No, they don't need to.
Akamai's model is to install a bunch of their own machines (a PoP) in each and every middle-to-large ISP. They then use source-based DNS to direct requests to the nearest PoP (with some luck, it'll be within your ISP's network).
They basically work as a smart reverse-proxy. You make your request to their PoP, and the PoP serves the content from cache. If you happen to be the first person requesting said content, the PoP will fetch it from the originating server (Apple, MS, CNN, whatever) and cache it to serve following request.
Both ndiswrapper (free and Free) and DriverLoader (commercial and proprietary) work on FC2. DL is more stable, though (ndiswrapper had a few occasional crashes every now and then, DL works flawlessly for me)
ShiteFinder abused the gTLD servers, but they still control the root servers. "master list of suffixes" is a good non-tech way of saying "root servers". So yes, VeriSlime controls the "master list". It is them who redirects.fr requests to AFNIC.
The difference between RH8 and RH9 isn't artificial. Most threaded apps break in RH9 due to the NPTL (there are workarounds, but ISVs don't support them)
A possible (depending on what people want) advantage in the new up2date is that you don't depend on RHN anymore, and if you don't depend on RHN you don't have to put up with overloaded servers: you can choose a nice quiet Fedora mirror near you and update from there.
Or, if you want more software, add different repositories (besides your main source for updates. As I stated earlier, you can use multiple repositories) like a mirror for http://fedora.us/ or http://rpm.livna.org/
Yeah, RedHat (Enterprise). Or Fedora.
Except now you can edit/etc/sysconfig/rhn/sources and add as many up2date, apt, AND yum sources as you wish. Found a new great apt repository? Fine, put it in there and keep using up2date.
Or switch to apt. Or run "yum update". Or whatever...
I honestly don't understand why so many people seem to think they must choose something else after RH dropped the "consumer market". I can understand going from RH to a commercial distro like SuSE due to fear or disconfort in using a "community-supported" distro like Fedora. But going from RH to Slackware or Gentoo because RH doesn't support the distro anymore? Are you people nuts?
I upgraded my home desktop straight from 7.2 to FC1. Worked like a charm, and was very much worth it. The GNOME upgrade alone justifies the minimal hassle.
I did that in a few machines (with yum instead of apt), with varied results.
Some machines booted perfectly, others had problems. The most successful case I had was a RH 7.2 to FC1 upgrade where the only thing failing was a threaded app that didn't like NPTL (LD_ASSUME_KERNEL solved that one in an instant)
From a binary/package-version perspective, FC1 is fully compatible with RHEL3 (including in the NPTL issues), so for most major apps you should actually look for RHEL support instead of Fedora.
I did this with a few apps (most notably Oracle 9i) with good results (packages and instructions for RHEL3, in my experience, always work for FC1)
By "we", I mean those who maintain moderately large mail systems. My own mailsystem gets about 8 million messages per day;
Of those 8 million, approximately 4 are rejected right at the SMTP level for various reasons (mostly inexistant domains or mail recipients); according to spamassassin, 43% of what does get in is also SPAM. That's almost another 2M. In other words, I (or my company) could have invested less than 25% of what I actually did in equipment (the spam filtering servers alone are almost a third of the total), and my nights and days would be *MUCH* less stressful.
They don't. There are 3 common WLAN chips nowadays: Centrino (Intel), BCM43xx (Broadcom) and ACX100 (TI). Some might argue a fourth chip into the list: Atheros, but there aren't many OEMs using it, at least where I buy my stuff.
Hey, it's been about 10 years since I last read them...:-)
Anyway, they _were_ about the same age, so Elijah Wood wouldn't be a problem. Wood, McKellan, and Rhys-Davies could play Bilbo, Gandalf, and Gloin. I don't think it would work, though; Well... maybe just Gandalf, and Andy Serkis as a slightly less crooked Smeagol for extra fun-points:)
Not necessarily... IIRC, both Frodo and Bilbo were about the same age when they left the Shire (30-something years old). That's one of the few things I didn't like about the movies; Frodo shouldn't have looked _so_ young...
if I want to make sure my app works with Microsofts OS I pay to them an extra $500 for the privilege.
In a commercial software world, it does make some sense; think of it from MS's POV: "If you're going to profit from an application which needs our applications/OS to work, you might as well pay for the priviledge of being able to market it as soon as our OS/SP comes out. Oh, free (as in beer) software? Screw you, then.". It's not like they need it, but it a good excuse
"Received" lines are already mandatory, and the contain the IP address of the server who places them.
As to validating them... It's tricky, if not impossible. Imagine this scenario: A user with a MUA, a frontend/relay combo at his ISP, and a remote MX receiving the message, and then actually forwarding it to an all-together different domain; respective IPs, 1.2.3.1, 4.5.6.1, 4.5.6.2, 10.10.10.10, 10.20.30.40
MTAs prepend the Received lines to the rest of the message, but for argument's sake let's write them top-down:
1st hop: Received by 4.5.6.1 from 1.2.3.1
2nd hop: Received by 4.5.6.2 from 4.5.6.1
3rd hop: Received by 10.10.10.10 from 4.5.6.2
4th hop: Received by 10.20.30.40 from 10.10.10.10
How do you propose 10.20.30.40 validates the original sender was actually 1.2.3.1? How can it even validate it came from the ISP whose mailservers are at 4.5.6.x? They never "talked" to eachother!
I've actually had this happening to me, several times: Spammers forging hops 1, 2, and the second part of 3 with IPs of mine, using their own IP in the first part of hop 3, and spamming along; the only thing giving them up was using IP addresses for the origin which I use as MXs. My MXs are always load-balancing switches and they could never _send_ e-mail. But the victims couldn't know that, and guess who gets the blame...
Re:Microsoft already did this about 5 years ago
on
Gates on Spam
·
· Score: 1
No, the math puzzle is when it spits out "Illegal divide by zero". "Illegal exception" means "I would gladly solve the puzzle, except you're using a Pentium(TM) and it can't divide!".
"Isn't this like saying that two out of four strains of ebola have been found to be susceptible to anthrax?"
Nope, RTFA; Using a variant of your own example, it's like saying two out of four types of pinworms enable anyone, anywhere, to place anthrax directly into your system just by telling the pinworms that "food" is coming along.:-)
Slashdot Mirror
No, they don't need to. Akamai's model is to install a bunch of their own machines (a PoP) in each and every middle-to-large ISP. They then use source-based DNS to direct requests to the nearest PoP (with some luck, it'll be within your ISP's network). They basically work as a smart reverse-proxy. You make your request to their PoP, and the PoP serves the content from cache. If you happen to be the first person requesting said content, the PoP will fetch it from the originating server (Apple, MS, CNN, whatever) and cache it to serve following request.
Both ndiswrapper (free and Free) and DriverLoader (commercial and proprietary) work on FC2. DL is more stable, though (ndiswrapper had a few occasional crashes every now and then, DL works flawlessly for me)
Remember that Verislime controls two major DNS infrastructures:
.com/.net gTLD servers
.fr requests to AFNIC.
1 - The
2 - The TLD/ccTLD root servers (all ccTLD)
ShiteFinder abused the gTLD servers, but they still control the root servers. "master list of suffixes" is a good non-tech way of saying "root servers". So yes, VeriSlime controls the "master list". It is them who redirects
I get around 50/60 pieces of SPAM each day in my abuse address. I eventually gave up and placed an address-confirmation script in front of it...
That's $3159 per machine, and those are today's prices... They weren't so low a couple of years ago...
The difference between RH8 and RH9 isn't artificial. Most threaded apps break in RH9 due to the NPTL (there are workarounds, but ISVs don't support them)
A possible (depending on what people want) advantage in the new up2date is that you don't depend on RHN anymore, and if you don't depend on RHN you don't have to put up with overloaded servers: you can choose a nice quiet Fedora mirror near you and update from there.
Or, if you want more software, add different repositories (besides your main source for updates. As I stated earlier, you can use multiple repositories) like a mirror for http://fedora.us/ or http://rpm.livna.org/
Yeah, RedHat (Enterprise). Or Fedora. /etc/sysconfig/rhn/sources and add as many up2date, apt, AND yum sources as you wish. Found a new great apt repository? Fine, put it in there and keep using up2date.
Except now you can edit
Or switch to apt. Or run "yum update". Or whatever...
I honestly don't understand why so many people seem to think they must choose something else after RH dropped the "consumer market". I can understand going from RH to a commercial distro like SuSE due to fear or disconfort in using a "community-supported" distro like Fedora.
But going from RH to Slackware or Gentoo because RH doesn't support the distro anymore? Are you people nuts?
I upgraded my home desktop straight from 7.2 to FC1. Worked like a charm, and was very much worth it. The GNOME upgrade alone justifies the minimal hassle.
I did that in a few machines (with yum instead of apt), with varied results.
Some machines booted perfectly, others had problems. The most successful case I had was a RH 7.2 to FC1 upgrade where the only thing failing was a threaded app that didn't like NPTL (LD_ASSUME_KERNEL solved that one in an instant)
From a binary/package-version perspective, FC1 is fully compatible with RHEL3 (including in the NPTL issues), so for most major apps you should actually look for RHEL support instead of Fedora.
I did this with a few apps (most notably Oracle 9i) with good results (packages and instructions for RHEL3, in my experience, always work for FC1)
By "we", I mean those who maintain moderately large mail systems. My own mailsystem gets about 8 million messages per day;
Of those 8 million, approximately 4 are rejected right at the SMTP level for various reasons (mostly inexistant domains or mail recipients); according to spamassassin, 43% of what does get in is also SPAM. That's almost another 2M. In other words, I (or my company) could have invested less than 25% of what I actually did in equipment (the spam filtering servers alone are almost a third of the total), and my nights and days would be *MUCH* less stressful.
Where do you think they get the cheap copies of Office FROM? ;)
Not really; it's more like "SCO seeks suckers down under"
'cause they don't have Nice Tits like the British, so they're jealous.
Better: Try defining FCC as Fucking Censoring Cunts.
They don't. There are 3 common WLAN chips nowadays: Centrino (Intel), BCM43xx (Broadcom) and ACX100 (TI). Some might argue a fourth chip into the list: Atheros, but there aren't many OEMs using it, at least where I buy my stuff.
Hey, it's been about 10 years since I last read them... :-) :)
Anyway, they _were_ about the same age, so Elijah Wood wouldn't be a problem. Wood, McKellan, and Rhys-Davies could play Bilbo, Gandalf, and Gloin. I don't think it would work, though; Well... maybe just Gandalf, and Andy Serkis as a slightly less crooked Smeagol for extra fun-points
Not necessarily... IIRC, both Frodo and Bilbo were about the same age when they left the Shire (30-something years old). That's one of the few things I didn't like about the movies; Frodo shouldn't have looked _so_ young...
Maybe... But having Rhys-Davies playing Gloin (Gimli's father) would be delicios and much more feasible...
if I want to make sure my app works with Microsofts OS I pay to them an extra $500 for the privilege.
In a commercial software world, it does make some sense; think of it from MS's POV: "If you're going to profit from an application which needs our applications/OS to work, you might as well pay for the priviledge of being able to market it as soon as our OS/SP comes out. Oh, free (as in beer) software? Screw you, then.". It's not like they need it, but it a good excuse
"Received" lines are already mandatory, and the contain the IP address of the server who places them.
As to validating them... It's tricky, if not impossible. Imagine this scenario: A user with a MUA, a frontend/relay combo at his ISP, and a remote MX receiving the message, and then actually forwarding it to an all-together different domain; respective IPs, 1.2.3.1, 4.5.6.1, 4.5.6.2, 10.10.10.10, 10.20.30.40
MTAs prepend the Received lines to the rest of the message, but for argument's sake let's write them top-down:
1st hop: Received by 4.5.6.1 from 1.2.3.1
2nd hop: Received by 4.5.6.2 from 4.5.6.1
3rd hop: Received by 10.10.10.10 from 4.5.6.2
4th hop: Received by 10.20.30.40 from 10.10.10.10
How do you propose 10.20.30.40 validates the original sender was actually 1.2.3.1? How can it even validate it came from the ISP whose mailservers are at 4.5.6.x? They never "talked" to eachother!
I've actually had this happening to me, several times: Spammers forging hops 1, 2, and the second part of 3 with IPs of mine, using their own IP in the first part of hop 3, and spamming along; the only thing giving them up was using IP addresses for the origin which I use as MXs. My MXs are always load-balancing switches and they could never _send_ e-mail. But the victims couldn't know that, and guess who gets the blame...
No, the math puzzle is when it spits out "Illegal divide by zero". "Illegal exception" means "I would gladly solve the puzzle, except you're using a Pentium(TM) and it can't divide!".
Interesting? How the hell does this get modded as interesting? It's a take on an April Fools RFC, ferchrissake! (p.s.: it's funny, laugh!)
"Isn't this like saying that two out of four strains of ebola have been found to be susceptible to anthrax?"
:-)
Nope, RTFA; Using a variant of your own example, it's like saying two out of four types of pinworms enable anyone, anywhere, to place anthrax directly into your system just by telling the pinworms that "food" is coming along.