To my understanding, blocking any ports isn't considered blocking based on _content_ which is what puts you at risk of losing common carrier protection.
I believe it would probably even be ok to block any website IP that had more than "X" traffic - as long as you did it consistently.
I am not a lawyer, but here's the very rough deal: Someone who is transmitting information as a "common carrier" isn't responsible for that content under very broad rules.
Someone who ISN'T a common carrier still isn't automatically responsible. There's lots of ways to NOT be liable, but this is one especially broad and reliable kind of protection and companies don't want to give it up.
1. A _hosting_ company can make whatever AUP they want, and they can enforce it; they aren't acting as a common carrier anyway. A hosting company can always be liable for what they're hosting (to some extent, after they know)
2. An _access_ company is protected by common carrier rules. So if your DSL provider prevents you from seeing certain sites then they become somewhat liable for all the content passed over their lines.
It doesn't count if the filtering is optional (most family-friendly) or if it is technical (most kinds of AV protection; anything supposedly to keep bandwidth down.)
So there's a narrow techincal distinction in there somewhere, but the rough idea is that people who are _bandwidth_ providers don't want to stop you from accessing something based on content, because it reduces their protection.
If you encrypt the Windows Filesystem then there is no trivial way to get the data without the decryption key. This is breakable given time, etc. depending on how strong the encryption is.
It also makes it a royal pain to recover if certain things go wrong.
If you DIDN'T encrypt the filesystem then it is absolutely trivial to change the admin password, to put the disk in another machine, to boot linux and read the HD... etc. Because the data is completely insecure.
This is COMPLETELY THE SAME on Linux. You can boot any normal linux install from a boot disk and reset the root password. Actually in Linux you don't even usually need a disk, you just need to enter init=bash in LILO (depending on LILO/Linux settings; this can certainly be turned off)
Security means not giving people physical access, or at a minimum restricting the boot sequence to only be harddrive, locking the CPU, locking the BIOS and locking the boot loader. This is true in any OS.
We should scream about the White House press corps - although I don't think we do it nearly enough. Because the fundamental core of democracy is that everyone needs unfettered access.
The government has an obligation to disclose as much information as possible and not to lie about it. An obligation they don't take seriously.
Google, on the other hand, is a publicly held corporation. They have specific reporting requirements to shareholders and the SEC, but none to the press. Smart investors should read the actual filings and not necessarily trust 3rd party analysis without cause. Talking to reporters is something Google does because it is good for Google.
Please note that Google did NOT ban CNET from search or change the rankings of CNET in google or google news. Google did not try to sue them.
Google simply said they would not talk to CNET reporters. That's EXACTLY the punishment that befits writing an article a company doesn't like - that they'll find someone else to give their press releases to.
For instance, the Microsoft Office interface is "obvious" - because they publicly released the software.
If someone made StarOffice (later OpenOffice) while under NDA at Sun that NDA would not keep them from releasing parts of that code unless the programming techniques used were different than what any programmer would expect.*
(Copyright WOULD prevent them from releasing it, of course!)
But the fact that a clone of MSOffice is "obvious" does not mean it is trivial, it was still a huge project.
*Curiously, I think reverse engineered MS protocols/formats probably ARE covered by NDA even at a company reverse-engineering them. But if the OpenOffice format is published it would no longer be protected.
In all likelyhood an NDA doesn't cover obvious works like this - anything that could be reasonably discovered publicly. Doubtless he couldn't post the _documents_ that he converted.
However, I am also not willing to just assume that no company would ever consider letting someone sourceforge a script like this. It is 1) worth good advertising and 2) clearly not important enough to be worth selling. Release it in the company's name, or not depending on what they prefer.
At a minimum a lot of small companies would be fine with this - big companies would vary wildly.
So if there's only a few templates and they were a pain to work out, how about releasing your regex scripts to sourceforge or similar? Or posting here?
IE is no longer supported on the Mac, so they are cutting out all supported mac browsers to (Firefox + Safari)
I think that "all users of Firefox" and "all users of Apple" has even more weight, because EVERYONE knows who Apple is, even if they don't really know what the difference is.
As a rule, Covad still leases the "last mile" of wire from the ILEC. But they don't lease any DSL equipment.
Earthlink (for instance, I think) actually leases the DSL line from the ILEC AND the spot in the ILEC's DSLAM. So earthlink can offer DSL by only providing uplink bandwidth and everything else is virtual.
I have DSL through Cyberonic. From talking to their techs, the impression I have is that Cyberonic has the same sort of setup Speakeasy does, but replace Covad with MCI. (And make the service thereofore be less reliable)
I therefore believe that Covad is not quite the last-man standing of DSL CLEC.
I certainly agree that it costs more to upgrade to a new major version of OSX*
"user support costs" = costs to suppor the users (seperate from the systems) That is, money you spend on people to answer the phone and explain to your users how to do something that was technically already functioning. I believe that no system has a lower user support cost for untrained users than OSX.
Whether or not this dwarfs the cost of the more expensive machines depends on your organization, the people in it, whether they actually do diverse things, and how much you actually help them. Ditto for software upgrades...
So, if I was going to make a lab of a hundred computers for students, I would make it Linux. If I was going to make a server setup I'd make it client-platform independent, and I'd generally have it run Linux. If I was going to choose a workstation OS for a small business that does a lot of varied work on their computers I'd choose OSX, because the users would be able to get more different things done without a support call.
*But also, there is little that doesn't function adequately in 10.1 and almost nothing in 10.2 The newer versions are snazzier, but not required. A lot of places EOL computers in 3-4 years...
I don't care about the karma (or the dogma) but I think this is kindof funny - I've finally gotten a troll mod! I think that was the only one I was missing.
I just said that there are a lot of individual factors going into what OS to choose, but in an absolute sense Linux is usually better than Windows.
1. Intelligently running a server is never really free, it requires admin time, power, hardware and software. Even if you can get some of these things pseudo-free you generally could apply the resources to something else instead.
2. It is likely that the lowest Total Cost of Ownership for your systems in the medium term is WHATEVER YOU HAVE NOW. Switching is expensive, and retraining your admins to know another system can be even more expensive. At first noone will be experienced with the new system, and that has its own headaches. Because switching is a big pain, generally the time to recover the cost of switching a _functional_ system is usually going to be measured in years.
3. If you are already upgrading the functionality of your system or creating a new system, it is usually cheapest to develop in whatever your inhouse admins are familiar with if they have time to work on the development.
So the actual number of cases where you can make a fair comparison between Windows and Linux is small because an existing company has an existing set of people.
But after that recovery Linux costs less to install new machines, less to admin servers, less to maintain and less to support equally-trained users on. (Versions of Linux generally have a UI intuitiveness slightly behind Windows, but this is more than outweighed by the number of issues that are legitimate windows and application/windows compatibility bugs)
Now, if you compare Linux to OSX on workstations you get even lower user support costs but higher upfront costs.
REGULATED monopolies can provide better service than a competitive market, IF the regulators are on the ball. Usually this still ends up being more expensive (in terms of provider cost) on average for the same service, but it may be less expensive for the consumer if the monopoly's ability to keep the money as profit is reduced.
UNREGULATED monopolies NEVER provide better service at a better price. They generally provide better profits for the shareholders in the monopoly.
This is giving a state-funded legal monopoly over DSL to the ILECs. If they are going to do this, individual municipalities and state should be able to vote that ALL of the phone monopoly in their area can go to someone else, since that's why it was originally put in, to serve the people.
The gap filler is needed to keep the tiles from rattling on LIFTOFF. Once in space, we don't need it.
The reentry has very different pressures/angles - I believe the pressure of the reentry keeps the tiles from moving enough to bump each other too badly.
It might be slighly less accurate than your way, but only if the time on the existing server really is hyperaccurate.
(That is, SOAP directly to an authoritative time server is probably more accurate than a tunneling proxy, but a tunneling proxy is probably more accurate than the two sets of drift calculations involved in syncing from a random server synced from somewhere else.)
My "confusion" isn't exactly about which word to use. I compose/. posts at about 50 wpm. When composing and typing like that my fingers sometimes insert homonyms and near homonyms, even to words I clearly know the difference between.
It is somewhat like my fingers have a substantial word-vocabulary but are hard of hearing.
I don't usually bother rereading my posts, because I don't care enough:) Sorry.
You have to address the "bad customers" problem.
on
The Case for Free WiFi?
·
· Score: 2, Insightful
If you have infinite tables, Free Services are great, and often better than metered services, because metering is a pain.
But you don't have infinite tables, and somebody sitting on your wifi for an hour with a cup of coffee might be LOOSING you money.
If I could wave a wand and create whatever system I wanted I would have BOTH. This way you get to have the best of both worlds.
A free system that had capped bandwidth (50k, perhaps) and was turned off during periods of peak patronage of the cafe. You start out with a sign that says "may be turned off during peak hours" - and then you turn it off whenever you're getting too full.
A paid system that doesn't have bandwidth caps or has much higher ones and never goes off. This should cost less than using one of your workstations, but not necessarily that much less.
Setting both up is pretty easy; just install two access points.
Also, if you're going to put in a nice system be sure to offer it to all your nearby business and residential neighbors to make a few extra bucks.
I am not a lawyer, but I don't think you have an obligation to keep those records. You have an obligation to turn them over if you have them, and an obligation not to selectively destroy them AFTER you know something bad happened. But I don't believe you are required to track all the details.
Actually, that's NOT a law, at least in IL. Enforcement is entirely voluntary on the part of the theater. The rule at my favorite theater is "No children under 6 in an R rated movie after 6"
yep, that's it. (It is a second-run theater, but it's a NICE second run theater, and $3)
Time to get the backup back up is a valid consideration, but it's not the only one. How easy it is to MAKE the backup (and therefore how current it will be) is arguably more important because having more frequent backups IS more important. Driving to a different town every day is probably not an efficient use of your time unless you can't get enough pipe to xfer just the changed files. Hence online backups.
If you're paranoid then run your own backup host over ssh at a trusted someone's personal connection. But there's no solution superior to online backups if the incremental changes in files can be met with 100% of your extra nightly bandwidth. Try backup PC on sourceforge. Try using more than 1 at different locations.
Parent has a good, but different point: If you have a lot of data it's going to take a ton of time to get it back up. If this is likely to be a problem, then by all means find faster ways to ship your data. One way would be to drive and get whichever of your mulitple backup machines is closest. (If you only have 1 backup machine make a copy to take with you and leave the original where it was)
But another way, especially if you don't have access to the online backups, is to drive a harddrive full of stuff somewhere. IF your backup provider can do an restore from a partially recovered backup (ie, rsync) you can keep extra physical backups lying around and still having the online "current" backup to save you. That is, you could bring in your extra HD from a month ago and just rsync the stuff that changed.
1. Each successive screening has been better than the previous. They're still in post-production and they're still improving stuff.
They probably have a relatively small advertising budget. So,
2. You don't want to compete with the summer blockbusters - because you don't want to compete with their advertising budget. So you schedule a fall release to not compete with them.
3. You want to magnify your small $ as much as possible. So you release trailers on RELEVANT movies that interested people are going to see. And you do what you can to make your fans advertise for you for free/cheap.
Precision Teaching is a method - or perhaps a metamethod - of teaching that is scientifically based and has been proven very effective repeatedly, including in federal studies. I also think it would appeal to many/.ers - it appeals to me as an engineer. It is basically about rewards, feedback loops and good measurement. I go into much more detail below, including a significant bibliography. PT is one area of Behavioral Psychology.
I believe PT is the most definitively and scientifically successful system, but there are other highly successful systems (such as Direct Instruction) The differences in success between these systems are much smaller than the benefit of using any of them.
Unfortunately, the systemic problems with our education system are much bigger than this. When very successful teaching methods are demonstrated they are usually dismantled to be more "fair" - mostly more fair to other teachers, not students. Any very significant success like this is too much like proof that the current system is deficient, and anyone making those decisions has a lot invested in the current system - and a lot of blame for its failure.
Interestingly, it seems as though the rise of Austism is increasing the visibility of effective teaching methods. In very rough terms, "normal" children learn a lot of stuff even if you teach pretty badly, and children with some disorders have a very hard time learning in any method. Roughly, children with moderate Austism disorders learn well if taught very well and learn very poorly if not taught very well. So the success of these children seems to provide a good measure of the capability of an educational system.
Note: the above paragraph paints some very, very broad strokes. It most certainly does not apply to all cases, etc. It's what I could fit in a paragraph. This is/.
In a "normal" population the highest correlated success factor is economic level of the neighboorhood - no matter WHAT goes on in the school. In a rich enough neighboorhood, the parents apparently get tutors.
Regardless, Precision Teaching based methods have been used to great effect in populations that have been marginalized by the "normal" system. That these marginalized "hard to teach" kids can quickly catch up and surpass their peers is a testament to how poor the mainstream system is. (A private center by me typically teaches a year of reading in 20 instructional hours.)
***** The rest of this post is excerpted from an in-progress website about the topic. Since it's not done, I'm not posting it on/. Consider the following a draft (and a darned long post) The bottom is a large number of (possibly better written) links. These links include three functioning schools. To my knowledge all receive some public funding via students who transfer there when the public system can't help them.
First and foremost PT is scientific measurement and the principle that "The Learner is always Right" - In PT we scientifically and numerically measure the behavior of the learner. Because we measure it consistently and numerically over time we are also measuring the change in this behavior. In general if the behavior is not meeting our goals and/or the change in behavior is not meeting our goals we alter the environment the learner is in somehow. We have established many details of how to take good data that is extremely sensitive at predicting changes in behavior.
The second pillar of PT is the Standard Celeration Chart. Beyond simply measuring in a scientific and numeric way we measure and display our data using a standard graphic display - the Standard Celeration Chart . This means that anyone familiar with Precision Teaching can immediately tell what is going on with someone else's chart at a glance often even without being familiar with that particular situation.
The third pillar of PT is interventions. Using our standard measurement and data sharing the Precision Teaching c
Parent (currently modded redundant) has a very valid point. OSX is the easiest to maintain desktop *nix, and to boot the best common desktop in the world. Since it can run OSX, classic MacOS, BSD* and Linux* software it has considerably more options available to users.
Personally, I'm very happy with SuSE 9.2 and I was very happy with SuSE 7.3. I view SuSE as the perfect middle ground of great support and installers and RedHat. More importantly, SuSE has always, without fail, just worked for everything. But I'll admit I haven't examined RedHat in awhile.
*(from source. Binaries work if they're targetted to PPC, but there are more BSD/Linux binaries targetted to i386 than PPC)
Slashdot Mirror
To my understanding, blocking any ports isn't considered blocking based on _content_ which is what puts you at risk of losing common carrier protection.
I believe it would probably even be ok to block any website IP that had more than "X" traffic - as long as you did it consistently.
I am not a lawyer, but here's the very rough deal:
Someone who is transmitting information as a "common carrier" isn't responsible for that content under very broad rules.
Someone who ISN'T a common carrier still isn't automatically responsible. There's lots of ways to NOT be liable, but this is one especially broad and reliable kind of protection and companies don't want to give it up.
1. A _hosting_ company can make whatever AUP they want, and they can enforce it; they aren't acting as a common carrier anyway. A hosting company can always be liable for what they're hosting (to some extent, after they know)
2. An _access_ company is protected by common carrier rules. So if your DSL provider prevents you from seeing certain sites then they become somewhat liable for all the content passed over their lines.
It doesn't count if the filtering is optional (most family-friendly) or if it is technical (most kinds of AV protection; anything supposedly to keep bandwidth down.)
So there's a narrow techincal distinction in there somewhere, but the rough idea is that people who are _bandwidth_ providers don't want to stop you from accessing something based on content, because it reduces their protection.
If you encrypt the Windows Filesystem then there is no trivial way to get the data without the decryption key. This is breakable given time, etc. depending on how strong the encryption is.
It also makes it a royal pain to recover if certain things go wrong.
If you DIDN'T encrypt the filesystem then it is absolutely trivial to change the admin password, to put the disk in another machine, to boot linux and read the HD... etc. Because the data is completely insecure.
This is COMPLETELY THE SAME on Linux. You can boot any normal linux install from a boot disk and reset the root password. Actually in Linux you don't even usually need a disk, you just need to enter init=bash in LILO (depending on LILO/Linux settings; this can certainly be turned off)
Security means not giving people physical access, or at a minimum restricting the boot sequence to only be harddrive, locking the CPU, locking the BIOS and locking the boot loader. This is true in any OS.
We should scream about the White House press corps - although I don't think we do it nearly enough. Because the fundamental core of democracy is that everyone needs unfettered access.
The government has an obligation to disclose as much information as possible and not to lie about it. An obligation they don't take seriously.
Google, on the other hand, is a publicly held corporation. They have specific reporting requirements to shareholders and the SEC, but none to the press. Smart investors should read the actual filings and not necessarily trust 3rd party analysis without cause. Talking to reporters is something Google does because it is good for Google.
Please note that Google did NOT ban CNET from search or change the rankings of CNET in google or google news. Google did not try to sue them.
Google simply said they would not talk to CNET reporters. That's EXACTLY the punishment that befits writing an article a company doesn't like - that they'll find someone else to give their press releases to.
For instance, the Microsoft Office interface is "obvious" - because they publicly released the software.
If someone made StarOffice (later OpenOffice) while under NDA at Sun that NDA would not keep them from releasing parts of that code unless the programming techniques used were different than what any programmer would expect.*
(Copyright WOULD prevent them from releasing it, of course!)
But the fact that a clone of MSOffice is "obvious" does not mean it is trivial, it was still a huge project.
*Curiously, I think reverse engineered MS protocols/formats probably ARE covered by NDA even at a company reverse-engineering them. But if the OpenOffice format is published it would no longer be protected.
I am not a lawyer.
Ben
In all likelyhood an NDA doesn't cover obvious works like this - anything that could be reasonably discovered publicly. Doubtless he couldn't post the _documents_ that he converted.
However, I am also not willing to just assume that no company would ever consider letting someone sourceforge a script like this. It is 1) worth good advertising and 2) clearly not important enough to be worth selling. Release it in the company's name, or not depending on what they prefer.
At a minimum a lot of small companies would be fine with this - big companies would vary wildly.
So if there's only a few templates and they were a pain to work out, how about releasing your regex scripts to sourceforge or similar? Or posting here?
IE is no longer supported on the Mac, so they are cutting out all supported mac browsers to (Firefox + Safari)
I think that "all users of Firefox" and "all users of Apple" has even more weight, because EVERYONE knows who Apple is, even if they don't really know what the difference is.
Ben
The following is my understanding...
As a rule, Covad still leases the "last mile" of wire from the ILEC. But they don't lease any DSL equipment.
Earthlink (for instance, I think) actually leases the DSL line from the ILEC AND the spot in the ILEC's DSLAM. So earthlink can offer DSL by only providing uplink bandwidth and everything else is virtual.
I have DSL through Cyberonic. From talking to their techs, the impression I have is that Cyberonic has the same sort of setup Speakeasy does, but replace Covad with MCI. (And make the service thereofore be less reliable)
I therefore believe that Covad is not quite the last-man standing of DSL CLEC.
I certainly agree that it costs more to upgrade to a new major version of OSX*
"user support costs" = costs to suppor the users (seperate from the systems) That is, money you spend on people to answer the phone and explain to your users how to do something that was technically already functioning. I believe that no system has a lower user support cost for untrained users than OSX.
Whether or not this dwarfs the cost of the more expensive machines depends on your organization, the people in it, whether they actually do diverse things, and how much you actually help them. Ditto for software upgrades...
So, if I was going to make a lab of a hundred computers for students, I would make it Linux. If I was going to make a server setup I'd make it client-platform independent, and I'd generally have it run Linux. If I was going to choose a workstation OS for a small business that does a lot of varied work on their computers I'd choose OSX, because the users would be able to get more different things done without a support call.
*But also, there is little that doesn't function adequately in 10.1 and almost nothing in 10.2 The newer versions are snazzier, but not required. A lot of places EOL computers in 3-4 years...
I don't care about the karma (or the dogma) but I think this is kindof funny - I've finally gotten a troll mod! I think that was the only one I was missing.
I just said that there are a lot of individual factors going into what OS to choose, but in an absolute sense Linux is usually better than Windows.
And got marked troll.
Neat.
1. Intelligently running a server is never really free, it requires admin time, power, hardware and software. Even if you can get some of these things pseudo-free you generally could apply the resources to something else instead.
2. It is likely that the lowest Total Cost of Ownership for your systems in the medium term is WHATEVER YOU HAVE NOW. Switching is expensive, and retraining your admins to know another system can be even more expensive. At first noone will be experienced with the new system, and that has its own headaches. Because switching is a big pain, generally the time to recover the cost of switching a _functional_ system is usually going to be measured in years.
3. If you are already upgrading the functionality of your system or creating a new system, it is usually cheapest to develop in whatever your inhouse admins are familiar with if they have time to work on the development.
So the actual number of cases where you can make a fair comparison between Windows and Linux is small because an existing company has an existing set of people.
But after that recovery Linux costs less to install new machines, less to admin servers, less to maintain and less to support equally-trained users on. (Versions of Linux generally have a UI intuitiveness slightly behind Windows, but this is more than outweighed by the number of issues that are legitimate windows and application/windows compatibility bugs)
Now, if you compare Linux to OSX on workstations you get even lower user support costs but higher upfront costs.
No.
REGULATED monopolies can provide better service than a competitive market, IF the regulators are on the ball. Usually this still ends up being more expensive (in terms of provider cost) on average for the same service, but it may be less expensive for the consumer if the monopoly's ability to keep the money as profit is reduced.
UNREGULATED monopolies NEVER provide better service at a better price. They generally provide better profits for the shareholders in the monopoly.
This is giving a state-funded legal monopoly over DSL to the ILECs. If they are going to do this, individual municipalities and state should be able to vote that ALL of the phone monopoly in their area can go to someone else, since that's why it was originally put in, to serve the people.
The gap filler is needed to keep the tiles from rattling on LIFTOFF. Once in space, we don't need it.
The reentry has very different pressures/angles - I believe the pressure of the reentry keeps the tiles from moving enough to bump each other too badly.
You don't want to make a new SOAP service and have to do all the delay calculations.
If you have access to an external server, just tunnel NTP over HTTP. (http://htun.runslinux.net/docs.html)
Essentially no programming required.
It might be slighly less accurate than your way, but only if the time on the existing server really is hyperaccurate.
(That is, SOAP directly to an authoritative time server is probably more accurate than a tunneling proxy, but a tunneling proxy is probably more accurate than the two sets of drift calculations involved in syncing from a random server synced from somewhere else.)
My "confusion" isn't exactly about which word to use. I compose /. posts at about 50 wpm. When composing and typing like that my fingers sometimes insert homonyms and near homonyms, even to words I clearly know the difference between.
:) Sorry.
It is somewhat like my fingers have a substantial word-vocabulary but are hard of hearing.
I don't usually bother rereading my posts, because I don't care enough
yes, yes it is.
/. comments long ago : )
I stopped previewing my
If you have infinite tables, Free Services are great, and often better than metered services, because metering is a pain.
But you don't have infinite tables, and somebody sitting on your wifi for an hour with a cup of coffee might be LOOSING you money.
If I could wave a wand and create whatever system I wanted I would have BOTH. This way you get to have the best of both worlds.
A free system that had capped bandwidth (50k, perhaps) and was turned off during periods of peak patronage of the cafe. You start out with a sign that says "may be turned off during peak hours" - and then you turn it off whenever you're getting too full.
A paid system that doesn't have bandwidth caps or has much higher ones and never goes off. This should cost less than using one of your workstations, but not necessarily that much less.
Setting both up is pretty easy; just install two access points.
Also, if you're going to put in a nice system be sure to offer it to all your nearby business and residential neighbors to make a few extra bucks.
I am not a lawyer, but I don't think you have an obligation to keep those records. You have an obligation to turn them over if you have them, and an obligation not to selectively destroy them AFTER you know something bad happened. But I don't believe you are required to track all the details.
Yeah, but at least by 7 they probably understand what you mean when you kick them out of the theater!
Actually, that's NOT a law, at least in IL. Enforcement is entirely voluntary on the part of the theater. The rule at my favorite theater is "No children under 6 in an R rated movie after 6"
yep, that's it. (It is a second-run theater, but it's a NICE second run theater, and $3)
I believe HD is the right way to store everything now. But there are two different questions and you tend to muddle them together.
I agree that I would choose "disks you own" over "disks some service owns" if I could only have one service.
But I would still choose "online transmission" over "disks I ship"
Hence my suggestion that you use "online transmission to disks you own" Try backuppc at sourceforge.
Time to get the backup back up is a valid consideration, but it's not the only one. How easy it is to MAKE the backup (and therefore how current it will be) is arguably more important because having more frequent backups IS more important. Driving to a different town every day is probably not an efficient use of your time unless you can't get enough pipe to xfer just the changed files. Hence online backups.
If you're paranoid then run your own backup host over ssh at a trusted someone's personal connection. But there's no solution superior to online backups if the incremental changes in files can be met with 100% of your extra nightly bandwidth. Try backup PC on sourceforge. Try using more than 1 at different locations.
Parent has a good, but different point: If you have a lot of data it's going to take a ton of time to get it back up. If this is likely to be a problem, then by all means find faster ways to ship your data. One way would be to drive and get whichever of your mulitple backup machines is closest. (If you only have 1 backup machine make a copy to take with you and leave the original where it was)
But another way, especially if you don't have access to the online backups, is to drive a harddrive full of stuff somewhere. IF your backup provider can do an restore from a partially recovered backup (ie, rsync) you can keep extra physical backups lying around and still having the online "current" backup to save you. That is, you could bring in your extra HD from a month ago and just rsync the stuff that changed.
1. Each successive screening has been better than the previous. They're still in post-production and they're still improving stuff.
They probably have a relatively small advertising budget. So,
2. You don't want to compete with the summer blockbusters - because you don't want to compete with their advertising budget. So you schedule a fall release to not compete with them.
3. You want to magnify your small $ as much as possible. So you release trailers on RELEVANT movies that interested people are going to see. And you do what you can to make your fans advertise for you for free/cheap.
Precision Teaching is a method - or perhaps a metamethod - of teaching that is scientifically based and has been proven very effective repeatedly, including in federal studies. I also think it would appeal to many /.ers - it appeals to me as an engineer. It is basically about rewards, feedback loops and good measurement. I go into much more detail below, including a significant bibliography. PT is one area of Behavioral Psychology.
/.
/. Consider the following a draft (and a darned long post) The bottom is a large number of (possibly better written) links. These links include three functioning schools. To my knowledge all receive some public funding via students who transfer there when the public system can't help them.
I believe PT is the most definitively and scientifically successful system, but there are other highly successful systems (such as Direct Instruction) The differences in success between these systems are much smaller than the benefit of using any of them.
Unfortunately, the systemic problems with our education system are much bigger than this. When very successful teaching methods are demonstrated they are usually dismantled to be more "fair" - mostly more fair to other teachers, not students. Any very significant success like this is too much like proof that the current system is deficient, and anyone making those decisions has a lot invested in the current system - and a lot of blame for its failure.
Interestingly, it seems as though the rise of Austism is increasing the visibility of effective teaching methods. In very rough terms, "normal" children learn a lot of stuff even if you teach pretty badly, and children with some disorders have a very hard time learning in any method.
Roughly, children with moderate Austism disorders learn well if taught very well and learn very poorly if not taught very well. So the success of these children seems to provide a good measure of the capability of an educational system.
Note: the above paragraph paints some very, very broad strokes. It most certainly does not apply to all cases, etc. It's what I could fit in a paragraph. This is
In a "normal" population the highest correlated success factor is economic level of the neighboorhood - no matter WHAT goes on in the school. In a rich enough neighboorhood, the parents apparently get tutors.
Regardless, Precision Teaching based methods have been used to great effect in populations that have been marginalized by the "normal" system. That these marginalized "hard to teach" kids can quickly catch up and surpass their peers is a testament to how poor the mainstream system is. (A private center by me typically teaches a year of reading in 20 instructional hours.)
*****
The rest of this post is excerpted from an in-progress website about the topic. Since it's not done, I'm not posting it on
First and foremost PT is scientific measurement and the principle that "The Learner is always Right" - In PT we scientifically and numerically measure the behavior of the learner. Because we measure it consistently and numerically over time we are also measuring the change in this behavior. In general if the behavior is not meeting our goals and/or the change in behavior is not meeting our goals we alter the environment the learner is in somehow. We have established many details of how to take good data that is extremely sensitive at predicting changes in behavior.
The second pillar of PT is the Standard Celeration Chart. Beyond simply measuring in a scientific and numeric way we measure and display our data using a standard graphic display - the Standard Celeration Chart . This means that anyone familiar with Precision Teaching can immediately tell what is going on with someone else's chart at a glance often even without being familiar with that particular situation.
The third pillar of PT is interventions. Using our standard measurement and data sharing the Precision Teaching c
Parent (currently modded redundant) has a very valid point. OSX is the easiest to maintain desktop *nix, and to boot the best common desktop in the world. Since it can run OSX, classic MacOS, BSD* and Linux* software it has considerably more options available to users.
Personally, I'm very happy with SuSE 9.2 and I was very happy with SuSE 7.3. I view SuSE as the perfect middle ground of great support and installers and RedHat. More importantly, SuSE has always, without fail, just worked for everything. But I'll admit I haven't examined RedHat in awhile.
*(from source. Binaries work if they're targetted to PPC, but there are more BSD/Linux binaries targetted to i386 than PPC)