After doing a lot of reading of the theory of P2P as a legitimate process in computing, I've come up with a plan:
Some enterprising programmer should consider writing a P2P program that lets each client system create a unique backup that every other client has a piece of, like a big worldwide RAID 5 or something. Given the ratio of (non-application stuff) data to drive space, this might work.
he's right...a t3 requires connection equipment, and let's say you're gonna try and set up a VLAN or something...that requires copper, dry line, whatever, that costs...by the month.
You need to pay someone a good wage to administer this, in fact several people to deal with complaints when suddenly everyone IS downloading the new linux ISO or new movie or whatever.
Say 3 support guys at least 25k a year for them, rental fees for dry lines...don't know, not terribly cheap say 2k for the year, equipment, electricity, maintenance...it goes on...bigger than you thought huh?
small ISPs can't really afford to be competitive against the big'uns, especially if they're about to get a big influx of customers trying to avoid the download caps from their existing ISP.
Someone still winds up stuck with the bill being sent out from a backbone.
Fair enough, let's not control what you do, but if you want real bandwidth, then let's talk.
Here's what we do at our ISP:
Normally we charge only for the bandwidth used. Our calculation for burstable connections is as follows: Samples are taken on your port approximately once a minute.
Each month the incoming and outgoing samples are sorted in ascending order. The 95th percentile entry in each list is examined. The higher of the incoming or outgoing figures is your monthly Burst rate.
And we charge you by your burst rate. This is fair policy, but you may not care for what the cost of your actual burst is.
Think on the order of $1000 per megabit total burst rate. How does your 1.5Mb RESIDENTIAL ADSL sound now?
This is an example of the problems of the broadband business model.
If you really want 512/256 service, expect to pay what we pay when we hook it up with a backbone.
And sure you have a right to do what we want, but you have to remember that it is quite easy for a customer to cost the ISP more money than they're making on your invoices. Then it doesn't exactly hurt to have you leave.
An old argument to be sure, but email attachments are bigger than they ought to be because of the MIME band-aid protocol.
Perhaps someone (read: not me) should draft a new mail transfer protocol that doesn't require binary files be converted into ASCII at a tremendous use of CPU and bandwidth.
my2c
no no no...it wasn't really done properly until the fifth element.
EVERYBODY knows you can't make robots that small, it's just silly...but we can surely mind control existing insects to do our bidding...that's the appROACH we need to take.
Slashdot Mirror
it's not in geosynchronous orbit, it's in low earth orbit...relatively reachable, but a pain in the ass nonetheless
This isn't really a bad idea.
After doing a lot of reading of the theory of P2P as a legitimate process in computing, I've come up with a plan:
Some enterprising programmer should consider writing a P2P program that lets each client system create a unique backup that every other client has a piece of, like a big worldwide RAID 5 or something. Given the ratio of (non-application stuff) data to drive space, this might work.
without newton's writings, I can describe the universe however I like and not be subject to any scathing editorials about 'consistency' jeez
he's right...a t3 requires connection equipment, and let's say you're gonna try and set up a VLAN or something...that requires copper, dry line, whatever, that costs...by the month.
You need to pay someone a good wage to administer this, in fact several people to deal with complaints when suddenly everyone IS downloading the new linux ISO or new movie or whatever.
Say 3 support guys at least 25k a year for them, rental fees for dry lines...don't know, not terribly cheap say 2k for the year, equipment, electricity, maintenance...it goes on...bigger than you thought huh?
small ISPs can't really afford to be competitive against the big'uns, especially if they're about to get a big influx of customers trying to avoid the download caps from their existing ISP.
Someone still winds up stuck with the bill being sent out from a backbone.
Fair enough, let's not control what you do, but if you want real bandwidth, then let's talk.
Here's what we do at our ISP:
Normally we charge only for the bandwidth used. Our calculation for burstable connections is as follows: Samples are taken on your port approximately once a minute.
Each month the incoming and outgoing samples are sorted in ascending order. The 95th percentile entry in each list is examined. The higher of the incoming or outgoing figures is your monthly Burst rate.
And we charge you by your burst rate. This is fair policy, but you may not care for what the cost of your actual burst is.
Think on the order of $1000 per megabit total burst rate. How does your 1.5Mb RESIDENTIAL ADSL sound now?
This is an example of the problems of the broadband business model.
If you really want 512/256 service, expect to pay what we pay when we hook it up with a backbone.
And sure you have a right to do what we want, but you have to remember that it is quite easy for a customer to cost the ISP more money than they're making on your invoices. Then it doesn't exactly hurt to have you leave.
An old argument to be sure, but email attachments are bigger than they ought to be because of the MIME band-aid protocol. Perhaps someone (read: not me) should draft a new mail transfer protocol that doesn't require binary files be converted into ASCII at a tremendous use of CPU and bandwidth. my2c
good stuff...needs to be more than a 1
no no no...it wasn't really done properly until the fifth element. EVERYBODY knows you can't make robots that small, it's just silly...but we can surely mind control existing insects to do our bidding...that's the appROACH we need to take.