That Matrix RAID thing is a total joke. I have it on my new machine...it doesn't make sense at all. For those of you who don't know, Matrix RAID is Intel's new technology that allows you to set up as many different types of RAID volume as the chipset supports across n disks. For instance, say you have two disks, and you chipset supports RAID-0 (striping) and RAID-1 (mirroring). Each of those types of RAID only require two disks, so Matrix RAID technology allows you to split those disks in to a RAID-0 volume of 200GB and a RAID-1 volume of 100GB. Supposedly, this would allow you to store your OS and apps on the stripe volume and your data on the mirror volume.
Why would you want to do this? Well, presumably, because it gives your data redundancy (and 2x read speed) and gives your OS and apps 2x read/write performance. But it doesn't work...like all RAID, it doesn't make sense to mix'n'match different RAIDs across the same physical drives because as soon as you copy a file from one volume to the other, you're trying to pull data off of both disks and write it to both disks. A bus-locking nightmare that really spams the processor with non-SCSI drives.
By the way, not to get off-topic, but if you had one machine with 4 200GB drives and another with 3 80GB drives, and you want some redundancy on each machine for important data and as much performance you can get out of the rest, how would you set things up?
I don't measure the technology bounce-back by stock prices or by how certain products are doing. I'm measuring it by how easy/difficult it is for me, as a software developer with several years' experience, to get a good job. By that measure, I saw activity increase markedly a little more than a year ago. That's when it happened. It's over now, and things are evening out--they will (hopefully) never restore to the days of the bubble blow-up, so as long as we keep our expectations reasonable...what more do we want?
The problem is electricity. If you have an electrical junction, current must flow through it as if at a stop light--one pathway goes while all other wait. This is the cause of 99.999% of waiting for stuff in a computer to happen. Light, on the other hand, does not have to wait for other light. Light beams simply pass right on through each other and keep on keepin' on. Furthermore, two light beams can interfere on a sensor in all sorts of complicated ways to convey all sorts of complicated information--one light pulse can convey one bit of information, but two can convey 4 or more, and so on.
Optical computing, when it finally gets here, will provide so much computing power it will not be consumable, even by Bill.
I've done a lot of thinking about this...I've come to the conclusion that too often, management tries to replace good ol' fashioned thinking with process. It doesn't work. People tend to get focused in on what they're doing to the exclusion of all else, and that means the smart people are cubbyholed and only have half the story and can't see where other parts of the project are failing, and dumb people have free reign over their little part.
If the ratio of intelligence to complexity is too low, then the project will fail no matter what process is in place or who is managing it. That's all there is to it. There's a lot of cool stuff out there to be done in development...sadly, most of the good ideas will never make it because the people working on them don't use common sense and best practices...they're just not smart enough to see what's important and what isn't.
This isn't one of those self-important diatribes from a holier-than-thou developer, either...true I'm a developer, but I admit when I'm too dumb to handle the particulars of a project; usually, that means the project is too complex for most people, but they press on anyway. Those projects always go down in flames eventually.
You have to know what the strenghs and weaknesses of your team and its members are, and exploit those to the fullest. Maybe, then, you can barely accomplish a project if the goal of that project is simple enough.
I have to say, this is a dumb idea. It merely attracts a different kind of thief, a much more prevalent kind. Those who are not generally bad people but will stoop low enough to steal a pizza they see lying unattended.
I think this study is flawed. Aren't the people predisposed to spending lots of time on the Internet actually *more* socially engaged (albeit virtually so) than they were previously? I think so...as I understand it, this study doesn't measure the demographics before and after Internet presence, they just compared the two. Likely you'll find that, before, these people weren't socializing anyway--they were on the computer. Now the only difference is, they're hardwired.
I'm guessing they'd cover the barcode of a $230 DeLonghi toaster with a barcode for a $40 Oster toaster. This is just an example--Wal*Mart, of course, doesn't sell $230 toasters.
Or, perhaps they didn't even switch brands. It's not hard to find two televisions in the same line that differ in price but don't differ in manufacturer, size, etc. I'm guessing if these people makde $1.5M, they had some kind of system.
Still, though, that's a long time standing in line to return stuff...
This is part of John Thompson's bid to make Symantec "six by six"--that's $6B revenue (or profit? I think revenue...) by end of 2006. It's a great promise to investors, but not if it requires reckless M&A's.
When I heard about this (when I worked there), I thought, "As long as it doesn't cause 'zero by seven.'"
Well, it is more intelligent to use other inboxes that exclusively get spam. The one drawback to Bayesian filtering schemes is that you have to diligently sort the ham from the spam, and the more mail you sort by hand, the more accurate it is. This is a huge problem for most users because the spam is getting so good at getting around these filters nowadays, you need to hand sort in the tens of thousands of messages for the Bayesian filter to be effective.
On the other hand, if you have several addresses you don't give out to your personal contacts, you know for a fact that any mail you receive there is unrequested, i.e., spam. So it can immediately be piped to your filter as such.
(By the way, you misread my message...I wasn't setting myself up in opposition to the parent, I was jumping on the parent's bandwagon. The "you guys" doesn't refer to the parent poster, it refers to the parent poster's audience. Like when you happen upon a conversation between person A and person B, and A says, "B, you're an idiot!" And then you say (to B), "You know, I hate to tell you this, but you are an idiot." That's why I used "you guys," a plural...I did not, in fact, believe the parent poster to be a group of people collaboratively typing.)
It doesn't work. You probably have seen a marked decrease in spam because your ISP got some anti-spam solution at their end and they're not forwarding along email anymore from known spammers. Trust me--there's no way your wife's actions had anything to do with it. I used to work for Brightmail (anti-spam company) right before they were bought by Symantec, I know this is definitely a trick spammers use.
Actually...I hate to tell you guys this, but most spammers use those unsubscribe requests all right. They use them to verify that the email address is active, and it goes into a higher priority hit list. Even if they're in the US where the law says they must honor your unsub request, there's nothing that says they can't sell the information to other spammers that this is an actively used email address with a real live person on the other end of it.
About 18 months ago I did a little experiment. I set up my own junk inboxes at different email services and started handing them out. Three of them I unsub'd every spam email I got, and the other three I didn't. Guess which one eventually ended up getting buried in 10 times more spam...
I have a friend that is quite intelligent. He did a spin on the same idea, and I recommend it to anyone that wants to cut their spam to one or two mails per week (or you could just get a gmail account--I only get a few spam messages per week over there). Here's how it works...
Go out to every free email service you can get your hands on that supports POP3 download. Hand those addresses out to every spam list you can get your hands on. Periodically (every hour or so) download those messages into your Bayesian spam filter, marking them as spam (salearn that comes with spam assassin, for instance). I know of no better way to train your filter system and keep your spam stats up-to-date.
Of course, this isn't totally free of manual intervention. There's the initial setup of all this, which is more or less a one-time thing, but for it to truly work well, you have to make sure you also pipe all your regular mail (ham, as spam assassin calls it) into your Bayesian filter as non-spam mail, and if any spam does show up at your regular address, make sure you sort it into a separate folder and deal with it as spam. The spammers are getting more and more clever every day, and the line between spam and ham gets ever fainter, requiring that much more learning by the filtering system to keep straight what's what. But it's really not more work than you go through anyway, and you'll collect far more stats to use against the spammers than you otherwise would.
And let's not forget the best part, either. Signing up for and collecting all that spam costs spammers a little change (though, you could argue it also costs the hosts of your spam accounts, though you can delete the downloaded messages off the server every hour as part of the d/l to try and minimize impact on them).
Yes, I am in that position. In much the same way my boss is in a position to judge the work I haven't yet completed (or haven't even yet been assigned). We are the consumers of the information--if we're not the ones to say whether it's done well or not, then who will?
This is a discussion of roles, not actual work product.
I am in a position to judge. I don't speak Hindi or Punjabi either, but I damn well want accurately digitized texts! Does that make me racist or something?
When is all this PC non-judgmental crap going to stop? They're being hired to do a job. They either do it well or they don't, and we have every right to judge the result as consumers of their work.
Ok, so here's my idea to fix this. Presumably, they're indexing all this information so people will actually use it. OCR errors are not like many other kinds of errors, in that they're easy to recognize most times (o instead of a zero, etc).
So, I would advocate loading the texts in two places...one in the searchable DB and one in a wiki format. See which one gets fixed more quickly and more accurately.:-)
Wellllll....not necessarily. Apparently Canon has increased the density of photosensors on the 20D from the 10D while *lowering* the noise levels. (I think they did this using noise reduction filters--not sure if they meant hardware or software...)
Well...that all digitals are "single lens" right now might or might not be true. There's no reason they can't make a digital twin lens reflex. (Yes, these exist--google the Rolleiflex Twin Lens Reflex for an example.)
Wow. These galleries are one of the most stark examples of growth I've ever seen. If you have a photography mentor/professor/teacher, show him/her these two galleries and thank that person profusely. If this was all solf-taught...well, wow.
Yea, I have to say that DoF preview is a feature of limited usefulness. Every pro I've ever talked to tells me the same thing--if you use it enough, you'll be able to associate the dim image you're seeing (especially with smaller apertures) with the final result. But I'm starting to think it requires you to take a few thousand shots a week like most pros do.:-)
I'm a photography buff, and I entered this thread hoping to kick off with some serious discussion. Instead, at least 20% of this page is about the idiotic use of flash on this article. Does it suck? Yes. Is it out of the way thanks to this post's parent? Yes! Let's stop talking about it! (And thanks to AC above for getting it out of the way...but now it's out of the way people!)
I checked that lens out, and it's a nice lens, but I have to admit I wanted to stay full Canon for my starter set (just bought the 20D with a 580EX flash, 10-22mm, 17-85mm, 50mm, and 75-300mm). Not of little importance was the 10-22mm's image stabilization...I plan to use it stopped down a bit past its wide-open state (vignetting--physics says it's unavoidable at some point) and in dawn/dusk lighting situations, so IS will definitely open a few more creative options for me. And what am I trying to do, save $200 at the expense of some shots when I'm already in 20 times that much?:-) At this point, after that much investment, if I lose shots even just occasionally that I didn't have to, I'd be kicking myself. That's the same reason I ended up going with the 17-85mm IS instead of the vastly cheaper 18-55mm that Canon is pushing. Less lens swaps plus IS means that the extra $400 investment there (
Well, that's fair, but I'll submit to you that you can't take decent quality normal pictures in daylight without the removal of that IR filter causing all sorts of haze problems (unless they have an internal one that can slide in front of the sensor and out...which would be clever--why don't all cameras have this, come to think of it?).
Slashdot Mirror
That Matrix RAID thing is a total joke. I have it on my new machine...it doesn't make sense at all. For those of you who don't know, Matrix RAID is Intel's new technology that allows you to set up as many different types of RAID volume as the chipset supports across n disks. For instance, say you have two disks, and you chipset supports RAID-0 (striping) and RAID-1 (mirroring). Each of those types of RAID only require two disks, so Matrix RAID technology allows you to split those disks in to a RAID-0 volume of 200GB and a RAID-1 volume of 100GB. Supposedly, this would allow you to store your OS and apps on the stripe volume and your data on the mirror volume.
Why would you want to do this? Well, presumably, because it gives your data redundancy (and 2x read speed) and gives your OS and apps 2x read/write performance. But it doesn't work...like all RAID, it doesn't make sense to mix'n'match different RAIDs across the same physical drives because as soon as you copy a file from one volume to the other, you're trying to pull data off of both disks and write it to both disks. A bus-locking nightmare that really spams the processor with non-SCSI drives.
By the way, not to get off-topic, but if you had one machine with 4 200GB drives and another with 3 80GB drives, and you want some redundancy on each machine for important data and as much performance you can get out of the rest, how would you set things up?
I don't measure the technology bounce-back by stock prices or by how certain products are doing. I'm measuring it by how easy/difficult it is for me, as a software developer with several years' experience, to get a good job. By that measure, I saw activity increase markedly a little more than a year ago. That's when it happened. It's over now, and things are evening out--they will (hopefully) never restore to the days of the bubble blow-up, so as long as we keep our expectations reasonable...what more do we want?
The problem is electricity. If you have an electrical junction, current must flow through it as if at a stop light--one pathway goes while all other wait. This is the cause of 99.999% of waiting for stuff in a computer to happen. Light, on the other hand, does not have to wait for other light. Light beams simply pass right on through each other and keep on keepin' on. Furthermore, two light beams can interfere on a sensor in all sorts of complicated ways to convey all sorts of complicated information--one light pulse can convey one bit of information, but two can convey 4 or more, and so on.
Optical computing, when it finally gets here, will provide so much computing power it will not be consumable, even by Bill.
If the guy's worth it, why wouldn't you pay that? It's a free market problem...if I do the work of ten men...why not make the money of as many?
I've done a lot of thinking about this...I've come to the conclusion that too often, management tries to replace good ol' fashioned thinking with process. It doesn't work. People tend to get focused in on what they're doing to the exclusion of all else, and that means the smart people are cubbyholed and only have half the story and can't see where other parts of the project are failing, and dumb people have free reign over their little part.
If the ratio of intelligence to complexity is too low, then the project will fail no matter what process is in place or who is managing it. That's all there is to it. There's a lot of cool stuff out there to be done in development...sadly, most of the good ideas will never make it because the people working on them don't use common sense and best practices...they're just not smart enough to see what's important and what isn't.
This isn't one of those self-important diatribes from a holier-than-thou developer, either...true I'm a developer, but I admit when I'm too dumb to handle the particulars of a project; usually, that means the project is too complex for most people, but they press on anyway. Those projects always go down in flames eventually.
You have to know what the strenghs and weaknesses of your team and its members are, and exploit those to the fullest. Maybe, then, you can barely accomplish a project if the goal of that project is simple enough.
I have to say, this is a dumb idea. It merely attracts a different kind of thief, a much more prevalent kind. Those who are not generally bad people but will stoop low enough to steal a pizza they see lying unattended.
I think this study is flawed. Aren't the people predisposed to spending lots of time on the Internet actually *more* socially engaged (albeit virtually so) than they were previously? I think so...as I understand it, this study doesn't measure the demographics before and after Internet presence, they just compared the two. Likely you'll find that, before, these people weren't socializing anyway--they were on the computer. Now the only difference is, they're hardwired.
I'm guessing they'd cover the barcode of a $230 DeLonghi toaster with a barcode for a $40 Oster toaster. This is just an example--Wal*Mart, of course, doesn't sell $230 toasters.
Or, perhaps they didn't even switch brands. It's not hard to find two televisions in the same line that differ in price but don't differ in manufacturer, size, etc. I'm guessing if these people makde $1.5M, they had some kind of system.
Still, though, that's a long time standing in line to return stuff...
This is part of John Thompson's bid to make Symantec "six by six"--that's $6B revenue (or profit? I think revenue...) by end of 2006. It's a great promise to investors, but not if it requires reckless M&A's.
When I heard about this (when I worked there), I thought, "As long as it doesn't cause 'zero by seven.'"
Well, it is more intelligent to use other inboxes that exclusively get spam. The one drawback to Bayesian filtering schemes is that you have to diligently sort the ham from the spam, and the more mail you sort by hand, the more accurate it is. This is a huge problem for most users because the spam is getting so good at getting around these filters nowadays, you need to hand sort in the tens of thousands of messages for the Bayesian filter to be effective.
On the other hand, if you have several addresses you don't give out to your personal contacts, you know for a fact that any mail you receive there is unrequested, i.e., spam. So it can immediately be piped to your filter as such.
(By the way, you misread my message...I wasn't setting myself up in opposition to the parent, I was jumping on the parent's bandwagon. The "you guys" doesn't refer to the parent poster, it refers to the parent poster's audience. Like when you happen upon a conversation between person A and person B, and A says, "B, you're an idiot!" And then you say (to B), "You know, I hate to tell you this, but you are an idiot." That's why I used "you guys," a plural...I did not, in fact, believe the parent poster to be a group of people collaboratively typing.)
It doesn't work. You probably have seen a marked decrease in spam because your ISP got some anti-spam solution at their end and they're not forwarding along email anymore from known spammers. Trust me--there's no way your wife's actions had anything to do with it. I used to work for Brightmail (anti-spam company) right before they were bought by Symantec, I know this is definitely a trick spammers use.
Actually...I hate to tell you guys this, but most spammers use those unsubscribe requests all right. They use them to verify that the email address is active, and it goes into a higher priority hit list. Even if they're in the US where the law says they must honor your unsub request, there's nothing that says they can't sell the information to other spammers that this is an actively used email address with a real live person on the other end of it.
About 18 months ago I did a little experiment. I set up my own junk inboxes at different email services and started handing them out. Three of them I unsub'd every spam email I got, and the other three I didn't. Guess which one eventually ended up getting buried in 10 times more spam...
I have a friend that is quite intelligent. He did a spin on the same idea, and I recommend it to anyone that wants to cut their spam to one or two mails per week (or you could just get a gmail account--I only get a few spam messages per week over there). Here's how it works...
Go out to every free email service you can get your hands on that supports POP3 download. Hand those addresses out to every spam list you can get your hands on. Periodically (every hour or so) download those messages into your Bayesian spam filter, marking them as spam (salearn that comes with spam assassin, for instance). I know of no better way to train your filter system and keep your spam stats up-to-date.
Of course, this isn't totally free of manual intervention. There's the initial setup of all this, which is more or less a one-time thing, but for it to truly work well, you have to make sure you also pipe all your regular mail (ham, as spam assassin calls it) into your Bayesian filter as non-spam mail, and if any spam does show up at your regular address, make sure you sort it into a separate folder and deal with it as spam. The spammers are getting more and more clever every day, and the line between spam and ham gets ever fainter, requiring that much more learning by the filtering system to keep straight what's what. But it's really not more work than you go through anyway, and you'll collect far more stats to use against the spammers than you otherwise would.
And let's not forget the best part, either. Signing up for and collecting all that spam costs spammers a little change (though, you could argue it also costs the hosts of your spam accounts, though you can delete the downloaded messages off the server every hour as part of the d/l to try and minimize impact on them).
Yes, I am in that position. In much the same way my boss is in a position to judge the work I haven't yet completed (or haven't even yet been assigned). We are the consumers of the information--if we're not the ones to say whether it's done well or not, then who will?
This is a discussion of roles, not actual work product.
I am in a position to judge. I don't speak Hindi or Punjabi either, but I damn well want accurately digitized texts! Does that make me racist or something?
When is all this PC non-judgmental crap going to stop? They're being hired to do a job. They either do it well or they don't, and we have every right to judge the result as consumers of their work.
Ok, so here's my idea to fix this. Presumably, they're indexing all this information so people will actually use it. OCR errors are not like many other kinds of errors, in that they're easy to recognize most times (o instead of a zero, etc).
So, I would advocate loading the texts in two places...one in the searchable DB and one in a wiki format. See which one gets fixed more quickly and more accurately. :-)
My solution to this problem is iron-clad. I keep all my banking accounts empty and have nothing of value on my computer, or in life.
Yea, I knew to stay out of this part of the discussion. Looks like we have our very own little circle of confusion going on right here...
Wellllll....not necessarily. Apparently Canon has increased the density of photosensors on the 20D from the 10D while *lowering* the noise levels. (I think they did this using noise reduction filters--not sure if they meant hardware or software...)
Well...that all digitals are "single lens" right now might or might not be true. There's no reason they can't make a digital twin lens reflex. (Yes, these exist--google the Rolleiflex Twin Lens Reflex for an example.)
Wow. These galleries are one of the most stark examples of growth I've ever seen. If you have a photography mentor/professor/teacher, show him/her these two galleries and thank that person profusely. If this was all solf-taught...well, wow.
Depending on the subject, sometimes I'll even go so far as to put a brown paper bag over their head.
Yea, I have to say that DoF preview is a feature of limited usefulness. Every pro I've ever talked to tells me the same thing--if you use it enough, you'll be able to associate the dim image you're seeing (especially with smaller apertures) with the final result. But I'm starting to think it requires you to take a few thousand shots a week like most pros do. :-)
Informative? How about off-topic?
I'm a photography buff, and I entered this thread hoping to kick off with some serious discussion. Instead, at least 20% of this page is about the idiotic use of flash on this article. Does it suck? Yes. Is it out of the way thanks to this post's parent? Yes! Let's stop talking about it! (And thanks to AC above for getting it out of the way...but now it's out of the way people!)
I checked that lens out, and it's a nice lens, but I have to admit I wanted to stay full Canon for my starter set (just bought the 20D with a 580EX flash, 10-22mm, 17-85mm, 50mm, and 75-300mm). Not of little importance was the 10-22mm's image stabilization...I plan to use it stopped down a bit past its wide-open state (vignetting--physics says it's unavoidable at some point) and in dawn/dusk lighting situations, so IS will definitely open a few more creative options for me. And what am I trying to do, save $200 at the expense of some shots when I'm already in 20 times that much? :-) At this point, after that much investment, if I lose shots even just occasionally that I didn't have to, I'd be kicking myself. That's the same reason I ended up going with the 17-85mm IS instead of the vastly cheaper 18-55mm that Canon is pushing. Less lens swaps plus IS means that the extra $400 investment there (
Oh, *that* one.
Well, that's fair, but I'll submit to you that you can't take decent quality normal pictures in daylight without the removal of that IR filter causing all sorts of haze problems (unless they have an internal one that can slide in front of the sensor and out...which would be clever--why don't all cameras have this, come to think of it?).