Nope, if you don't have a valid MX record and the server sending the message isn't yours then you don't get through. I have kept with this ISP for over a decade precisly because they do what the customer wants and today the number one thing a residential ISP can do to keep customers happy is to reduce and minimize spam. After having their servers repeatedly crushed under the onslaught of spammers they finally became somewhat militant about blocking spammers and I congratulate them for doing so. A very large %age of spam comes from hijacked boxes on cable or dsl lines so one of their first checks is to see if the line is in a registered dialup/consumer dsl/cablemodem netblock so you would never even get to the MX check =) Sorry but spammers have abused the system so much that blindly following RFC's written in a MUCH more innocent time just doesn't work.
Cool sounds like SPEWS did it's job, your ISP was forced to remove a spammer from their network. That's one less server sending gigs of unwanted email into the internet. Sorry but spam is a tragedy of the commons. We have this great public shared resource and a few people are pissing in the well. If we have to make some people taste the piss to get their neighbors to stop pissing then so be it.
Have you made sure to unmark every incorrect email? Also turn off auto-placement and select all your legit email, if any of it gets flaged as spam untag it, then do the opposite with spam, if any of it gets unmarked by the filter remark it as spam. This will significantly increase your hit ratio and lower your false positives. For me it is about 99.7% effective and about 1.5% false positives, none of which have been important emails (mass mailings from companies I have a legitimate business relationship with, but their mass mailing address wasn't yet in my personal address book to get whitelisted)
My ISP already does this, all incoming emails are checked to confirm that the email address's MX record is legit and the server that the message is coming from matches one of those MX records. This sometimes trips up bad mail admins or people running new mail servers, but so what?
Mailing lists have to be whitelisted manually, they get no special consideration. If you sign up for a mailing list without adding it to the whitelist then it is your own faux pa (sp?). People can live with a little inconvenience like adding addresses to whitelist and getting challenges on their mailing lists, what people can't live with is thousands of pieces of spam a month.
You block outbound 25 and limit the number of messages per account to something reasonable (even 1K messages would severly curtail the problem and should not affect many legitimate users). Also by using legit mail relays all of the messages will have lefit FROM: headers so tracking down the infected machine(s) will be much easier.
Check out Better than diamond They have full created diamonds, diamond coated synthetics, and processed diamonds that are some of the most flaw free natural diamonds.
Try to make a computer without gold, won't happen. Uranium or Plutonium is needed for the kicker charge in thermonuclear weapons, the intense heat, pressure, and nutron energy needed can only be achieved through conventional nuclear decay or in a tokamak style magnetic containment reactor (but so far no one has broken even on energy input consistantly)
Well since he isn't planning on selling them until at LEAST January I would say they don't exist right now on the market. I did state that there were companies targeting this fall, guess the schedule slipped a bit =) Besides the prices quoted are 10-50% off, or an average of ~2/3rds which is what I stated. The cost savings aren't that significant. My comment was on the current state of the commercial gem business, not what might or might not be on the market going forward. BTW the earnings I am buying are Asha diamond simulants from Better than diamonds They have diamond simulants, artificial diamonds, and improved natural diamonds where impurities are driven out through physical processes. The Asha's are the only "cheap" thing they sell. They are a high quality CZ that are diamond plated through plasma vapor deposition which gives them the sparkle are wear of diamonds at a fraction of the price.
There is only one substance harder than diamond, that is there is only one thing that will scratch a diamond and it is a fairly recent synthetic. Diamonds however will readily cleave along faults in their lattice. As to prettyness that is completely in the eye of the beholder but a firey diamond is definitly a beutiful stone.
Gem quality artificial diamons of significant size don't yet exist (at least not colorless although one company claims to have them coming out this fall), they will NOT be $5, but rather about 2/3rds the cost of the natural ones. And beyond that DeBeers has a flourescense test that will detect artificial diamonds. The real battle will be in convincing Jane Q Public that her man isn't being a cheapskate looser if he gets her an artificial stone for her ring. I have convinced my wife that the replacements to the 1/2 carrat earings that she lost will be 1 carrat synthetics with a gas deposited diamond coating, they are as pretty in the light as real diamons and cost about 1/4th what the smaller "real" stones cost me =) Of course not every woman is so ameniable as the misses.
The difference is that other than a tiny amount of time to process the junk commercial bulk mail does not cost the recipient anything. On the other hand spam has HUGE costs for the recipient (talk to your mail admin sometime to get an idea, fighting spam can be a full time job and makes the job of keeping the mail servers humming much more difficult).
The registry has per user hives (always has) so it is not the OS's fault that app writers are lazy/stupid. Also under 2K and above you can right click an app and select "run as" and then use any users security context. Under XP this is even more enhanced by having a sandboxed mode that prevents many destructive actions from being performed even if the app is run as admin.
For the 9x codebase this is correct, for NT this is completely bull. The NT kernel and even the userspace system are designed for multiple users. Otherwise profiles and NTFS permissions for the local filesystem would be superfelous.
Actually the NT based systems WERE designed for security, especially with win2k and up and code signing. It is the piss poor implementation that has blown windows security. Office on the other hand is not designed with security in mind and it is Office and particularly Outlook which is the biggest windows security vulnerability. The next biggest security problem is IIS and it's poor default configs, but IIS 6 which ships with Server 2003 has MUCH more sane defualts, including it not being installed at all by defualt!
PGP sign all your email, that way you will be able to prove that an infecting email did not originate from you. Also the very fact that it is a windows worm and you run Linux should indemnify you.
I was the Linux technical resource for a desktop support team that supported ~50K desktops, of those 10K were Solaris, 36K were windows (mostly 2K), and 4K were RH Linux. Linux was only 8% of the total but still a lot more than 20 systems =) And Sun was showing this desktop running on their thin terminals so I don't think you have to worry about resources too much =) Oh yeah and anything is an improvement from CDE.
I'm sure that's why EMC, IBM, Hitachi data systems, Netapp, StorageTek, etc all have enterprise class disk backup systems. In fact enterprises are the MOST likely to use disk backup, their backup windows are extremely small and are basically going towards zero so they need the ability to copy data extremely quickly, there is no tape or optical system that can keep up with RAID arrays, even the gigantic stuff like a StorageTek Powderhorn don't have the agregate throughput of the larger disk systems. Of course most of the time these systems are part of a staged system that eventually ends up on tape for archival purposes, but some enterprises are even contemplating eliminating tape alltogether. And yes, large companies like Progressive DO backup everything over the WAN, in fact they replicate DB transactions across the WAN in realtime so that should one datacenter go offline nothing is lost.
Actually most of the DLT partners along with the major backup vendors (Veritas and a couple others) got together just a couple months ago to define a standard for drive error stats and a software interface that would allow standardized tracking of DLT media and predictable early warning. This was in an effort to reduce error rates on restore and save tape from being almost completely replaced with arrays of IDE disks in the enterprise sector.
Or how about just using a medium meant for archival storage like high quality CD-R's from name brand vendors and tape backup like DLT, LTO, or even DAT. Or for a cheap home solution get a decent sized external hdd and just do differential backups ocassionally.
Same here, I have CD-R's from my Ricoh CDRW (the first one on the market) that are now 6 1/2 years old and all of them that aren't scratched to hell are still readable. I wonder if this isn't indicitive of a falloff in quality of blanks. Back then blank CDR's were over a buck a piece and to get ones that would read in most normal cdrom or audio players you had to pay around $4-5 a piece, now you can get CD-R's for around 5 cents a piece. Obviously scales of economy have dropped the price but I'm wondering if cheapening of materials has not also been a big part of it.
Gaming generally doesn't take a lot of bandwidth unless you are running the server. A single T-1 can handle over 100 NWN players for instance. Figure $50/user * 400 game players = $20,000 that more than pays for the 4-5 T-1's they will need while playing and when they aren't playing that bandwidth can be used for web sessions. The only thing that really kills ISP's is people who run servers nonstop.
You forget that it is basically impossible for CLEC's to install DSLAMs at all POP's because there is "insufficient room" yet when the ILEC wants to expand their equipment there is plenty of rack space. While I agree that the ILEC should not be required to provide networking services for less than the cost of deployment I DO think they should be required to provide undundled access to the DSLAM. Also every expert that isn't employed by the ILECs has stated that this will do NOTHING to speed up broadband rollout and will result in higher prices.
Yep, it's called IBA or Intel Boot Agent, it allows booting of a diskless system through PXE. It's actually where Paladium came from origionally. In order to pull a boot image over the network and be sure it was not tampered with on the wire through a man in the middle attack you need hardware crypto with a signed boot image. Basically every PC made in the last 4-5 years or so supports it (there are exceptions but they are usually consumer only oriented only PC's, corporate PC's almost all support it). It's how many enterprises install linux and windows if they don't use ghost.
Slashdot Mirror
Nope, if you don't have a valid MX record and the server sending the message isn't yours then you don't get through. I have kept with this ISP for over a decade precisly because they do what the customer wants and today the number one thing a residential ISP can do to keep customers happy is to reduce and minimize spam. After having their servers repeatedly crushed under the onslaught of spammers they finally became somewhat militant about blocking spammers and I congratulate them for doing so. A very large %age of spam comes from hijacked boxes on cable or dsl lines so one of their first checks is to see if the line is in a registered dialup/consumer dsl/cablemodem netblock so you would never even get to the MX check =) Sorry but spammers have abused the system so much that blindly following RFC's written in a MUCH more innocent time just doesn't work.
Cool sounds like SPEWS did it's job, your ISP was forced to remove a spammer from their network. That's one less server sending gigs of unwanted email into the internet. Sorry but spam is a tragedy of the commons. We have this great public shared resource and a few people are pissing in the well. If we have to make some people taste the piss to get their neighbors to stop pissing then so be it.
Have you made sure to unmark every incorrect email? Also turn off auto-placement and select all your legit email, if any of it gets flaged as spam untag it, then do the opposite with spam, if any of it gets unmarked by the filter remark it as spam. This will significantly increase your hit ratio and lower your false positives. For me it is about 99.7% effective and about 1.5% false positives, none of which have been important emails (mass mailings from companies I have a legitimate business relationship with, but their mass mailing address wasn't yet in my personal address book to get whitelisted)
My ISP already does this, all incoming emails are checked to confirm that the email address's MX record is legit and the server that the message is coming from matches one of those MX records. This sometimes trips up bad mail admins or people running new mail servers, but so what?
Mailing lists have to be whitelisted manually, they get no special consideration. If you sign up for a mailing list without adding it to the whitelist then it is your own faux pa (sp?). People can live with a little inconvenience like adding addresses to whitelist and getting challenges on their mailing lists, what people can't live with is thousands of pieces of spam a month.
You block outbound 25 and limit the number of messages per account to something reasonable (even 1K messages would severly curtail the problem and should not affect many legitimate users). Also by using legit mail relays all of the messages will have lefit FROM: headers so tracking down the infected machine(s) will be much easier.
Check out Better than diamond They have full created diamonds, diamond coated synthetics, and processed diamonds that are some of the most flaw free natural diamonds.
Try to make a computer without gold, won't happen. Uranium or Plutonium is needed for the kicker charge in thermonuclear weapons, the intense heat, pressure, and nutron energy needed can only be achieved through conventional nuclear decay or in a tokamak style magnetic containment reactor (but so far no one has broken even on energy input consistantly)
Well since he isn't planning on selling them until at LEAST January I would say they don't exist right now on the market. I did state that there were companies targeting this fall, guess the schedule slipped a bit =) Besides the prices quoted are 10-50% off, or an average of ~2/3rds which is what I stated. The cost savings aren't that significant. My comment was on the current state of the commercial gem business, not what might or might not be on the market going forward. BTW the earnings I am buying are Asha diamond simulants from Better than diamonds They have diamond simulants, artificial diamonds, and improved natural diamonds where impurities are driven out through physical processes. The Asha's are the only "cheap" thing they sell. They are a high quality CZ that are diamond plated through plasma vapor deposition which gives them the sparkle are wear of diamonds at a fraction of the price.
There is only one substance harder than diamond, that is there is only one thing that will scratch a diamond and it is a fairly recent synthetic. Diamonds however will readily cleave along faults in their lattice. As to prettyness that is completely in the eye of the beholder but a firey diamond is definitly a beutiful stone.
Gem quality artificial diamons of significant size don't yet exist (at least not colorless although one company claims to have them coming out this fall), they will NOT be $5, but rather about 2/3rds the cost of the natural ones. And beyond that DeBeers has a flourescense test that will detect artificial diamonds. The real battle will be in convincing Jane Q Public that her man isn't being a cheapskate looser if he gets her an artificial stone for her ring. I have convinced my wife that the replacements to the 1/2 carrat earings that she lost will be 1 carrat synthetics with a gas deposited diamond coating, they are as pretty in the light as real diamons and cost about 1/4th what the smaller "real" stones cost me =) Of course not every woman is so ameniable as the misses.
The difference is that other than a tiny amount of time to process the junk commercial bulk mail does not cost the recipient anything. On the other hand spam has HUGE costs for the recipient (talk to your mail admin sometime to get an idea, fighting spam can be a full time job and makes the job of keeping the mail servers humming much more difficult).
The registry has per user hives (always has) so it is not the OS's fault that app writers are lazy/stupid. Also under 2K and above you can right click an app and select "run as" and then use any users security context. Under XP this is even more enhanced by having a sandboxed mode that prevents many destructive actions from being performed even if the app is run as admin.
For the 9x codebase this is correct, for NT this is completely bull. The NT kernel and even the userspace system are designed for multiple users. Otherwise profiles and NTFS permissions for the local filesystem would be superfelous.
Actually the NT based systems WERE designed for security, especially with win2k and up and code signing. It is the piss poor implementation that has blown windows security. Office on the other hand is not designed with security in mind and it is Office and particularly Outlook which is the biggest windows security vulnerability. The next biggest security problem is IIS and it's poor default configs, but IIS 6 which ships with Server 2003 has MUCH more sane defualts, including it not being installed at all by defualt!
PGP sign all your email, that way you will be able to prove that an infecting email did not originate from you. Also the very fact that it is a windows worm and you run Linux should indemnify you.
I was the Linux technical resource for a desktop support team that supported ~50K desktops, of those 10K were Solaris, 36K were windows (mostly 2K), and 4K were RH Linux. Linux was only 8% of the total but still a lot more than 20 systems =) And Sun was showing this desktop running on their thin terminals so I don't think you have to worry about resources too much =) Oh yeah and anything is an improvement from CDE.
small fry????
I'm sure that's why EMC, IBM, Hitachi data systems, Netapp, StorageTek, etc all have enterprise class disk backup systems. In fact enterprises are the MOST likely to use disk backup, their backup windows are extremely small and are basically going towards zero so they need the ability to copy data extremely quickly, there is no tape or optical system that can keep up with RAID arrays, even the gigantic stuff like a StorageTek Powderhorn don't have the agregate throughput of the larger disk systems. Of course most of the time these systems are part of a staged system that eventually ends up on tape for archival purposes, but some enterprises are even contemplating eliminating tape alltogether. And yes, large companies like Progressive DO backup everything over the WAN, in fact they replicate DB transactions across the WAN in realtime so that should one datacenter go offline nothing is lost.
Actually most of the DLT partners along with the major backup vendors (Veritas and a couple others) got together just a couple months ago to define a standard for drive error stats and a software interface that would allow standardized tracking of DLT media and predictable early warning. This was in an effort to reduce error rates on restore and save tape from being almost completely replaced with arrays of IDE disks in the enterprise sector.
Or how about just using a medium meant for archival storage like high quality CD-R's from name brand vendors and tape backup like DLT, LTO, or even DAT. Or for a cheap home solution get a decent sized external hdd and just do differential backups ocassionally.
Same here, I have CD-R's from my Ricoh CDRW (the first one on the market) that are now 6 1/2 years old and all of them that aren't scratched to hell are still readable. I wonder if this isn't indicitive of a falloff in quality of blanks. Back then blank CDR's were over a buck a piece and to get ones that would read in most normal cdrom or audio players you had to pay around $4-5 a piece, now you can get CD-R's for around 5 cents a piece. Obviously scales of economy have dropped the price but I'm wondering if cheapening of materials has not also been a big part of it.
Gaming generally doesn't take a lot of bandwidth unless you are running the server. A single T-1 can handle over 100 NWN players for instance. Figure $50/user * 400 game players = $20,000 that more than pays for the 4-5 T-1's they will need while playing and when they aren't playing that bandwidth can be used for web sessions. The only thing that really kills ISP's is people who run servers nonstop.
You forget that it is basically impossible for CLEC's to install DSLAMs at all POP's because there is "insufficient room" yet when the ILEC wants to expand their equipment there is plenty of rack space. While I agree that the ILEC should not be required to provide networking services for less than the cost of deployment I DO think they should be required to provide undundled access to the DSLAM. Also every expert that isn't employed by the ILECs has stated that this will do NOTHING to speed up broadband rollout and will result in higher prices.
Yep, it's called IBA or Intel Boot Agent, it allows booting of a diskless system through PXE. It's actually where Paladium came from origionally. In order to pull a boot image over the network and be sure it was not tampered with on the wire through a man in the middle attack you need hardware crypto with a signed boot image. Basically every PC made in the last 4-5 years or so supports it (there are exceptions but they are usually consumer only oriented only PC's, corporate PC's almost all support it). It's how many enterprises install linux and windows if they don't use ghost.
$600,000 dollars / 1,200 Gflops = $500/Gflop. I think you misplaced a decimal in the $Cad->$US conversion =)