are equipped to handle the future, even if they do turn out to be liberal arts types. At least this way they know when they're getting screwed on a book contract, and can tell when a demonstration on a TV ad is complete BS from a chemistry point of view.
Finding the image edges from her bulk scans is one of the more trivial operations you can do on an image. Grab a handy-dandy image library for your chosen format (pnglib, jpeg, whatever) and write a couple pages of code and you're done. GPL it and help others with the same problem.
I agree, but he may be onto something in that his solution may be easier than most to cheaply embed in low end computer hardware.
Although I seem to remember hearing a long time ago how someone had built a circuit (which I presume could be put easily on an expansion card) that obtained entropy from the small (mostly temperature-fluctuation-induced) changes in capacitance and resistance on some standard crappy-grade capacitors and resistors in a simple circuit. While your case's overall temp might be a bit predictable to a tenth of a degree, I think this was supposed to be sensitive to the very small seemingly random fluctuations of a much smaller degree.
Footnote: If you do ever suffer failed drives from this scenario (drives have been running for extended periods, like months to years, you power them off and allow them to cool, then try to use them again), there's a trick to reviving them that works a lot of the time:
Hold the drive in one hand, with the flat metal surface which covers the disc platters facing up, and give it a good solid whack with the handle of a screwdriver. Another variant is to drop it on the floor from a height of 1-2 feet, again hitting the metal cover surface over the platters, not the circuit board side or the edges.
I've seen many an IBM technician use this technique on service calls succesfully, and I've done it myself a few times, it does work. The supposed theory is that the heads get stuck in the parked position, and you're jarring them loose again.
I've twice taken breaks from home connectivity, about 6-9 months each time, roughly same time frame as the above poster (net since around 94, BBSes before that). I didn't do any excersizing or woods-walking, but I do get a lot more productive on my home computers. I would work on real code a lot more often, instead of getting sucked into mindless browsing and slashdotting. If I needed some faqs, manuals, source code, etc from the web, I'd just sneaker-net home whatever I needed on an LS-120 floppy from work, or from the free net access at the community college down the street from my apartment.
Yeah but how much dev $$ could it take to re-implement 8-bit nintendo games? The gameplay and graphics can be gratuitously ripped, so there's no design costs. The code is so simple and small by modern standards it couldn't be all that much $$.
Again, those (heat, power) are external events though, not random failure, as in what the MTBF is indicating. Controlling the external factors is the sysadmins' and datacenter ops peoples' job. As for wear and tear, no it won't cause simul-failure. As for hardware bugs - if they're in the drive's firmware it's remotely possible, but I don't know... never heard of it happening.
Yeah - lesson #1, never power off your disk arrays until you're about to throw them away. Should've brought in a generator. Making the mistake of letting your drives go cold is an external circumstance that will cause multiple failure.
You just have to learn how to use a search engine. Most "Ask Slashdot" questions are so simple that just typing in 1-2 keywords gets what you need - if your query is more complicated, chances are Ask Google still works, you just need a more complicated query to extract the right data.
And for those that have said Google isn't good enough because you want personal opinion and not a howto.... you're likely to find what you're looking for on Google if you would just look. Check out that guy's website with his grueling story of how he did it, it took him a week and a lot of duct tape. Read the damn howto as well, I'm sure it's based on someone's personal experience.
Come to Ask Slashdot when you question actually borders on technical opinion instead. But even then, don't ask people to do you damn job for you. So many of the non "Ask Google" questions are obviously just that - usually what tips these off is that they're too specific to a special situation the asker is dealing with. Ask the broad general question instead.
Simultaneous disk failure is about as rare as winning the lotto while simultaneously having you and your friend on the other side of the planet get struck by lightning - unless there's a common larger problem (e.g. power surge to both drives or something).
As a practical example, 4 or 5 years ago I had large amount of disks attached to some large oracle servers, roughly on the order of 600 or so hard drives in several arrays taking up several racks, all the same manufacturer/model, with a handful of groupings of revision/lot/date.
This set of disks was seeing fairly constant and heavy activity for a few years while I was there. As you can imagine, with 600 disks and the usual MTBF numbers, we quite regularly had disk failure. We kept a few spares onsite and replaced them as they failed, then exchanged the dead drive for a new spare. As I roughly remember it, we probably averaged about one disk failure every 2-3 weeks. Two, perhaps three times, we had a double disk failure during a 24 hour period - but they were never close enough that we didn't have plenty of time to replace the first (and in any case, odds are slim that two failed out of 600 would happen to affect the same data).
Of course, another point back at the original guy with the failed disks - don't use raid 5, chunk out some more money (disks are cheap) and do proper mirroring - and if you stripe use 1+0, not 0+1.
They could be using a hybrid of the two techniques:
Perhaps they've stored a whole library of generic graphics/sound/"ai" routines on the dohickey that given good common coverage to the legacy games, and the cards just store sprite and gameplay/flow data in a highly compressed format.
In the old 8-bit nintendo (probably other and later consoles as well) cartridge programmers implemented bank switching to put more data in the cartiridge than the architecture was really designed to handle. They are known as "mappers", and it's what you hear about when you read about NES emulators and whatnot and what "mappers" they support - they're referring to memory addressing schemes used by games that couldn't fit normally.
I disagree that gaming is the sole pusher. I think Porn has always been a big technology factor, and continues to be so, at least on level footing with games. Sad thing is that most porn needs are met by even older machines than current games. Porn hasn't been keeping up the tech pace the past few years, instead sticking to low-res video and whatnot. When they get their asses in gear and starting releasing photorealistic 3-d porn worlds, computer hardware will start jumping again.
You're obviously out of your league if you think writing a regression suite for an html user interface is really difficult or money-sucking, and you're a clueless consumer of bullshit if you actually go out and buy such products from other software companies.
The only rollling you should do is another joint to take your mind off of the deadbeat life you've had since your overpaid useless position was eliminated in the dot-bomb blowout.
I don't ever go to XYZ.slashdot.org, I think that whole naming system is a bunch of crap. Without looking at the URL, there was no way to know - for the most part I know an article is in Ask Slashdot beacuse the article usually says so somewhere.
If you want the auto manufacturers to play - you need to get more awareness among the public of your research, and then perhaps get US sciency TV channels like discovery to cover it, etc...
On another note - the amaccs area of the web site doesn't appear to be linked from the main page. The direct link your provided works - but if I go back to the company main page, there's no way to find it again.
Contrary to the slashdot post, you only need to be up to 0.9.6e to be safe. If you happen to just now be upgrading past this bug, 0.9.6g is even better, but if you're already running "e" you are safe. The article kinda alarmed me at first when I saw the "g", thinking there was a new exploit in "e" and I needed to upgrade again.
This would have made the first decent Ask Slashdot in a while, since it actually seeks a broad range of opinion and experiential knowledge from the community that probably can't be found in 5 seconds on Google. Too bad it wasn't in the Ask Slashdot section though:)
Actually it would probably be the best solution. Train a lab maze rat to climp up a similar shaft in a lab, and strap leightweight scientific equipment on his back.
Slashdot Mirror
Fight! Fight!
Yeah, Slowlaris does suck, but it has its good points, and I still think it's on top of the rest in the commercial market.
Now AIX I have a bone to pick with. I've spent a lot of time with AIX in the 4.x versions, on a large number of RS6ks (SP/2 frames actually). I think AIX is a horrible implementation of unix, totally departing from the ideals of unix. I could go on for days on end (and I have, causing some of my co-workers to die of boredom), but to state a few simpler ones:
Commandline text tools have arbitrary limits. AIX's built-in text tools (grep, sed, awk, etc) all have arbitrary limitations, like "no lines longer than 2048 chars". This flies in the face of the unix philosophy. Normally I wouldn't have noticed this as I don't regularly deal with lines that long. However, I regularly encountered such lines in the cluttered command output of AIX's configuration tool "smit", which is a whole other rant on it's own.
System config data that should be in human (and unix tool) parseable config files in etc, is instead in a binary object oriented database. Further, the tools to manipulate the database are poorly documented if at all. Forget grepping or vi-ing your config. It's even worse specifically on AIX on SP/2s, all the SP/2 stuff is similarly in binary object databases.
Their standard linker is really nonstandard and jacked, making configure and make scripts have many headaches.
I have no opinion on MacOS X, except that I don't except it to be much of a datacenter contender.
I'm gonna go out on a limb here and be very opinionated and biased - but hey, it is an opinion question.
I think Solaris will be the only commercial unix to survive more than a few short years from now (here I define survive as being useful, current and having a decent market share and attracting new users - technically Netware still survives today, but it hardly meets my definition, it's just slowly dying off as legacy users switch away).
That being said, I don't think Solaris has much of a future either if they don't change their ways soon. They've already been trounced for web and application servers by thin cheap commodity linux stuff. They're only real foothold at the moment is large databases (think E10K-E15K class machines, 20TB databases, etc), and highly available databases at that.
They're currently in the process of losing this to Oracle9's RAC linux clusters, which blow Sun away in terms of bang for the buck, and can scale just as well in overall bang and reliability.
Now - all of the above is from the perspective of someone who believes only in the technical data and is willing to be on the slightly bleeding edge. When you factor in typical corporate environments/attitudes and whatnot, the picture slows down and pushes off into the future a bit further for the mainstream unix consumers.
None the less, I think my assessment here will prove to be accurate over time, with all commercial unices eventually falling to Linux, with Sun being the longest and strongest holdout.
Of course, in this long term sense, I really mean "the idea of Linux" when I say Linux. Linux could be supplanted by some other GPL (or GPL-ish) kernel down the line that re-uses a lot of the drivers and OS componentry from Linux and not really change my point.
And for one final caveat, since I can't really see the future, I don't know for sure that Sun won't manage to correct their currently tragic course and get back in shape and survive. If they were smart, they'd stop trying to marginalize linux as a thin edge device, and start contributing large amounts of their man-hours to perfecting linux on the sparc64 platform as a way of protecting their hardware and support businesses against the demise of Solaris.
Yeah, but in many branches of science, our theories have settled down to some degree, and change as a course of slow refinements (Newtonian physics still works great for most human-sized macro events, for instace - everything afterward merely refines it for small/huge and special cases).
IMHO, theories of the universe are still in their infancy, and every new theory seems to be at complete odds with previous theories, we're still in very much of a guessing game stage without much solidity.
In a few years a different theory will emerge, ad infinitum. Astrophysics is a black art at this stage. It's important that we pursue it, but we all know that the "current theories" are just working models to help us along, they're almost always disproved and new models are born. It will be along time before we settle into a model that we can believe with some degree of certainty.
Yes, I suggested he fix his own problem. Consider it my small protest against Ask Slashdot. This is a great do it yourself job.
I agree, let them choose their own course. But I don't think he's trying to force hsi kids into being math/sci geeks, he just wants them to enjoy the subject and be decent at it. US public schools do a fairly poor job of educating kids on math/sci, especially in the early grades, which makes a big difference down the road. It's important that he as a parent take over where the school leaves off to ensure that his children are equipped to handle the future, even if they do turn out to be liberal arts types. At least this way they know when they're getting screwed on a book contract, and can tell when a demonstration on a TV ad is complete BS from a chemistry point of view.
Finding the image edges from her bulk scans is one of the more trivial operations you can do on an image. Grab a handy-dandy image library for your chosen format (pnglib, jpeg, whatever) and write a couple pages of code and you're done. GPL it and help others with the same problem.
I vote for intelligent design
I agree, but he may be onto something in that his solution may be easier than most to cheaply embed in low end computer hardware.
Although I seem to remember hearing a long time ago how someone had built a circuit (which I presume could be put easily on an expansion card) that obtained entropy from the small (mostly temperature-fluctuation-induced) changes in capacitance and resistance on some standard crappy-grade capacitors and resistors in a simple circuit. While your case's overall temp might be a bit predictable to a tenth of a degree, I think this was supposed to be sensitive to the very small seemingly random fluctuations of a much smaller degree.
Footnote: If you do ever suffer failed drives from this scenario (drives have been running for extended periods, like months to years, you power them off and allow them to cool, then try to use them again), there's a trick to reviving them that works a lot of the time:
Hold the drive in one hand, with the flat metal surface which covers the disc platters facing up, and give it a good solid whack with the handle of a screwdriver. Another variant is to drop it on the floor from a height of 1-2 feet, again hitting the metal cover surface over the platters, not the circuit board side or the edges.
I've seen many an IBM technician use this technique on service calls succesfully, and I've done it myself a few times, it does work. The supposed theory is that the heads get stuck in the parked position, and you're jarring them loose again.
I've twice taken breaks from home connectivity, about 6-9 months each time, roughly same time frame as the above poster (net since around 94, BBSes before that). I didn't do any excersizing or woods-walking, but I do get a lot more productive on my home computers. I would work on real code a lot more often, instead of getting sucked into mindless browsing and slashdotting. If I needed some faqs, manuals, source code, etc from the web, I'd just sneaker-net home whatever I needed on an LS-120 floppy from work, or from the free net access at the community college down the street from my apartment.
Yeah but how much dev $$ could it take to re-implement 8-bit nintendo games? The gameplay and graphics can be gratuitously ripped, so there's no design costs. The code is so simple and small by modern standards it couldn't be all that much $$.
Again, those (heat, power) are external events though, not random failure, as in what the MTBF is indicating. Controlling the external factors is the sysadmins' and datacenter ops peoples' job. As for wear and tear, no it won't cause simul-failure. As for hardware bugs - if they're in the drive's firmware it's remotely possible, but I don't know... never heard of it happening.
Yeah - lesson #1, never power off your disk arrays until you're about to throw them away. Should've brought in a generator. Making the mistake of letting your drives go cold is an external circumstance that will cause multiple failure.
You just have to learn how to use a search engine. Most "Ask Slashdot" questions are so simple that just typing in 1-2 keywords gets what you need - if your query is more complicated, chances are Ask Google still works, you just need a more complicated query to extract the right data.
And for those that have said Google isn't good enough because you want personal opinion and not a howto.... you're likely to find what you're looking for on Google if you would just look. Check out that guy's website with his grueling story of how he did it, it took him a week and a lot of duct tape. Read the damn howto as well, I'm sure it's based on someone's personal experience.
Come to Ask Slashdot when you question actually borders on technical opinion instead. But even then, don't ask people to do you damn job for you. So many of the non "Ask Google" questions are obviously just that - usually what tips these off is that they're too specific to a special situation the asker is dealing with. Ask the broad general question instead.
Simultaneous disk failure is about as rare as winning the lotto while simultaneously having you and your friend on the other side of the planet get struck by lightning - unless there's a common larger problem (e.g. power surge to both drives or something).
As a practical example, 4 or 5 years ago I had large amount of disks attached to some large oracle servers, roughly on the order of 600 or so hard drives in several arrays taking up several racks, all the same manufacturer/model, with a handful of groupings of revision/lot/date.
This set of disks was seeing fairly constant and heavy activity for a few years while I was there. As you can imagine, with 600 disks and the usual MTBF numbers, we quite regularly had disk failure. We kept a few spares onsite and replaced them as they failed, then exchanged the dead drive for a new spare. As I roughly remember it, we probably averaged about one disk failure every 2-3 weeks. Two, perhaps three times, we had a double disk failure during a 24 hour period - but they were never close enough that we didn't have plenty of time to replace the first (and in any case, odds are slim that two failed out of 600 would happen to affect the same data).
Of course, another point back at the original guy with the failed disks - don't use raid 5, chunk out some more money (disks are cheap) and do proper mirroring - and if you stripe use 1+0, not 0+1.
They could be using a hybrid of the two techniques:
Perhaps they've stored a whole library of generic graphics/sound/"ai" routines on the dohickey that given good common coverage to the legacy games, and the cards just store sprite and gameplay/flow data in a highly compressed format.
In the old 8-bit nintendo (probably other and later consoles as well) cartridge programmers implemented bank switching to put more data in the cartiridge than the architecture was really designed to handle. They are known as "mappers", and it's what you hear about when you read about NES emulators and whatnot and what "mappers" they support - they're referring to memory addressing schemes used by games that couldn't fit normally.
I disagree that gaming is the sole pusher. I think Porn has always been a big technology factor, and continues to be so, at least on level footing with games. Sad thing is that most porn needs are met by even older machines than current games. Porn hasn't been keeping up the tech pace the past few years, instead sticking to low-res video and whatnot. When they get their asses in gear and starting releasing photorealistic 3-d porn worlds, computer hardware will start jumping again.
You're obviously out of your league if you think writing a regression suite for an html user interface is really difficult or money-sucking, and you're a clueless consumer of bullshit if you actually go out and buy such products from other software companies.
The only rollling you should do is another joint to take your mind off of the deadbeat life you've had since your overpaid useless position was eliminated in the dot-bomb blowout.
I don't ever go to XYZ.slashdot.org, I think that whole naming system is a bunch of crap. Without looking at the URL, there was no way to know - for the most part I know an article is in Ask Slashdot beacuse the article usually says so somewhere.
If you want the auto manufacturers to play - you need to get more awareness among the public of your research, and then perhaps get US sciency TV channels like discovery to cover it, etc...
On another note - the amaccs area of the web site doesn't appear to be linked from the main page. The direct link your provided works - but if I go back to the company main page, there's no way to find it again.
Contrary to the slashdot post, you only need to be up to 0.9.6e to be safe. If you happen to just now be upgrading past this bug, 0.9.6g is even better, but if you're already running "e" you are safe. The article kinda alarmed me at first when I saw the "g", thinking there was a new exploit in "e" and I needed to upgrade again.
This would have made the first decent Ask Slashdot in a while, since it actually seeks a broad range of opinion and experiential knowledge from the community that probably can't be found in 5 seconds on Google. Too bad it wasn't in the Ask Slashdot section though
Actually it would probably be the best solution. Train a lab maze rat to climp up a similar shaft in a lab, and strap leightweight scientific equipment on his back.