Issuing a 550 will typically generate a bounce message, yes, but it's not YOUR machine generating the bounce message, it's the MTA that's currently talking to YOUR MTA.
If it's the spammer (or, say, a compromised machine) sending to you, then it's likely that no bounce is detected. Instead its POSSIBLE that your name will be removed from a list because the sending MTA thinks that a) the address doesn't work anymore, or b) you're using something like spamcop, making your address 'poison'. (see below)
If it's a open relay talking to you, then the sending MTA will generate a bounce and attempt to send it to the 'envelope from'. This will also likely bounce, and then the message goes to the open relay administrator's mailbox. And that, in turn, should give that admin a real big hint that their server's an open relay.
(I've found that since using spamcop, the number of spams that are sent to me - before the filters - has actually gone down. I suspect that some spammers try to determine which addresses are likely to cause a spamcop notification, and avoid those addresses... pure speculation, but it's the only one that could explain the drop in spams, rather than gain)
Even before sitefinder, Verisign were well able to collect typos by modifying the DNS server to record any query on domains that don't exist. If something got hit a lot, they could just 'squat the domain' themselves, and since they're a registrar, they don't have to pay any money to do so.
This is why I feel that registrars need to pay a small stipend for every domain they themselves host, perhaps to the IETF or some other body. THis way, they're not encouraged to give themselves free domains, and it can pay for a bit of network infrastructure for the common good.
Although HP's disclaimer to indemnify your machine if you make modifications to the Linux kernel yourself, I think HP had every right to do this.
After all, what would stop you INTENTIONALLY adding in copyright SysV code to the kernel, and then asking HP to legally protect you from something that you should be responsible for yourself.
This is, of course, pointed out in the NewsForge article, but I doubt that more than half of./ readers will get that far into the article:)
I don't think that's relevant. In most circumstances, you ALREADY KNOW the sizes of your vectors, since they're fixed by the type of algorithm you're programming.
In other words, in these 'most cases' each position in the vector is NOT a new sample in a dataset, but rather a new variable, or axis of variability, in an equation.
Oh... I never meant to imply that I thought this was a GOOD thing. I'm just trying to make sense of the Verisign rep's comments, and figure out WHAT he could possibly mean by 'it'll only redirect web queries'.
Doing the redirect thing on www. domains is the only thing I could think of.
IT could very well be that they're saying that queries for www.sometyponame.com will return an IP address, but sometyponame.com will return a negative result.
The clearly the best mouse ever was the hemispherical, red-clown-nose mouse that came with the AT&T/Teletype 5620 terminal. What a buzz using that thing was:)
If you're on web page A, click on a link to B and it redirects to C, some browsers will, when fetching C, have a referrer of A, and some will have a referrer of B.
Wow, taken to the extreme, the exploitation of their systems could have caused a train collision and injury or death to hundreds of Maryland and Virginia commuters.
Will Microsoft's new 'Indemnification push' cover the legal costs for something like this, I wonder?:)
It seems strange, though, that they still have all their glitzy web pages advertising the SCO Forum, but there's one tiny little part that says the forum is postponed. Here's my analysis:
1. It was some kind of crack.
2. It really is postponed, But they don't want to advertise the fact, both for the obvious reasons of negative PR, but also because they want all the anti-SCO crowd to turn up anyway. They've probably emailed all the people who have actually registered to make sure they're not cheesed off. However, this will seriously impact the people who legitimately wanted to turn up, but were going to sign-in on the day.
Slashdot Mirror
Issuing a 550 will typically generate a bounce message, yes, but it's not YOUR machine generating the bounce message, it's the MTA that's currently talking to YOUR MTA.
If it's the spammer (or, say, a compromised machine) sending to you, then it's likely that no bounce is detected. Instead its POSSIBLE that your name will be removed from a list because the sending MTA thinks that a) the address doesn't work anymore, or b) you're using something like spamcop, making your address 'poison'. (see below)
If it's a open relay talking to you, then the sending MTA will generate a bounce and attempt to send it to the 'envelope from'. This will also likely bounce, and then the message goes to the open relay administrator's mailbox. And that, in turn, should give that admin a real big hint that their server's an open relay.
(I've found that since using spamcop, the number of spams that are sent to me - before the filters - has actually gone down. I suspect that some spammers try to determine which addresses are likely to cause a spamcop notification, and avoid those addresses... pure speculation, but it's the only one that could explain the drop in spams, rather than gain)
I think you're mistaken. When he says 'bounce spam' he doesn't mean composing a new message and sending it to the 'envelope from'.
He means ensuring the spam message gets a 550 code, or something similiar, rather than 'accepting' it and trashing it later.
The problem with intentionally removing support for one OS is that you're never really sure how many OTHER OS's it's going to effect.
A better effect would be to announce that no MORE support will be done for a particular OS, and just let it atrophy.
Even before sitefinder, Verisign were well able to collect typos by modifying the DNS server to record any query on domains that don't exist. If something got hit a lot, they could just 'squat the domain' themselves, and since they're a registrar, they don't have to pay any money to do so.
This is why I feel that registrars need to pay a small stipend for every domain they themselves host, perhaps to the IETF or some other body. THis way, they're not encouraged to give themselves free domains, and it can pay for a bit of network infrastructure for the common good.
However, if you look at SCO's stock price over the past 24 hours, the indemnification and/or SCO's response to it has done SCO more harm than good.
Hooray!
Absolutely!
The disclaimer is mostly certainly a CYA move by HP, but it's a SENSIBLE CYA move.
Although HP's disclaimer to indemnify your machine if you make modifications to the Linux kernel yourself, I think HP had every right to do this.
./ readers will get that far into the article :)
After all, what would stop you INTENTIONALLY adding in copyright SysV code to the kernel, and then asking HP to legally protect you from something that you should be responsible for yourself.
This is, of course, pointed out in the NewsForge article, but I doubt that more than half of
I don't think that's relevant. In most circumstances, you ALREADY KNOW the sizes of your vectors, since they're fixed by the type of algorithm you're programming.
In other words, in these 'most cases' each position in the vector is NOT a new sample in a dataset, but rather a new variable, or axis of variability, in an equation.
Not to put a too fine a point on it, No :)
ILM is Lucas
Dreamworks is Spielberg
Except for the fact that using the 1000MB=1GB definition is the accurate one.
Actually, 1000MB == 1GB...
you're probably thinking 1024MiB = 1GiB
If someone is suing Apple, etc, over the definition of 'mega', then they're going to lose.
It certainly helps when 'random.scramble()' is part of the standard libraries :)
Here's a similar working program in python:
/usr/bin/env python
#!
import fileinput, re, random
def scramble_word ( mo ):
word = mo.group(0)
innards = list(word[1:-1])
random.shuffle(innards)
return word[0] + "".join(innards) + word[-1]
for l in fileinput.input():
l = re.sub ( "\w{4,}", scramble_word, l )
print l
Oh... I never meant to imply that I thought this was a GOOD thing. I'm just trying to make sense of the Verisign rep's comments, and figure out WHAT he could possibly mean by 'it'll only redirect web queries'.
Doing the redirect thing on www. domains is the only thing I could think of.
IT could very well be that they're saying that queries for www.sometyponame.com will return an IP address, but sometyponame.com will return a negative result.
Replying to my own post... how crass...
There's a better picture of the mouse on this brochure for the successor, the 5630.
The clearly the best mouse ever was the hemispherical, red-clown-nose mouse that came with the AT&T/Teletype 5620 terminal. What a buzz using that thing was :)
To expand... on one file, md5sums for the following groups of lines are calculated:
... and so on.
1,2,3
2,3,4
3,4,5
4,5,6
Once the code spits out the md5sums, all matching md5sums are compared, regardless of which line they're on (or started on, since it's 3-line chunks).
So, no, an extra line, 10, or 1000 won't throw it off.
It TOTALLY depends on the browser you're using.
If you're on web page A, click on a link to B and it redirects to C, some browsers will, when fetching C, have a referrer of A, and some will have a referrer of B.
Wow, taken to the extreme, the exploitation of their systems could have caused a train collision and injury or death to hundreds of Maryland and Virginia commuters.
:)
Will Microsoft's new 'Indemnification push' cover the legal costs for something like this, I wonder?
Well, it appears that SCO have updated their calendar software to remove the 'postpoinment'[sic] message :)
I'll read in that someone can't spell :)
It seems strange, though, that they still have all their glitzy web pages advertising the SCO Forum, but there's one tiny little part that says the forum is postponed. Here's my analysis:
1. It was some kind of crack.
2. It really is postponed, But they don't want to advertise the fact, both for the obvious reasons of negative PR, but also because they want all the anti-SCO crowd to turn up anyway. They've probably emailed all the people who have actually registered to make sure they're not cheesed off. However, this will seriously impact the people who legitimately wanted to turn up, but were going to sign-in on the day.
About 4 million hits, or 400 thousand pageviews, per day, here :)
Just under 40,000 hits in the busiest day... this is a slashdotting? Come back when you get into the millions. :)