The way to tell is to measure how long it takes for the sysadmin to a) notice that it runs sendmail and b) changes it for something else. Personally I use exim, but just about anything is better than sendmail.
Having said that: I would not touch qmail with a bargepole either.
Which question were people answering? Are you intending to use an "embedded linux" or just "linux" in a small standalone device? We use a "standard" linux in several standalone devices. We have no need for, nor do we want to use a "specialist" distribution because we do want to be locked in. It is no coincidence that one of the more fertile areas of cpu support development in the kernel, at the moment, is for ARM devices.
Ofcom in the UK have measured the emissions from the PLT (as it is called here) system trials being held here. The results are here on the Ofcom website.
Basically they show just how noisy these signals are as well as spectrum trashing ability they have. It appears that, even at 100m from the powerline the systems exceed the German emission standards (the only applicable ones that the EU has currently) by a wide margin.
It is unlikely that PLT will be licenced in the UK with these characteristics.
As usual, the practicalities of actual use are not stated in the article. In a multiuser, single frequency, satellite system such as this *may* be, the problems are formidable. They are also quite well researched (see Computer Networks by Prof Tanenbaum). Hopefully any practical system would have more than one channel with separate uplink and downlink frequencies.
Every consumer oriented satellite system has, so far, failed (such as the one touted by BT here in the UK before ADSL became more wide spread) because of the inherent problems like latency induced lack of syncronisation, 'hidden stations' and the resultant exponential decline in throughput with added users.
These are made much worse by the fact that there is a large latency which tends to affect the more efficient packet protocols more (eg slotted aloha).
The only advantage this system has is that at least it starts at 100Mbits raw speed. But since theory suggests that, without some serious cleverness, you will only be able to utilise about 18% of that. In the limit, how much use will that be in 2015?
And we have solved all the hidden station problems have we? Shannon has been found to be tosh? Nyquist a mathematical bungler?
Yes, you may get 40 Mbps if you are close enough, both sides are running enough (legal) power and you are the only two ends on the channel without any interference of any kind. Note the word "may".
Look, there are some fundamental problems with using wireless for network connectivity. They are to do with the medium, the nature of transmitters and receivers (and the delays inherent), the lack of bandwidth (because we all want two pieces and there is way not enough to go around) and then there is terrain (which helps for frequency re-use but in all other ways doesn't) and the fact that, even in an ideal world, more than two ends to a conversation on a channel degrades the experience geometrically for everyone when you start to add more users.
In the home / small office for a few units not doing much simultanious work then fine. But please, don't expect anything even approaching some of the figures people quote because, in practice you won't ever get them (and that's assuming WiMAX is full duplex which 802.11[b-g] aren't).
Remember all those wonderful thruput figures for GPRS, 3G etc. Notice how they have all come down to numbers that make ISDN look fast? Do yourself a favour: get yourself an erlang, use a piece of wet string (or wire or fibre) or better still - two.
Never mind whether the thing will fly. Nor whether the radius of the coverage area is 5, 75 or 305 miles. The crucial question with any radio based system is: "how many erlangs can it achieve"?
An Erlang is the name for a full-duplex [telephone] connection. It is a unit of [simultaneous] capacity for a [telephone] system.
The problem you have with [all] wireless systems is the limited bandwidth available and the fact that you run out of it very, very quickly. This is why the cellular approach is so effective (believe it or not). The idea of cellular is that you re-use channels by carefully siting small, low power base stations then assigning your limited pool of frequencies in a specific way so that, with care, you have a scalable system.
In principle, if you need more capacity, you can extend a cellular system by sub dividing cells, using even lower power base stations and replicating the frequency assignment plan.
Any air/space borne system that covers huge areas will have a problem with capacity. There are not enough frequencies and thus channels available to make such a system viable - and that is before you address issues like being obscured (eg in a building).
Slashdot Mirror
The way to tell is to measure how long it takes for the sysadmin to a) notice that it runs sendmail and b) changes it for something else. Personally I use exim, but just about anything is better than sendmail.
Having said that: I would not touch qmail with a bargepole either.
Which question were people answering? Are you intending to use an "embedded linux" or just "linux" in a small standalone device? We use a "standard" linux in several standalone devices. We have no need for, nor do we want to use a "specialist" distribution because we do want to be locked in. It is no coincidence that one of the more fertile areas of cpu support development in the kernel, at the moment, is for ARM devices.
The IP address of the server is in Amazon's netblock and the webserver is obviously also an (just one) amazon machine:-
host mturk.amazon.com
mturk.amazon.com is an alias for rewrite.amazon.com.
rewrite.amazon.com has address 207.171.163.18
whois 207.171.163.18
OrgName: Amazon.com, Inc.
OrgID: AMAZON-4
Address: 605 5th Ave S
City: SEATTLE
StateProv: WA
PostalCode: 98104
Country: US
NetRange: 207.171.160.0 - 207.171.191.255
CIDR: 207.171.160.0/19
NetName: AMAZON-01
NetHandle: NET-207-171-160-0-1
Parent: NET-207-0-0-0-0
NetType: Direct Assignment
NameServer: NS-1.AMAZON.COM
NameServer: NS-2.AMAZON.COM
NameServer: NS-3.AMAZON.COM
NameServer: AUTH00.NS.UU.NET
Comment:
RegDate: 1999-09-23
Updated: 2002-03-19
RTechHandle: AC6-ORG-ARIN
RTechName: Amazon.com, Inc.
RTechPhone: +1-206-266-2187
RTechEmail: NOC@amazon.com
OrgTechHandle: ROLEA19-ARIN
OrgTechName: Role Account
OrgTechPhone: +1-206-266-2187
OrgTechEmail: noc@amazon.com
# ARIN WHOIS database, last updated 2005-11-03 19:10
# Enter ? for additional hints on searching ARIN's WHOIS database.
Seems legit, if rather strange branding, to me.
Dirk
Ofcom in the UK have measured the emissions from the PLT (as it is called here) system trials being held here. The results are here on the Ofcom website.
Basically they show just how noisy these signals are as well as spectrum trashing ability they have. It appears that, even at 100m from the powerline the systems exceed the German emission standards (the only applicable ones that the EU has currently) by a wide margin.
It is unlikely that PLT will be licenced in the UK with these characteristics.
Every consumer oriented satellite system has, so far, failed (such as the one touted by BT here in the UK before ADSL became more wide spread) because of the inherent problems like latency induced lack of syncronisation, 'hidden stations' and the resultant exponential decline in throughput with added users.
These are made much worse by the fact that there is a large latency which tends to affect the more efficient packet protocols more (eg slotted aloha).
The only advantage this system has is that at least it starts at 100Mbits raw speed. But since theory suggests that, without some serious cleverness, you will only be able to utilise about 18% of that. In the limit, how much use will that be in 2015?
Sigh....
And we have solved all the hidden station problems have we? Shannon has been found to be tosh? Nyquist a mathematical bungler?
Yes, you may get 40 Mbps if you are close enough, both sides are running enough (legal) power and you are the only two ends on the channel without any interference of any kind. Note the word "may".
Look, there are some fundamental problems with using wireless for network connectivity. They are to do with the medium, the nature of transmitters and receivers (and the delays inherent), the lack of bandwidth (because we all want two pieces and there is way not enough to go around) and then there is terrain (which helps for frequency re-use but in all other ways doesn't) and the fact that, even in an ideal world, more than two ends to a conversation on a channel degrades the experience geometrically for everyone when you start to add more users.
In the home / small office for a few units not doing much simultanious work then fine. But please, don't expect anything even approaching some of the figures people quote because, in practice you won't ever get them (and that's assuming WiMAX is full duplex which 802.11[b-g] aren't).
Go an read up on it. Computer Networks by Andrew S Tanenbaum is probably a good place to start.
Remember all those wonderful thruput figures for GPRS, 3G etc. Notice how they have all come down to numbers that make ISDN look fast? Do yourself a favour: get yourself an erlang, use a piece of wet string (or wire or fibre) or better still - two.
Er.. I believe the article was on the BBC which AFAIK is an organ of your former colonial masters circa 1776.
Never mind whether the thing will fly. Nor whether the radius of the coverage area is 5, 75 or 305 miles. The crucial question with any radio based system is: "how many erlangs can it achieve"?
An Erlang is the name for a full-duplex [telephone] connection. It is a unit of [simultaneous] capacity for a [telephone] system.
The problem you have with [all] wireless systems is the limited bandwidth available and the fact that you run out of it very, very quickly. This is why the cellular approach is so effective (believe it or not). The idea of cellular is that you re-use channels by carefully siting small, low power base stations then assigning your limited pool of frequencies in a specific way so that, with care, you have a scalable system.
In principle, if you need more capacity, you can extend a cellular system by sub dividing cells, using even lower power base stations and replicating the frequency assignment plan.
Any air/space borne system that covers huge areas will have a problem with capacity. There are not enough frequencies and thus channels available to make such a system viable - and that is before you address issues like being obscured (eg in a building).