Some adsl providers also impose caps downstream and don't offer higher than 1.5mbit service, mainly because it's too much trouble.
If you're a big business (*cough*cough*ameritech/sbc), the best way to make money with the least trouble is to make an aritifically small dsl radius, cap the lines at something way below the DMT standard (cap it at oh, 1.5mbit or so), so it's essentially impossible to have any signal problems, thus no need to spend money troubleshooting connections. They wanted something canned and easy (that's the ameritech way) and a small-radius low-bandwidth connection is the best way to do it. That's why you don't see a lot of places selling 6mbit dsl.
I think using a calculator is a horrible idea. I used it throughout my middle school and into highschool - but as i progressed later into highschool you realize how worthless it really is.
Any good test that's going to test your understanding of something isn't going to use numbers at all, so you better be _freaking_good_ at how the algebra works. I know ppl that got through factoring in alg1 by graphing the equation and finding the zeros. I'm sure they're doing real well now attempting to factor Legendre polynomials.
As time goes on formula memorization becomes worthless too (i haven't had a closed book test so far this semester), if you know a formula, you better know how to derive it (or at LEAST how to prove it's true), where it came from, and what it means. They do not emphasize this enough in middle/high school, but if you learn it, you'll be in great shape for absolutely anything you do (I can derive microeconomic formulas from F=ma, it's that powerful.)
> but rather refuted the underlying claim that
> practitioners could sense the energy fields of
> their subjects
Incorrect. The study provided evidence that it was probable (as in > 50% chance) that the claim was false, however was nowhere near the 90-95% level of significance which is generally accepted by acedemia as a level that provides 'interesting' and probable results.
If i remember correctly, her sample size was incredibly small (sub-100), leading to an incredibly large confidence interval, and no statistical significance.
Kudos to the PR people that pulled this off, but the study proved absolutely nothing. It did make JAMA, but having read the article i can definately say that although good for a 14 year old, was one of the poorest published statistical studies i have ever seen.
She made the news because of her age, in reality her study was far from perfect. The study ultimately was inconclusive at a decent level of signifiance, failed to protect itself from outside variables, and was obviously biased.
While it may have been 'good for her age', she really proved nothing, claiming she 'proved' something is just sensationalism. It did make a good 60 minutes piece however, because no one is going to go out and say 'oh, her work was worthless' on national tv.
Re:trying to compile this thing..
on
Low-Bandwidth X
·
· Score: 1
ldconfig
The linker keeps a cache of the libraries it has, you need to rebuild the cache when you add libraries.
On the topic of latency, which ppl brought up earlier, 802.11b is probably one of the scariest latent technologies. Given the vendor implementation of 802.11 (frequency hopping vs. fixed frequency, the latter is more popular because of easy of use), if it uses frequency hopping, you're looking at modem-caliber latencies, just to your firsthop (the access pt).
Having 1mbit (or 5.5, or 11, whatever you're using) of bandwidth per client isn't going to be the biggest factor in the performance of the device, the latency is going to be a very big issue.
>With all this.NET and remote X stuff, I'm >wondering, why would you want to run such things >remote?
Cost. Rather than buying N equal workstations, with N harddrives and N of every peripheral, you can buy one rather large server, then deploy diskless workstations with a network connection, then run all the apps on the server over the network. Sun has a product line called the Sun Rays that do it (and they've been doing it for a while, we have some 5 year old diskless workstations lying around too). It makes administration easier because you have one server to care about, rather than N workstations, and because all the cpu use is on the serverside, assuming you get decently beefy workstations, you'll never need to upgrade them.
Performance wise, on a 100mbit network, the latency isn't really an issue, it is tolerable, if even noticeable.
There also is that server thing, if you're administrating lots of servers, using a dtlogin or a terminal service makes things so much easier, instead of you having to move each server to your desk, or you doing work right next to the racks.
Actaully yes and no. Transaction support requires you to use BDB, install a patched BDB, and even the webpage has stated that this part of mysql is still in it's beta stages. As much as i love mysql and the idea of transaction support, using any beta software in a production environment is incredibly risky and i wouldn't do it.
>Tell him you understand that he may have a
>preference toward Access, but you know from past
>experience that Access will not meet your long-
>term needs and you recommend something different.
But also, does mysql really meet your needs? Mysql doesn't have stable transaction support (they're on b7..and i would never use beta *anything* in a production machine). Mysql also sucks when doing a lot of heavy inserts because of the table locking.
If you want to use mysql (maybe it is better in this situation) you should have a list of reasons on why it's better than mssql/access, not just 'because i want to'. Maybe your boss has a reason why access will do the job better.
You're the DBA, make the intelligent decision, not just mysql because it looks cool and is free.
Yeah. I'm not saying that what these guys are saying is a great idea at all. I frequently use lynx because i do most of my sysadmin work in virtualconsoles, so i hate having to boot X just to load a stupid webpage to get the docs for one small program/product.
> He didn't recieve a lick of scholorship money,
> though, and he needed the money, so he didn't go.
MIT scholarships are entirely need based, so that argument doesn't really hold. I know a lot of people with ~1300 SAT's that have made it into/are going to MIT, thankfully MIT also looks a LOT at teacher recs, which is really the best way to find a student's potential (though it's also the most time consuming from the admissions standpoint).
As working for an ISP, we were more than willing to offer copies on cd of Netscape and/or IE for our dialup clients, for whom a ~15mb download generally isn't an option. However when investigating licensing issues,we realized that redistribution was either a) subject to a lot of licensing issues which we could not agree with, or b) was cost-prohibitive.
Simply removing the red-tape on browser redistribution would solve a lot of the problems.
It wouldn't solve the problem of ppl with 386s running windows 3.1 or whatever, but really there isn't much of a solution for them.
hi. who said he couldn't have 'telnetted' in as a normal user, then buffer overflowed to get root?
Though i'm pretty sure this was the one where he got in from a suid httpd (which actaully is probably equally as retarded now, but...) no he did not telnet and login as root like the past 20 people are complaining about.
> In some places in there, they have bandwidth >that makes OC48 look like a dialup modem.
No. Try again. Much of MIT's campus network is old repeated 10mbit (repeated/hubbed btw, not even switched - collisions gallore). Having class B subnets doesn't help traffic much either.
They do not have an oc-48. They don't have a terribly lot of bandwidth, hell even their vBNS link is smaller than most. Furthermore their network is largely unregulated so most of the bandwidth is mostly taken up by the fservs and such.
These "network hubs" in academia you talk about do have a bit of bandwidth, but most of that is through abilene/vbns/i2/etc where it's inter-school (and with some gov labs) and certainly not to europe. And it's certainly not an oc-48. Commodity internet is expensive.
And the idea of any school having an oc-48 of commodity bandwidth is preposterous, uunet's backbone is largely oc-48 (and only between hubs, everything else is lower) the idea of mit (or anything in academia) having this kind of link is just stupid.
>A firewall would help, some, but not solve the >problem (FreeBSD ipfw cost $30 486 w/8-16mb ram >and 500 mb harddrive,).
No. First of all your crap based tulip card or even the 486 is even going to be remotely close to handling the amount of data going through.
And that's assuming it's ethernet i'll bet my lunch it's fddi, which throws the entire cost estimate out the window because now you need fddi cards. Not to mention the 486 would drop packets left and right.
Your 486 may be doing your masquerading for your home network friggen wel, but it's not going to work at mae-east (or whatever).
while a (useable, decent) firewall won't hurt, bgp filtering is probably the way to go.
Is it unfair? I'm inclined to say no. At least it's not anymore unfair than other ISP services.
A normal local ISP cannot afford the cost of a DSLAM, nor that can bare the cost of buying n+1 normal DSL 'modems'. In this respect, they have to rely on the phone company. Also the distance radius from the CO, w/o the DSLAM, the distance radius is reduced by the distance from CO->ISP office, which equates to less consumer base. I think having the phone company control the DSLAM makes sense. Considering the large startup cost, it's shares the cost among all the local ISPS sharing the DSLAM (and other related equipment, i dunno what's exactly involved). The high fixed yet low marginal costs of providing dsl makes this a logical idea.
I think the problem is that the phone company is venturing into the ISP market, which puts them in a really really wierd situtation. They want to both sell DSL to resellers (ie the local ISP) as well as themelves sell it to consumers. Imho, the only reason they're even allowing resellers into the picture is to get any monopolistic-hunting government people off their back and nothing more, especially when you take into account the 'discount' (or lackthereof) that they're giving to their resellers.
I don't see any alternative. DSl works the same way w/ any dedicated connection that you have w/ your local ISP (frac T1s or xdsl, whatever), the ISP is still at the whim of the phone company for any physical problems. Is the alternative to have the ISP control all of the dsl service. You don't want that, we have a HDSL running to an apt building in town, it goes down whenever it rains. The phone company says it's fine, the ISP is at the whim of the phone company, we're forced to play a waiting game until the line dries out, then it spontaneously starts working again. At least if you give the phone company some vested interest (ie if they own the DSLAM) they'll care a little more about whether or not it works.
Yes, if they had known that cat reads a file, or that the | redirects stdout to the stdin of the next command, then perhaps i would be proud of them for learning something.
At HS they decided to give everyone an email account (and webaccess). I'd say a good 80-90% (students and teachers) of people use it. Step into the computer lab at a given day, you'll see 1 (maybe two, if a paper is due) people working on homework, and about 12-15 people writing email. Of that 12-15 people, I'd say (purely guesstimate, but i don't think i'm that far off), 8-10 of them are reading and forwarding flame mail. If you ever need to type out a paper, good luck, it's almost impossible to get a computer during peak times. It's rediculous, all the 7th-8th graders (7-12grade is on one campus) sitting down at a computer email the person next to them.
They don't know how to use the computer, they know how to click on 'Connect to ', then type in their username and password. If all the macs were full, we'd get questions, 'how do i get to e-mail on this windows computer?' I've seen people write down directions, (Start->programs,etc) so they could get to it every time.
Then they learned how to ytalk to the person next to you. It got to be so disruptive we removed permissions, then they found the write command, and did cat | write . It was follow the leader, one person borrowed a book from the library, and found all the obnoxious things that could be done, and told everyone else about it. There was a printout of every obnoxious command a person found, he typed it out and distributed it out to everyone else. We finally through them in a custom unix shell, where you could enter pine, or finger. So now they sit around mailbombing each other with those windows mailbomber programs.
My point is, it's not going to work. Not for 7th graders, even up to 10th graders were doing these things and being generally obnoxious. It's quieted down some, but sitll people spend all their free time reading and forwarding chainmail. Current/var/mail size is 1.4G for ~1800 users (cat/etc/passwd | wc), while at the ISP i now work at,/var/mail is ~2G, for about ~6000 users.
From what I can tell, this is a purely middle-highschool phenominon. In college, there is the few people that still forward chainmail and the like, but it's few and far between. On the student cluster, i don't regularly get spammed w/ banners and lastlogs from other people using the 'write' cmd, even though they have perfect permissions to do so. 7th grade (even 9th imho), is way too early to be giving students net access while expecting them to use it responsibly.
So we give them a laptop/webpad/whatever, now what are they supposed to do with it? When i was highschool, i took a particularly hard courseload, and i had maybe say, 1-2 papers due a week. That translates into 2-3 nights actually banging at a keyboard with it. So they're going to spend $200+/person for a crippled computer to do a job that they could have easily done in a computer lab, at a fraction of the cost.
In order to deploy this program (or at least to convince me that it's a goodidea) they have to justify the cost, and that the benefits outweigh it. They have to make classes web-aware.
How easy is it to make a class web-aware? Easy. Scan in your assignment sheets, pdf it, do some adobe pagemill (or whatever), and *bam*, you have a webpage. Internet-aware? Yes. Useful? Marginally. Effective teaching tool? No.
I've seen web-based homework, and never found it effective. Sure, you can do multiple-choice, true/false questions easily (hell, i wrote a program that does that) but anything past that is very difficult, or has bad implementation. I particularly love my physics web-based homework, where i have to integrate some godawful function, and only get credit if i get the answer within 1%. Partial credit doesn't exist.
I had a Chem Teacher who felt this push to 'use the internet' from above, and came up with this (imho) stupid idea to put all our lab reports on the web. Neat, possibly. Useful to us? No. She's a great teacher, we did all our other work by conventional means, and we actually learned from it. She knew of this internet thing, and wanted to use it, but didn't know how. Our school administration really didn't give a flying fark what it was, as long as something was internet-based.
I'm not saying the internet can't work as learning tool, just that 1) i've never seen it yet, and 2) these schools certainly can't do it correct yet. Until the schools get the infrastructure set (train the teachers, do research, find what works) so that the computers would be an effective learning tool, giving every kid a laptop is frivolous.
Slashdot Mirror
Some adsl providers also impose caps downstream and don't offer higher than 1.5mbit service, mainly because it's too much trouble.
If you're a big business (*cough*cough*ameritech/sbc), the best way to make money with the least trouble is to make an aritifically small dsl radius, cap the lines at something way below the DMT standard (cap it at oh, 1.5mbit or so), so it's essentially impossible to have any signal problems, thus no need to spend money troubleshooting connections. They wanted something canned and easy (that's the ameritech way) and a small-radius low-bandwidth connection is the best way to do it. That's why you don't see a lot of places selling 6mbit dsl.
Any good test that's going to test your understanding of something isn't going to use numbers at all, so you better be _freaking_good_ at how the algebra works. I know ppl that got through factoring in alg1 by graphing the equation and finding the zeros. I'm sure they're doing real well now attempting to factor Legendre polynomials.
As time goes on formula memorization becomes worthless too (i haven't had a closed book test so far this semester), if you know a formula, you better know how to derive it (or at LEAST how to prove it's true), where it came from, and what it means. They do not emphasize this enough in middle/high school, but if you learn it, you'll be in great shape for absolutely anything you do (I can derive microeconomic formulas from F=ma, it's that powerful.)
> practitioners could sense the energy fields of
> their subjects
Incorrect. The study provided evidence that it was probable (as in > 50% chance) that the claim was false, however was nowhere near the 90-95% level of significance which is generally accepted by acedemia as a level that provides 'interesting' and probable results.
If i remember correctly, her sample size was incredibly small (sub-100), leading to an incredibly large confidence interval, and no statistical significance.
Kudos to the PR people that pulled this off, but the study proved absolutely nothing. It did make JAMA, but having read the article i can definately say that although good for a 14 year old, was one of the poorest published statistical studies i have ever seen.
While it may have been 'good for her age', she really proved nothing, claiming she 'proved' something is just sensationalism. It did make a good 60 minutes piece however, because no one is going to go out and say 'oh, her work was worthless' on national tv.
ldconfig
The linker keeps a cache of the libraries it has, you need to rebuild the cache when you add libraries.
Having 1mbit (or 5.5, or 11, whatever you're using) of bandwidth per client isn't going to be the biggest factor in the performance of the device, the latency is going to be a very big issue.
>wondering, why would you want to run such things
>remote?
Cost. Rather than buying N equal workstations, with N harddrives and N of every peripheral, you can buy one rather large server, then deploy diskless workstations with a network connection, then run all the apps on the server over the network. Sun has a product line called the Sun Rays that do it (and they've been doing it for a while, we have some 5 year old diskless workstations lying around too). It makes administration easier because you have one server to care about, rather than N workstations, and because all the cpu use is on the serverside, assuming you get decently beefy workstations, you'll never need to upgrade them.
Performance wise, on a 100mbit network, the latency isn't really an issue, it is tolerable, if even noticeable.
There also is that server thing, if you're administrating lots of servers, using a dtlogin or a terminal service makes things so much easier, instead of you having to move each server to your desk, or you doing work right next to the racks.
Actaully yes and no. Transaction support requires you to use BDB, install a patched BDB, and even the webpage has stated that this part of mysql is still in it's beta stages. As much as i love mysql and the idea of transaction support, using any beta software in a production environment is incredibly risky and i wouldn't do it.
>preference toward Access, but you know from past
>experience that Access will not meet your long-
>term needs and you recommend something different.
But also, does mysql really meet your needs? Mysql doesn't have stable transaction support (they're on b7..and i would never use beta *anything* in a production machine). Mysql also sucks when doing a lot of heavy inserts because of the table locking.
If you want to use mysql (maybe it is better in this situation) you should have a list of reasons on why it's better than mssql/access, not just 'because i want to'. Maybe your boss has a reason why access will do the job better.
You're the DBA, make the intelligent decision, not just mysql because it looks cool and is free.
Yeah. I'm not saying that what these guys are saying is a great idea at all. I frequently use lynx because i do most of my sysadmin work in virtualconsoles, so i hate having to boot X just to load a stupid webpage to get the docs for one small program/product.
> He didn't recieve a lick of scholorship money,
> though, and he needed the money, so he didn't go.
MIT scholarships are entirely need based, so that argument doesn't really hold. I know a lot of people with ~1300 SAT's that have made it into/are going to MIT, thankfully MIT also looks a LOT at teacher recs, which is really the best way to find a student's potential (though it's also the most time consuming from the admissions standpoint).
As working for an ISP, we were more than willing to offer copies on cd of Netscape and/or IE for our dialup clients, for whom a ~15mb download generally isn't an option. However when investigating licensing issues ,we realized that redistribution was either a) subject to a lot of licensing issues which we could not agree with, or b) was cost-prohibitive.
Simply removing the red-tape on browser redistribution would solve a lot of the problems.
It wouldn't solve the problem of ppl with 386s running windows 3.1 or whatever, but really there isn't much of a solution for them.
Though i'm pretty sure this was the one where he got in from a suid httpd (which actaully is probably equally as retarded now, but...) no he did not telnet and login as root like the past 20 people are complaining about.
>that makes OC48 look like a dialup modem.
No. Try again. Much of MIT's campus network is old repeated 10mbit (repeated/hubbed btw, not even switched - collisions gallore). Having class B subnets doesn't help traffic much either.
They do not have an oc-48. They don't have a terribly lot of bandwidth, hell even their vBNS link is smaller than most. Furthermore their network is largely unregulated so most of the bandwidth is mostly taken up by the fservs and such.
These "network hubs" in academia you talk about do have a bit of bandwidth, but most of that is through abilene/vbns/i2/etc where it's inter-school (and with some gov labs) and certainly not to europe. And it's certainly not an oc-48. Commodity internet is expensive.
And the idea of any school having an oc-48 of commodity bandwidth is preposterous, uunet's backbone is largely oc-48 (and only between hubs, everything else is lower) the idea of mit (or anything in academia) having this kind of link is just stupid.
>problem (FreeBSD ipfw cost $30 486 w/8-16mb ram
>and 500 mb harddrive,).
No. First of all your crap based tulip card or even the 486 is even going to be remotely close to handling the amount of data going through.
And that's assuming it's ethernet i'll bet my lunch it's fddi, which throws the entire cost estimate out the window because now you need fddi cards. Not to mention the 486 would drop packets left and right.
Your 486 may be doing your masquerading for your home network friggen wel, but it's not going to work at mae-east (or whatever).
while a (useable, decent) firewall won't hurt, bgp filtering is probably the way to go.
A normal local ISP cannot afford the cost of a DSLAM, nor that can bare the cost of buying n+1 normal DSL 'modems'. In this respect, they have to rely on the phone company. Also the distance radius from the CO, w/o the DSLAM, the distance radius is reduced by the distance from CO->ISP office, which equates to less consumer base. I think having the phone company control the DSLAM makes sense. Considering the large startup cost, it's shares the cost among all the local ISPS sharing the DSLAM (and other related equipment, i dunno what's exactly involved). The high fixed yet low marginal costs of providing dsl makes this a logical idea.
I think the problem is that the phone company is venturing into the ISP market, which puts them in a really really wierd situtation. They want to both sell DSL to resellers (ie the local ISP) as well as themelves sell it to consumers. Imho, the only reason they're even allowing resellers into the picture is to get any monopolistic-hunting government people off their back and nothing more, especially when you take into account the 'discount' (or lackthereof) that they're giving to their resellers.
I don't see any alternative. DSl works the same way w/ any dedicated connection that you have w/ your local ISP (frac T1s or xdsl, whatever), the ISP is still at the whim of the phone company for any physical problems. Is the alternative to have the ISP control all of the dsl service. You don't want that, we have a HDSL running to an apt building in town, it goes down whenever it rains. The phone company says it's fine, the ISP is at the whim of the phone company, we're forced to play a waiting game until the line dries out, then it spontaneously starts working again. At least if you give the phone company some vested interest (ie if they own the DSLAM) they'll care a little more about whether or not it works.
Yes, if they had known that cat reads a file, or that the | redirects stdout to the stdin of the next command, then perhaps i would be proud of them for learning something.
They don't know how to use the computer, they know how to click on 'Connect to ', then type in their username and password. If all the macs were full, we'd get questions, 'how do i get to e-mail on this windows computer?' I've seen people write down directions, (Start->programs,etc) so they could get to it every time.
Then they learned how to ytalk to the person next to you. It got to be so disruptive we removed permissions, then they found the write command, and did cat | write . It was follow the leader, one person borrowed a book from the library, and found all the obnoxious things that could be done, and told everyone else about it. There was a printout of every obnoxious command a person found, he typed it out and distributed it out to everyone else. We finally through them in a custom unix shell, where you could enter pine, or finger. So now they sit around mailbombing each other with those windows mailbomber programs.
My point is, it's not going to work. Not for 7th graders, even up to 10th graders were doing these things and being generally obnoxious. It's quieted down some, but sitll people spend all their free time reading and forwarding chainmail. Current /var/mail size is 1.4G for ~1800 users (cat /etc/passwd | wc), while at the ISP i now work at, /var/mail is ~2G, for about ~6000 users.
From what I can tell, this is a purely middle-highschool phenominon. In college, there is the few people that still forward chainmail and the like, but it's few and far between. On the student cluster, i don't regularly get spammed w/ banners and lastlogs from other people using the 'write' cmd, even though they have perfect permissions to do so. 7th grade (even 9th imho), is way too early to be giving students net access while expecting them to use it responsibly.
In order to deploy this program (or at least to convince me that it's a goodidea) they have to justify the cost, and that the benefits outweigh it. They have to make classes web-aware.
How easy is it to make a class web-aware? Easy. Scan in your assignment sheets, pdf it, do some adobe pagemill (or whatever), and *bam*, you have a webpage. Internet-aware? Yes. Useful? Marginally. Effective teaching tool? No.
I've seen web-based homework, and never found it effective. Sure, you can do multiple-choice, true/false questions easily (hell, i wrote a program that does that) but anything past that is very difficult, or has bad implementation. I particularly love my physics web-based homework, where i have to integrate some godawful function, and only get credit if i get the answer within 1%. Partial credit doesn't exist.
I had a Chem Teacher who felt this push to 'use the internet' from above, and came up with this (imho) stupid idea to put all our lab reports on the web. Neat, possibly. Useful to us? No. She's a great teacher, we did all our other work by conventional means, and we actually learned from it. She knew of this internet thing, and wanted to use it, but didn't know how. Our school administration really didn't give a flying fark what it was, as long as something was internet-based.
I'm not saying the internet can't work as learning tool, just that 1) i've never seen it yet, and 2) these schools certainly can't do it correct yet. Until the schools get the infrastructure set (train the teachers, do research, find what works) so that the computers would be an effective learning tool, giving every kid a laptop is frivolous.