Obviously, this is a great achievement deserving of the/. homepage...
However, I'm more interested in hearing about how they are going to process/archive/use that much data!
I'll be honest and say that I'd never heard (or at least remembered) anything about the LSST, so I just did a brief lookover of their site and it seems like a ridiculously cool project.
LSST will rapidly scan the sky, charting objects that change or move
That means it will have to store multiple versions (history) to be able to do trend analysis. So at multiple TB's of data, how exactly are they planning on processing/analyzing it?!?
As is traditional in the US for many large ground-based telescopes, the LSST is a public-private project. Private support leverages even larger federal support. This traditionally has been true even for facilities where the data was not public immediately. LSST breaks with that tradition in that the data and data products from LSST are immediately public, without a proprietary time period. Thus, private funding for LSST supports open access.
It also mentioned that they are hoping that individuals (and groups) do interesting things with the data...but seriously, I was talking with my brother the other day about how cheap 1TB disks were now. But even if they were only $100 a piece, you'd still be dropping some major $$$ just to be able to begin to do anything (comprehensive) with it.
You are missing the point. I had a whole post (with numbers/calculations) done, but it all boils down to this. If you are using your laptop as your main storage point (probably a bad idea), then yes 16GB is probably not enough. But if you are smart about it, then you can easily make it work. Worst case scenario for your CD's (320kbps) would come out to over a week of continuous music. If you are one of those "I need all of my songs with me at all times" people, then again 16GB may not work for you. But I go through phases and want to hear different styles at different times, so I can easily swap out what I want to hear from my main storage, out of my +20GB worth of music I probably only list to maybe a GB at a time, and probably 60-70% is stuff that I will never listen to again.
You can take that rough principle above and apply it to all of your other examples as well. You can easily use more than 16GB worth of storage. But with just a tiny bit of effort you can also easily live happily with less, and probably much less. Does that mean I'm jumping on the SSD bandwagon at this point, nope...I'll wait for it to be cost effective for me and my usage habits, and I don't need a/. article to let me know when that time will be.
Isn't the Library of Congress continually growing? If so, doesn't that need to be a dynamic algorithm to adjust for its rate of growth? I couldn't find any documentation or any historical data, but I would think its out there somewhere...then we can start working on this algorithm.
If you don't like regular expressions, that's your choice, but it really has nothing to do with Perl, and you should try to understand them, as they are very useful, even outside programming.
Completely agree. However, for some reason when something comes up that I'm going to need to use regex's with I go to perl first. If it is just something procedural (if this, do that) then I'll typically use something else, especially if it will have to be maintained by someone else later on.
I can't explain exactly why that is the case, but since (I)we tend to use perl for its regex capabilities (because its fantastic) it probably gets associated with them and hence regex's (percieved) complexity especially to people who are unfamiliar with them. Heck, I'm familiar with them but my no means an expert and that "example" looks a little daunting until you start breaking it down piece by piece.
As much as I love perl, I personally feel as though if you put a well written and commented perl script beside a python or php (I actually use php from the command line every now and then too), that the perl just looks more complex. And to some (management) its more about perception than performance...hence the distaste for perl.
Yes, that's a bug...and one I recently encountered. But the fix is easy (but admittedly not obvious). You just kill the processes that hang and then reinstall the locales package once you restart. I fully agree that manually killing processes is not something you want your average user to have to do, but the workarounds are out there and (for the most part) clearly documented in those links you speak of.
I think this discussion started from someone wondering if Dell will continue to support this laptop several years down the road. To me it seems that as hardware matures its support just gets better. Just think of how many posts you hear about people putting xubuntu on their "old P2 sitting around collecting dust" and it "just works" because that hardware is well understood. So as long as Dell provides enough information about the hardware at the onset, then it will be supported well at first and as bugs/issues arise they will be incorporated into the mainstream codebases.
I'm in the process of doing the same. However, it may be short lived. She was having issues with her USB controller under XP, so I told her to give Ubuntu a try to see if it was a hardware or software issue. Its looking like its a hardware issue, but I just got her setup to dual-boot and haven't had a chance to really see what is going on. If she does buy something new, I doubt she'll stick with it "just because."
Not that it's bloaty for a non-geek, who's probably used to winXP
I think it is about "just right" for the non-geek but (as always) there is always room for improvement. If she does make "the switch" I'm sure that we'll end up using the system completely different and both end up learning more as a result. Example: I'm pretty sure it can be done via the GUI, but the only way I know how to find my IP address is via ifconfig...so after I was talking to her about how Ubuntu is so easy is to use within the first 5 minutes of her using the liveCD I had to tell her to break out the terminal. Like I said, there is probably a way to do it with the GUI that makes it look nice and pretty, but I don't know how to. Either way, I was able to login with Remote Desktop and get he at least up and running. This weekend will be the big challenge...wish me luck!
Something about the article itself. There is that saying (you don't have to agree, but still): "Ubuntu on the desktop, Debian on the server" and the summary talks mostly about KDE4 and games?!? This is just my personal opinion but as a (somewhat) Debian sysadmin, I'd really prefer that they focus on keeping things fast and stable instead of trying to throw in extra eye-candy, because the only Debian installs I've ever done were all going to be headless...
If you spent $1.99 for a ring tone, you aren't going to call for tech support.
Haha, I think I've just found a new pastime:
Tech Support: Hello, how can I help you. Me: I'm having difficulty with my ring tone. Tech Support: What seems to be the problem? Me: Well I installed it and now its playing Fergie. Tech Support: Umm, isn't that what you purchased? Me: Yes, but I want it to play a different song now, so how can I file a bug report? Tech Support: Wait, what? Me: I tried to download some patches and recompile, but it is not working. Do you have a wiki? Tech Support: What patches, what are you talking about and no we don't have a wiki. Me: Can you hold on a sec, I need to speak with my manager. [sit phone down and wait several minutes]. Me: Sorry about that, where were we. How do I configure my ringtone to play Miley Cryus? Tech Support: You have to buy that ringtone. Me: No, I've already bought one I just want to change it now. Tech Support: Umm, I don't think you understand how this works. Me: This ringtone sucks! I want my money back! [click]
Really? You mean to tell me that you can actually produce a superior product in micro quantities to a highly sophisticated and (almost) fully automated continuous brewing process that can also buy their raw resources in bulk quantities?
You might be able to produce a similar product at roughly the same cost that ends up costing almost the same amount to you the consumer, but you are negating both their (and the retailer's) markup as well as ignoring your time as a monetary resource.
FYI, I know exactly what you meant but that particular statement is either wrong or you've got a multi-billion dollar process at your disposal...either way, I'd love to try your brew. Cheers.
but OTOH Atom is quite a bit faster than even the fastest Atom.
In that case I'll take the fastest Atom, which is faster than the fastest Atom, which is faster than the fastest Atom....
So long as the power consumption holds constant, after a few iterations this thing should be able to run circles around even the BlueGene/L...all in a sub-compact notebook/UMPC!
There are only 'consequences' of actions if you do something wrong or get caught. Kids have to make mistakes and take chances in order to grow. If you parents are so over protective, how are you going to learn....and being young is the time to be a little reckless and have fun. You get to be 'resposible' and adult acting soon enough...
That is truly great advice, but can be misused in the hands of idiot parents who substitute "Kids have to make mistakes" with "Kids should have no rules or discipline." Part of being a (good) parent is setting rules and limits, then justly punishing when their children do (and they always will) break the rules. What makes it difficult is that no two children are the same, so there isn't a one-size-fits-all crime/punishment rulebook to follow.
Great parents respect their kids, but they also "instill" respect in their kids for themselves and other adults. Bottom line, having/raising kids is a HUGE responsibility, and it should be treated as such...a HUGE responsibility.
It's an arms race, and this system won't work forever, but it's fairly easy to implement and fairly difficult to overcome.
Not really, its all about scale. That system wouldn't last more than just a few seconds if a full "attack" were performed by a large botnet. The number of permutations is relatively finite, therefore with a large number of computers trying to "solve" the problem, once the correct answers were "cracked" then they could be shared and eventually the bots either know all of the answers, or you removed *all* of the questions from the list.
I'm not saying this is an ineffective system for small/medium sites, but it wouldn't cut it for really large sites.
This got modded funny, but its actually insightful. You'd be surprised...or not, guess it just depends on how tech savvy the people you know are (everything is relative).
I think that its short sighted to focus only on ninja and laser integration. We need an open weapon framework. It should be generic enough to work with the following: Units
Ninjas (obviously...reference design)
Pirates
Playboy models
Sharks
Bill Gates (he's retired, what else is he going to do)
(War) Kittens
Weapons
Lasers
Laser-television-beer coolers
nun chucks
rubber chickens
iPhones (be creative with the usage)
Frying Pans
beer
I can hear it now..."ZOMG, Playboy Bunnies with frickin beer"...wait
I honestly don't think that it has anything to do with GPS. I mean, have you seen the size of trucks these days? Its probably a side effect of the hormone filled meats they ate when they were just little personal transports. Also, is it just me or are the new rigs getting crazier and crazier with their axles. Back in my day, 4 was enough for anything...now you're seeing all of these high fangled 6'ers and 8'ers.
unless a computer science graduate degree works WAY differently
Yes and no. To some extent, CS (like mathematics) can be considered an absolute science. You either have a correct or incorrect answer. Yes, you still go through intermediate steps to arrive at the final solution which can be achieved through more than one method/algorithm. Yes, there are also other "human factors" like code style, comments, etc. Yes, a quick quizzing about those underlying principles/step can probably identify cheaters.
However, I think that the article is talking about undergraduate level courses which (in general) are graded less stringently that graduate and above assignments like what you are talking about. I went to a (well-respected) private university, and while you could probably get away with this for some aspects of CS101, the finals were actually written. You had to analyze code, and determine the output. As the courses got more advanced, so did the student/professor interaction. All of this makes me (and hopefully my employers) appreciate the fact that my degree is actually worth more than all of these "pay $X and attend for 4 years" knockoffs.
This is like the "game" between drm and hackers. They can put up roadblocks but they will eventually be "cracked." What is stopping the students from outsourcing the original papers and then outsourcing (locally) the "handwriting?" Furthermore they could just using some of the really great "handwriting fonts."
The problem isn't stopping the act of cheating, its stopping the underlying desire/incentive to cheat. If all it takes to get a degree is to have papers turned in on time for 4 years, then you get magically get paid more because you can put a piece of paper on your wall why not outsource? Your increased income will surely offset the outsourcing costs. But if you are in a job/industry where you actually have to use/perform using your "learned" skillsets *OR YOU GET FIRED* then I think the incentive to actually learn/develop will be maintained.
Slashdot Mirror
The 30 terabytes of data obtained each night
Obviously, this is a great achievement deserving of the /. homepage...
However, I'm more interested in hearing about how they are going to process/archive/use that much data!
I'll be honest and say that I'd never heard (or at least remembered) anything about the LSST, so I just did a brief lookover of their site and it seems like a ridiculously cool project.
LSST will rapidly scan the sky, charting objects that change or move
That means it will have to store multiple versions (history) to be able to do trend analysis. So at multiple TB's of data, how exactly are they planning on processing/analyzing it?!?
As is traditional in the US for many large ground-based telescopes, the LSST is a public-private project. Private support leverages even larger federal support. This traditionally has been true even for facilities where the data was not public immediately. LSST breaks with that tradition in that the data and data products from LSST are immediately public, without a proprietary time period. Thus, private funding for LSST supports open access.
It also mentioned that they are hoping that individuals (and groups) do interesting things with the data...but seriously, I was talking with my brother the other day about how cheap 1TB disks were now. But even if they were only $100 a piece, you'd still be dropping some major $$$ just to be able to begin to do anything (comprehensive) with it.
I would think Google would be open to all of those options except #1...
You are missing the point. I had a whole post (with numbers/calculations) done, but it all boils down to this. If you are using your laptop as your main storage point (probably a bad idea), then yes 16GB is probably not enough. But if you are smart about it, then you can easily make it work. Worst case scenario for your CD's (320kbps) would come out to over a week of continuous music. If you are one of those "I need all of my songs with me at all times" people, then again 16GB may not work for you. But I go through phases and want to hear different styles at different times, so I can easily swap out what I want to hear from my main storage, out of my +20GB worth of music I probably only list to maybe a GB at a time, and probably 60-70% is stuff that I will never listen to again. /. article to let me know when that time will be.
You can take that rough principle above and apply it to all of your other examples as well. You can easily use more than 16GB worth of storage. But with just a tiny bit of effort you can also easily live happily with less, and probably much less. Does that mean I'm jumping on the SSD bandwagon at this point, nope...I'll wait for it to be cost effective for me and my usage habits, and I don't need a
Isn't the Library of Congress continually growing? If so, doesn't that need to be a dynamic algorithm to adjust for its rate of growth? I couldn't find any documentation or any historical data, but I would think its out there somewhere...then we can start working on this algorithm.
However, their flexibility will quickly decline, so some of the allure will be lost.
If you don't like regular expressions, that's your choice, but it really has nothing to do with Perl, and you should try to understand them, as they are very useful, even outside programming.
Completely agree. However, for some reason when something comes up that I'm going to need to use regex's with I go to perl first. If it is just something procedural (if this, do that) then I'll typically use something else, especially if it will have to be maintained by someone else later on.
I can't explain exactly why that is the case, but since (I)we tend to use perl for its regex capabilities (because its fantastic) it probably gets associated with them and hence regex's (percieved) complexity especially to people who are unfamiliar with them. Heck, I'm familiar with them but my no means an expert and that "example" looks a little daunting until you start breaking it down piece by piece.
As much as I love perl, I personally feel as though if you put a well written and commented perl script beside a python or php (I actually use php from the command line every now and then too), that the perl just looks more complex. And to some (management) its more about perception than performance...hence the distaste for perl.
Yes, that's a bug...and one I recently encountered. But the fix is easy (but admittedly not obvious). You just kill the processes that hang and then reinstall the locales package once you restart. I fully agree that manually killing processes is not something you want your average user to have to do, but the workarounds are out there and (for the most part) clearly documented in those links you speak of.
I think this discussion started from someone wondering if Dell will continue to support this laptop several years down the road. To me it seems that as hardware matures its support just gets better. Just think of how many posts you hear about people putting xubuntu on their "old P2 sitting around collecting dust" and it "just works" because that hardware is well understood. So as long as Dell provides enough information about the hardware at the onset, then it will be supported well at first and as bugs/issues arise they will be incorporated into the mainstream codebases.
I put my Mom on Ubuntu
I'm in the process of doing the same. However, it may be short lived. She was having issues with her USB controller under XP, so I told her to give Ubuntu a try to see if it was a hardware or software issue. Its looking like its a hardware issue, but I just got her setup to dual-boot and haven't had a chance to really see what is going on. If she does buy something new, I doubt she'll stick with it "just because."
Not that it's bloaty for a non-geek, who's probably used to winXP
I think it is about "just right" for the non-geek but (as always) there is always room for improvement. If she does make "the switch" I'm sure that we'll end up using the system completely different and both end up learning more as a result. Example: I'm pretty sure it can be done via the GUI, but the only way I know how to find my IP address is via ifconfig...so after I was talking to her about how Ubuntu is so easy is to use within the first 5 minutes of her using the liveCD I had to tell her to break out the terminal. Like I said, there is probably a way to do it with the GUI that makes it look nice and pretty, but I don't know how to. Either way, I was able to login with Remote Desktop and get he at least up and running. This weekend will be the big challenge...wish me luck!
Something about the article itself. There is that saying (you don't have to agree, but still): "Ubuntu on the desktop, Debian on the server" and the summary talks mostly about KDE4 and games?!? This is just my personal opinion but as a (somewhat) Debian sysadmin, I'd really prefer that they focus on keeping things fast and stable instead of trying to throw in extra eye-candy, because the only Debian installs I've ever done were all going to be headless...
If you spent $1.99 for a ring tone, you aren't going to call for tech support.
Haha, I think I've just found a new pastime:
Tech Support: Hello, how can I help you.
Me: I'm having difficulty with my ring tone.
Tech Support: What seems to be the problem?
Me: Well I installed it and now its playing Fergie.
Tech Support: Umm, isn't that what you purchased?
Me: Yes, but I want it to play a different song now, so how can I file a bug report?
Tech Support: Wait, what?
Me: I tried to download some patches and recompile, but it is not working. Do you have a wiki?
Tech Support: What patches, what are you talking about and no we don't have a wiki.
Me: Can you hold on a sec, I need to speak with my manager. [sit phone down and wait several minutes].
Me: Sorry about that, where were we. How do I configure my ringtone to play Miley Cryus?
Tech Support: You have to buy that ringtone.
Me: No, I've already bought one I just want to change it now.
Tech Support: Umm, I don't think you understand how this works.
Me: This ringtone sucks! I want my money back! [click]
Hate to self promote here...but why not buy the book?
Quite the entertaining read.
and for much less money bottle for bottle
Really? You mean to tell me that you can actually produce a superior product in micro quantities to a highly sophisticated and (almost) fully automated continuous brewing process that can also buy their raw resources in bulk quantities?
You might be able to produce a similar product at roughly the same cost that ends up costing almost the same amount to you the consumer, but you are negating both their (and the retailer's) markup as well as ignoring your time as a monetary resource.
FYI, I know exactly what you meant but that particular statement is either wrong or you've got a multi-billion dollar process at your disposal...either way, I'd love to try your brew. Cheers.
but OTOH Atom is quite a bit faster than even the fastest Atom.
In that case I'll take the fastest Atom, which is faster than the fastest Atom, which is faster than the fastest Atom....
So long as the power consumption holds constant, after a few iterations this thing should be able to run circles around even the BlueGene/L...all in a sub-compact notebook/UMPC!
My car doesn't suddenly quit for no reason.
...have patience my friend, its just a matter of time.
That is truly great advice, but can be misused in the hands of idiot parents who substitute "Kids have to make mistakes" with "Kids should have no rules or discipline." Part of being a (good) parent is setting rules and limits, then justly punishing when their children do (and they always will) break the rules. What makes it difficult is that no two children are the same, so there isn't a one-size-fits-all crime/punishment rulebook to follow.
Great parents respect their kids, but they also "instill" respect in their kids for themselves and other adults. Bottom line, having/raising kids is a HUGE responsibility, and it should be treated as such...a HUGE responsibility.
It's an arms race, and this system won't work forever, but it's fairly easy to implement and fairly difficult to overcome.
Not really, its all about scale. That system wouldn't last more than just a few seconds if a full "attack" were performed by a large botnet. The number of permutations is relatively finite, therefore with a large number of computers trying to "solve" the problem, once the correct answers were "cracked" then they could be shared and eventually the bots either know all of the answers, or you removed *all* of the questions from the list. I'm not saying this is an ineffective system for small/medium sites, but it wouldn't cut it for really large sites.
Axle, is that you?
3: ??? *&@(*@(@*))
Does that step involve boobs? If it does, great! Because that is where the real profits are.
This got modded funny, but its actually insightful. You'd be surprised...or not, guess it just depends on how tech savvy the people you know are (everything is relative).
I think that sentiment can be applied to just about anything these days...
So you are saying that this guy whoever he is, is a heteronym? Cause I still can't see a difference.
Units
Weapons
I can hear it now..."ZOMG, Playboy Bunnies with frickin beer" ...wait
And then when the gov't does try it they put the sign about 100 yards in front of the bridge that says "If you hit this sign, duck!"
I honestly don't think that it has anything to do with GPS. I mean, have you seen the size of trucks these days? Its probably a side effect of the hormone filled meats they ate when they were just little personal transports. Also, is it just me or are the new rigs getting crazier and crazier with their axles. Back in my day, 4 was enough for anything...now you're seeing all of these high fangled 6'ers and 8'ers.
Get off my roadway!
However, I think that the article is talking about undergraduate level courses which (in general) are graded less stringently that graduate and above assignments like what you are talking about. I went to a (well-respected) private university, and while you could probably get away with this for some aspects of CS101, the finals were actually written. You had to analyze code, and determine the output. As the courses got more advanced, so did the student/professor interaction. All of this makes me (and hopefully my employers) appreciate the fact that my degree is actually worth more than all of these "pay $X and attend for 4 years" knockoffs.
This is like the "game" between drm and hackers. They can put up roadblocks but they will eventually be "cracked." What is stopping the students from outsourcing the original papers and then outsourcing (locally) the "handwriting?" Furthermore they could just using some of the really great "handwriting fonts."
The problem isn't stopping the act of cheating, its stopping the underlying desire/incentive to cheat. If all it takes to get a degree is to have papers turned in on time for 4 years, then you get magically get paid more because you can put a piece of paper on your wall why not outsource? Your increased income will surely offset the outsourcing costs. But if you are in a job/industry where you actually have to use/perform using your "learned" skillsets *OR YOU GET FIRED* then I think the incentive to actually learn/develop will be maintained.