SpamAssassin, what you mean the filter with a rule that says CommuniGate is spam software? Doh. By that reasoning, they should also have rules for sendmail, qmail, postfix, exchange, [...]
The other possibility is a net-block equivalent of ORBS. Some on the Sec-Focus Incidents list (and other fora, over the years) have bounced around the idea of blocking netoblocks who'#s POCs don't work, or who don't have or respond to mail to the RFC-mandated abuse@, security@, hostmaster@,.. standard mail accounts.
IMO domain names should be revoked unless at least the postmaster@ is verified working and being listened to, maybe by pinging on a random basis with a key that must be returned...that way, blocking all mail from domains that don't resolve would actually be quite effective.
So, blackhole-route or firewall it rather than rejecting mails. Lots of buggy mail sending software will repeatedly hammer your server if it receives certain SMTP reject codes.
not quite - if you're using an OS which HLTs the CPU at idle (Linux, BSD, NT, or win9x with CPUidle/Rain/Waterfall/[...]), you will be using more power by having the CPU do work. if s@h causes you not to use suspend mode, you will also be using more power.
might be that the aliens will come and visit and give us some help before global warming is too much of a problem, you never know, but I think it's just a little bit unfriendly to other inhabitants of this planet to use more power than you need, unless you are sure it comes from a clean source...
the more people participating, the more new people will get introduced to the project - for example, friends seeing the screensaver and asking about it, etc.
Of course, the good magazines eventually started putting checksums through the data, so you could tell which line the mistakes were in. What fantastic fun it was when they just had one big checksum for the whole lot and it just said 'data wrong, check typing'...
Would have been a lot nicer if they'd used base64...
Wouldn't it possible to initiate an SMTP transaction and then abort that transaction just before the email was actually sent, while still verifying that email could be sent?
You can't verify whether an email would be sent without sitting and waiting to see if it is actually sent. Given that you might be trying to test a server which is open relay and being abused by spammers, you might have to wait quite some time for that message to actually get delivered.
Some mail servers reject unauthorised relaying at the SMTP port. Others bounce the mail. Others just drop the mail.
Also, by starting and aborting the connection, on some fairly common configurations you would cause noise to the sysadmin/postmaster (either in logs or by email - some mail server software goes out of its way to log errors in the SMTP transaction, which can be quite helpful if people are having problems sending mail).
In addition I think that exposure to non-linear onscreen editing at an early age may cause some difficulties with normal speech (although it might have benefits in other areas).
FreeBSD's natd can rewrite to a specific address rather than the named address
Right, I'm doing that at the moment.
but there was nothing in my cursory inspection that said it could do so when operating in pure bridged mode.
Hmm, maybe still worth playing with then - I haven't played much with pure bridging yet. Still, at least on *BSD we have securelevel=3 which at least requires activity at the console to make any changes (assuming all scripts and binaries run before the securelevel is raised to 3 have been set "chflags schg").
It's quite useful to log both on and off the box (even better if you set it up so the off-box logging is reasonably invisible on the firewall), and compare the logs between the two for differences.
WAP11 tuning can (and should) be done a bit more carefully than just opening up the SNMP utility and typing '80' in all the boxes.
Looking at different values and monitoring with wlanexpert
I see that on my WAP11s, near the factory setting the adjustment is very sensitive (i.e. small change in CR31 = large change in signal strength). The 20-30 values around it (maybe something like B0-C8 on the AP I have been testing) account for about 7-8dBm of difference.
CR31 settings outside this range have much
less effect on signal strength - perhaps 1-2dBm.
I would be interested to know how clean the output
is when the amplifier is set to the lowest amount (i.e. highest CR31 value) for the maximum signal strength measured.
I assume that above this value there will be a lot of distortion. (I'm not an RF engineer and would appreciate comments from anyone who is, but I assume it is similar to audio amplification - if so, imagine you have an amplifier and the inputs are turned up much louder than can be handled - the output doesn't get louder, it just gets more and more distorted. I assume that the situation here is similar.)
The question I would like to have answered is, at this value, is there still a serious amount of power into the sidebands? (Answering this requires access to a spectrum analyser - so this is just a question not a suggestion! Still, setting like this is at least not likely to cause worse problems than setting at 80, and isn't going to reduce the range).
Values below 80 react quite strangely, I didn't test very much since I found many values reducing power below the card's sensitivity (so I had to run up and down several flights of stairs to reset CR31 from the wired lan, which was very good exercise!). So...
People who want to reduce the power output to the
minimum, possibly to keep the footprint of their WLAN as low as possible maybe to avoid interfering with neighbours, or so that passers-by are less likely to stumble across it, should definitely try different values below 80 as well as above 80 - at least on my boxes <80 is not a mirror of >80. (and use carefully positioned carefully chosen antennas, turn off SSID broadcasts, enable WEP, etc.)
I hope that everybody noted their default settings
before modifying CR31;-) My two boxes (bought at the same time) came set to...
So this definitely seems to be done per-unit and not per-batch. (And, these are different to figures I've seen quoted in mailing list posts).
Presumably they are factory-tuned for the best trade-off between good range and a clean signal, without putting too much power into the sidebands, and probably with a safety margin so that this remains true while the unit ages and if it's operated in different temperatures (electronic components are not at exactly the rated value, they are usually within a certain tolerance, the software setting is to account for this - in other designs this might be done using, for example, variable resistors). And obviously the factory settings will be tuned to ensure that the unit is within FCC limits (for example, ensuring that transmissions stay within the ISM band so you're not broadcasting into licensed bands without a license, which you might be if you adjust CR31 without testing with proper equipment or filtering to remove out-of-band transmissions).
5GHz (802.11a) is blocked more quickly by obstacles than 2.4GHz (802.11b/g). Also the radio design of individual cards and access points plays a big part - so the fact that 2.4GHz parts have had chance for designs to mature may be a factor too.
It will be interesting to see how OFDM does in the 2.4GHz band (and don't forget that CCK isn't the only coding specified by 802.11b, PBCC is also specified as an option at 5.5 and 11Mbps, and although this is not widely available yet, it gives better range than CCK).
The APs already include the required 'tertiary' (host-ap) firmware on the PC cards.
However this is not the firmware on 'normal' PC cards that would be used as wireless clients and might be used with homebrew hardware.
Normal cards have 'primary' (ap-client) firmware not host-AP firmware (which is owned by Intersil and subject to NDA). Some code is floating around on the net but of dubious heritage and probably best avoided for any commercial projects:)
Surely stopping illegal commerce in a certain currency would just mean that some people would use another for some of their business (thus decreasing value of the one and increasing value of the other, presumably USD, which doesn't have such restrictions)..?
What they normally use is a database with all the addresses which don't have a TV licence.
...combined with a list of people who've bought televisions/videos/satellite receivers/tv cards, and given their real address (it's a requirement for shops to request details and pass them on to tv licensing otherwise they can be fined), plus presumably address lists from subscription services (sky/etc).
£10 discount for blind people is a discount of about a third on the £35 a black-and-white license costs. (Unless for some strange reason they're using a colour tv, such as maybe so that non-blind people living in the same house can watch programs in colour, in which case the charge is really kind-of fair enough).
Mind you, they could always just buy a radio covering the relevant bands and pay nothing... (Though it would probably be easier to pay to avoid having to convince licensing officials they really don't have a tv - that's probably one of the worst things about the licensing system).
Though, some of those blocks which disappear shortly after spam has been sent from them just *might* have been RBL'd, which would of course cause them to disappear from above.net, teleglobe etc...
Slashdot Mirror
So, blackhole-route or firewall it rather than rejecting mails. Lots of buggy mail sending software will repeatedly hammer your server if it receives certain SMTP reject codes.
might be that the aliens will come and visit and give us some help before global warming is too much of a problem, you never know, but I think it's just a little bit unfriendly to other inhabitants of this planet to use more power than you need, unless you are sure it comes from a clean source...
Would have been a lot nicer if they'd used base64 ...
What, just like you don't have to do with 'diskcopy a: b:' from a DOS window?
You can't verify whether an email would be sent without sitting and waiting to see if it is actually sent. Given that you might be trying to test a server which is open relay and being abused by spammers, you might have to wait quite some time for that message to actually get delivered. Some mail servers reject unauthorised relaying at the SMTP port. Others bounce the mail. Others just drop the mail.
Also, by starting and aborting the connection, on some fairly common configurations you would cause noise to the sysadmin/postmaster (either in logs or by email - some mail server software goes out of its way to log errors in the SMTP transaction, which can be quite helpful if people are having problems sending mail).
In addition I think that exposure to non-linear onscreen editing at an early age may cause some difficulties with normal speech (although it might have benefits in other areas).
Wow, that's enormous ...... (bet it sounds good!)
You've not read the cunning plan in Cryptonomicon then...
It's quite useful to log both on and off the box (even better if you set it up so the off-box logging is reasonably invisible on the firewall), and compare the logs between the two for differences.
Surely you can NAT using an address which is not assigned to an interface?
Looking at different values and monitoring with wlanexpert I see that on my WAP11s, near the factory setting the adjustment is very sensitive (i.e. small change in CR31 = large change in signal strength). The 20-30 values around it (maybe something like B0-C8 on the AP I have been testing) account for about 7-8dBm of difference.
CR31 settings outside this range have much less effect on signal strength - perhaps 1-2dBm.
I would be interested to know how clean the output is when the amplifier is set to the lowest amount (i.e. highest CR31 value) for the maximum signal strength measured.
I assume that above this value there will be a lot of distortion. (I'm not an RF engineer and would appreciate comments from anyone who is, but I assume it is similar to audio amplification - if so, imagine you have an amplifier and the inputs are turned up much louder than can be handled - the output doesn't get louder, it just gets more and more distorted. I assume that the situation here is similar.)
The question I would like to have answered is, at this value, is there still a serious amount of power into the sidebands? (Answering this requires access to a spectrum analyser - so this is just a question not a suggestion! Still, setting like this is at least not likely to cause worse problems than setting at 80, and isn't going to reduce the range).
Values below 80 react quite strangely, I didn't test very much since I found many values reducing power below the card's sensitivity (so I had to run up and down several flights of stairs to reset CR31 from the wired lan, which was very good exercise!). So...
I hope that everybody noted their default settings before modifying CR31 ;-) My two boxes (bought at the same time) came set to ...
c7-c7-c7-c7-c5-c3-c1-c1-bf-bf-bf-bf-bf-be
c7-c7-c7-c7-c7-c5-c3-c3-c1-c1-c1-c1-c1-c1
So this definitely seems to be done per-unit and not per-batch. (And, these are different to figures I've seen quoted in mailing list posts).
Presumably they are factory-tuned for the best trade-off between good range and a clean signal, without putting too much power into the sidebands, and probably with a safety margin so that this remains true while the unit ages and if it's operated in different temperatures (electronic components are not at exactly the rated value, they are usually within a certain tolerance, the software setting is to account for this - in other designs this might be done using, for example, variable resistors). And obviously the factory settings will be tuned to ensure that the unit is within FCC limits (for example, ensuring that transmissions stay within the ISM band so you're not broadcasting into licensed bands without a license, which you might be if you adjust CR31 without testing with proper equipment or filtering to remove out-of-band transmissions).
Although in Europe, where I am, at the moment 0W power is permitted for 802.11a in either band ;-)
It will be interesting to see how OFDM does in the 2.4GHz band (and don't forget that CCK isn't the only coding specified by 802.11b, PBCC is also specified as an option at 5.5 and 11Mbps, and although this is not widely available yet, it gives better range than CCK).
However this is not the firmware on 'normal' PC cards that would be used as wireless clients and might be used with homebrew hardware.
Normal cards have 'primary' (ap-client) firmware not host-AP firmware (which is owned by Intersil and subject to NDA). Some code is floating around on the net but of dubious heritage and probably best avoided for any commercial projects :)
Perhaps this may be part of the reason why there's an agreement to continue using DM (search for 'emerged').
Mind you, they could always just buy a radio covering the relevant bands and pay nothing... (Though it would probably be easier to pay to avoid having to convince licensing officials they really don't have a tv - that's probably one of the worst things about the licensing system).
It really *really* sucks that DVD players don't have a 15-pin VGA connector on them though.
Though, some of those blocks which disappear shortly after spam has been sent from them just *might* have been RBL'd, which would of course cause them to disappear from above.net, teleglobe etc...