Re: SERIAL NUMBER PROBLEM *AGAIN*!!!
This WebDNA talk-list message is from 1998
It keeps the original formatting.
numero = 21279
interpreted = N
texte = If anyone feels uncomfortable with the SN issue - watch the little lights onyour hub when your server launches. Whatever is going on, it certainly isdamn efficient - which stands to reason, in light of how well the rest ofthe product is designed.The folks at PCS have found a way of validating SNs without having to 'phonehome' or depend on another server for validation. It does not have to passSNs back and forth; PCS probably wouldn't have done something arbitrarilymoronic like broadcasting unencrypted SNs, nor does it cache SNs anywherebut in your prefs.It seems logical to me - and I am only operating on pure hunch - that thepackage broadcasts a checksum based on SN, and leaves it parked in the TCPstack. Any other copy will do the same. If two copies with the samechecksum find each other, one returns to demo mode.When WebCat quits, it clears the checksum info from the stack. A servercrash will leave that info stranded in the TCP stack. When the WebCat copyon that very same machine is re-launched, it sees the stranded data andassumes that it is there because of another copy running nearby or even onthe same machine, and it reverts to demo mode.Just thinking about this a little bit, I wonder what sort of magic it wouldtake to make sure the stack info is cleaned even after a bomb. Would yourun a separate process for just this? Would you keep setting and clearingthe data in hopes that the crash occurs while the data is cleared? It's alot of work for only one purpose - to be able to restart the server after acrash, even though the computer has not been restarted.That's nuts. The people who are doing the most bitching about this areprobably the same ones running the server on the same machine with theirbrowser and editing software, so rebooting takes too long. Many of theemails are worded to make it sound as if being unable to relaunch softwareon an unstable machine somehow does a disservice to customers. PCS would dowell to preserve this as a feature since it forces a re-launch and preventsunstable operation from inaccurately tarnishing the reputation of theproduct.As for me, I paid a bunch for the software, so I want the SN scheme to makepiracy a damn pain in the ass. Rebooting the server for me is very rarebecause a properly configured server, on which actual customers actuallyrely, should be stripped of everything that can possibly be removed, andcarefully tuned. If a server does go down, I not only assume that I'm notgoing to just re-launch, but I'm going to divert all traffic to a backupserver, tear the unit that crashed apart and scour the thing to find out howit happened.When did we WebCat users start getting irrationally paranoid? The lastthing we need to do is generate false rumors about the product and makemarketing to our customers that much tougher. Maybe we all need a vacationat Ken's place! I'll bring the badmitton.Now as for this goofy thread - you get the ankles and I'll get the wrists,Pat McCormick----- Original Message -----From: Kenneth Grome
To: Sent: Tuesday, November 24, 1998 5:49 PMSubject: Re: SERIAL NUMBER PROBLEM *AGAIN*!!!>>So webcat is using bandwidth and cycles systematically culling millions>>of ip addresses for some secret handshake? I find that highly unlikely.>> If it is true, we're all in for a very strange ride.>>If that's what it's doing, what happens when it finds an IP address runningwebcat? Does it pass the SN back and forth between the two copies ofwebcat -- unencrypted -- to determine whether the SN is the same? Does itlog the IP address of the other machine when it finds a matching SN? Whatother goodies does this new version of webcat offer?>>This sounds like something Microsoft would do ...>>Sincerely,>Ken Grome>808-737-6499>WebDNA Solutions>mailto:ken@webdna.net>http://www.webdna.net>>>>>>
Associated Messages, from the most recent to the oldest:
If anyone feels uncomfortable with the SN issue - watch the little lights onyour hub when your server launches. Whatever is going on, it certainly isdamn efficient - which stands to reason, in light of how well the rest ofthe product is designed.The folks at PCS have found a way of validating SNs without having to 'phonehome' or depend on another server for validation. It does not have to passSNs back and forth; PCS probably wouldn't have done something arbitrarilymoronic like broadcasting unencrypted SNs, nor does it cache SNs anywherebut in your prefs.It seems logical to me - and I am only operating on pure hunch - that thepackage broadcasts a checksum based on SN, and leaves it parked in the TCPstack. Any other copy will do the same. If two copies with the samechecksum find each other, one returns to demo mode.When WebCat quits, it clears the checksum info from the stack. A servercrash will leave that info stranded in the TCP stack. When the WebCat copyon that very same machine is re-launched, it sees the stranded data andassumes that it is there because of another copy running nearby or even onthe same machine, and it reverts to demo mode.Just thinking about this a little bit, I wonder what sort of magic it wouldtake to make sure the stack info is cleaned even after a bomb. Would yourun a separate process for just this? Would you keep setting and clearingthe data in hopes that the crash occurs while the data is cleared? It's alot of work for only one purpose - to be able to restart the server after acrash, even though the computer has not been restarted.That's nuts. The people who are doing the most bitching about this areprobably the same ones running the server on the same machine with theirbrowser and editing software, so rebooting takes too long. Many of theemails are worded to make it sound as if being unable to relaunch softwareon an unstable machine somehow does a disservice to customers. PCS would dowell to preserve this as a feature since it forces a re-launch and preventsunstable operation from inaccurately tarnishing the reputation of theproduct.As for me, I paid a bunch for the software, so I want the SN scheme to makepiracy a damn pain in the ass. Rebooting the server for me is very rarebecause a properly configured server, on which actual customers actuallyrely, should be stripped of everything that can possibly be removed, andcarefully tuned. If a server does go down, I not only assume that I'm notgoing to just re-launch, but I'm going to divert all traffic to a backupserver, tear the unit that crashed apart and scour the thing to find out howit happened.When did we WebCat users start getting irrationally paranoid? The lastthing we need to do is generate false rumors about the product and makemarketing to our customers that much tougher. Maybe we all need a vacationat Ken's place! I'll bring the badmitton.Now as for this goofy thread - you get the ankles and I'll get the wrists,Pat McCormick----- Original Message -----From: Kenneth Grome To: Sent: Tuesday, November 24, 1998 5:49 PMSubject: Re: SERIAL NUMBER PROBLEM *AGAIN*!!!>>So webcat is using bandwidth and cycles systematically culling millions>>of ip addresses for some secret handshake? I find that highly unlikely.>> If it is true, we're all in for a very strange ride.>>If that's what it's doing, what happens when it finds an IP address runningwebcat? Does it pass the SN back and forth between the two copies ofwebcat -- unencrypted -- to determine whether the SN is the same? Does itlog the IP address of the other machine when it finds a matching SN? Whatother goodies does this new version of webcat offer?>>This sounds like something Microsoft would do ...>>Sincerely,>Ken Grome>808-737-6499>WebDNA Solutions>mailto:ken@webdna.net>http://www.webdna.net>>>>>>
Pat McCormick
DOWNLOAD WEBDNA NOW!
Top Articles:
Talk List
The WebDNA community talk-list is the best place to get some help: several hundred extremely proficient programmers with an excellent knowledge of WebDNA and an excellent spirit will deliver all the tips and tricks you can imagine...
Related Readings:
WebCatalog can't find database (1997)
intel mac (2006)
emailer (1997)
Javascript Form Validation (2002)
WebCat & WebTen (1997)
Site Search Concepts (2003)
IIS4b2 and WebCatalog b19 (1997)
HEADER AND FOOTER (1997)
Weird problems with [SHOWIF]s (1997)
Problems with ^ could be solved with [REPLACE CHARACTERS] (1997)
OT: Real time Stats (2003)
Hard Questions ? (1997)
Advice (2003)
Formating found categories (1997)
b12 cannot limit records returned and more. (1997)
notification solutions (1997)
There's a bug in the math context ... (1997)
Crash Log Question (2006)
2nd WebCatalog2 Feature Request (1996)
Is there an alternative to emailer? (1999)