If you run your own mail server or rely on filtering at your client end the spam uses up your bandwidth, your storage, your CPU resources to filter it, etc.
True, but if most mail admins made the effort, spam would die on the vine for the same reason that vaccinations are effective even if some people aren't immune. Lower the return on spam far enough and it becomes more hassle than it's worth.
Fine, I'm happy for you. You obviously don't own an active domain, or a business. Because otherwise I could guarantee that it gets to be a problem for you.
I do both (well, I work for a guy who owns a business), but neither my home account nor my coworkers' inboxes get nontrivial amounts of spam. I've written instructions on how I did it, and if you follow them, you can probably get rid of your spam problem as well.
It's not easy if you're J. Random Enduser, but any qualified system administrator should be able to take the steps needed to win back control of his servers. You can choose to do this - with today's software - if you're willing to exert a modest amount of effort.
OK, that seems pretty legitimate and I'd probably have done the same thing while arranging the permanent fix. I didn't mean to bag on you, but I have to deal with broken mailservers on a near-daily basis ("What do you mean you won't accept our mail if we send "HELO localhost"? That's it's name!") and hear plenty of other admins demand that I fix their problem. It's a reflex for me to want to smite people who say things like that.:-)
It was impossible to totally close the open relay in version 5.08 I think it was. I had an on-going argument with the orbs blacklist on this, begging them to cut me some slack as users on my network could not route email to certain servers running the blacklist. The issue was finally resolved by taking away lotes as the public mail gateway.
Out of curiosity, why did you expect a database of broken mailservers to whitelist your broken mailserver? Although inconvenient for you, it was very convenient for people who didn't want the spam that was probably relayed through your machine.
When IPv6 is finally implemented, IP address will be much longer than your phone numbers Then you will appreciate the use of FQDN
Dang. You beat me to it.
Scenario One:
Me: OK, mom, the pictures of your grandkids are at 2001:470:1f01:224:1::2. Wait, do you have a pen? That's two zero zero one colon... oh, OK, I'll slow down. That's two. Zero. Zero. One. Colon. No, not "dot". Yeah, I know they use a lot of dots, but this one's a colon. No, you have to press shift first - that's a semicolon.
Scenario Two:
Me: Pictures of the kids are online at www.strauser.com. That's right! Your own last name, right there on the Internet! No, they can't get your social security number. No, seriously. I bought it. The government didn't just assign it to us. No, you don't have to change your email address. Come on, ma, it's supposed to be cute, and no one's gonna steal your identity just because I registered our last name. Ma? Hello?
You know, now that I think of it, he may have a point.
The above was for Just some guy's "lets use all our main memory for hookers and blow because the swap file will save us" idea.
When you first load an application, the VM system (not swap - VM!) loads the parts of that program needed to get it up and running. Then, after the application opens, parses, and loads its data files, you're ready to begin working.
When you revisit an application that's been idle for a while, the VM system (not swap - VM!) loads the parts of that program needed to get it "active" again, typically just the GUI code until you start doing stuff. You're ready to begin working.
The main difference between the two scenarios is that in the first case, you're guaranteed to have to reload the data from disk. In the latter, there's a chance that it'll still be in RAM. Even if it's not, though, you only have to load the preprocessed data (and only enough of it to begin your operations on that data).
You guys seem to think that your applications and data live on fiber channel SANs, while your swap space consists of a floppy striped with a Zip drive. In reality, the only difference is that your swapped out applications are pre-initialized and in the same state you left them. I don't understand why you insist that's a bad thing.
Suppose that eBay's customers were clamoring for their server farm to be powered by cold fusion. eBay responds by making a working cold fusion reactor and patents it. The fact that their customers were the impetus for its creation doesn't mean that eBay wouldn't deserve the patent.
Having said that, this case closer J. Random Troll getting a patent for "barter - in space!" and then suing eBay because they have customers with satellite linkups. It's still a stupid patent lawsuit, but not for the reason you gave.
To that end, I've always wondered what would be more disruptive to the human populace: longer days or longer seconds?
Longer seconds. The change in length of a day is extremely gradual ("glacial" is fast by comparison), but as seconds are defined in terms of physical constants, a changing second means that our physics have ripped and we're fixin' to die.
Seriously... about how many people out there actually need to know NTP to this degree?
Oh, about 10. But how many weird things do you know that not many others would value?
Some people are really, really into keeping time. It's a hobby for them. This book is for that sort of person. Besides, although my company didn't need to hire a person to do nothing but NTP, they certainly needed at least one person on staff with that skillset (hint: Active Directory, Kerberos, "clockskew") to keep everything else working. How fortunate for me that my boss needs the skills that I picked up out of personal curiosity!
Ah, I think I see the problem. You're using the Microsoft "VM == swap" definition instead of the standard OS theory definition. There's actually quite a huge difference between the two concepts.
Why allocate, deallocate, and reallocate when you can just allocate?
That would seem to support my position: why allocate the memory, deallocate it at program exit, then reallocate it when loading the program and data file again?
Seems simpler and more reliable to me. I've been known to be a little crazy from time to time though.
I'm lost at you connection between desktops and Oracleland. I'll take you at your word and trust that you're an expert DBA, but that doesn't make you an expert at OS design.
rrrrright, just try and open photoshop with 1.8gig image loaded in virtual memory. Reloading the app is definitely quicker regardless of your platform.
Ain't no way. A drive with a sustained 30MB/s transfer rate would take 60 seconds to load that image into RAM, ignoring all the malloc()s and other processing that would be required. With VM, however, revisiting that image only requires pulling enough of it from swap back into RAM to start whatever process you're wanting to do on it.
It damn well does matter what is running and what isn't. Every program you have spinning its wheels in the background is eating up memory.
What third-rate OS are you using that doesn't support proper virtual memory? On a modern system, the primary difference between an idle application and one that's not loaded at all is that the idle app's data (but not the app itself) might be moved from RAM to swap. Should you need to access that data, paging it back in is almost certainly going to be faster than starting the app from scratch (thereby having to re-initialize everything, load the data file from disk, parse it into its in-memory representation, etc.).
I'm not really clear on what supposed benefits I should be getting from constantly closing and later restarting my applications. Next time I feel I'm being too productive, I'll give it a shot.
Incidentally, the instructions for installing Firefox 1.5 on Ubuntu are here if you want to subject them to the mother-test yourself.
Next time my mom knows or cares which version of which web browser she's using, I'll take a look at that.
In the mean time, I switched my in-laws from a WebTV to a little quiet PC running Kubuntu. It's easy, it Just Works, and they like it. The only time anything will ever be installed on that PC is if I do it, so talk of the relative difficulty of situations that will never come up is fairly pointless.
Please provide a source that doesn't include the words "marijuana" or "cannabis" in the title or domain name. Every single one of the first ten Google hits for "hemp faq" has at least one of them, which doesn't help a lot on the credibility-o-meter.
Seriously, is anyone not pro-marijuana in favor of industrial hemp farming? I'm genuinely curious.
How much money can we save by forgetting that the middle east exists? Even if E85 were no cheaper per mile than gasoline, the savings from not counting on a bunch of psychopaths to support our economy should be impressive.
Slashdot Mirror
True, but if most mail admins made the effort, spam would die on the vine for the same reason that vaccinations are effective even if some people aren't immune. Lower the return on spam far enough and it becomes more hassle than it's worth.
I do both (well, I work for a guy who owns a business), but neither my home account nor my coworkers' inboxes get nontrivial amounts of spam. I've written instructions on how I did it, and if you follow them, you can probably get rid of your spam problem as well.
It's not easy if you're J. Random Enduser, but any qualified system administrator should be able to take the steps needed to win back control of his servers. You can choose to do this - with today's software - if you're willing to exert a modest amount of effort.
Absolutely: see them in court. There's a lot of financial value to be had in demonstrating that you won't voluntarily submit to extortion.
OK, that seems pretty legitimate and I'd probably have done the same thing while arranging the permanent fix. I didn't mean to bag on you, but I have to deal with broken mailservers on a near-daily basis ("What do you mean you won't accept our mail if we send "HELO localhost"? That's it's name!") and hear plenty of other admins demand that I fix their problem. It's a reflex for me to want to smite people who say things like that. :-)
Out of curiosity, why did you expect a database of broken mailservers to whitelist your broken mailserver? Although inconvenient for you, it was very convenient for people who didn't want the spam that was probably relayed through your machine.
Dang. You beat me to it.
Scenario One:
Me: OK, mom, the pictures of your grandkids are at 2001:470:1f01:224:1::2. Wait, do you have a pen? That's two zero zero one colon... oh, OK, I'll slow down. That's two. Zero. Zero. One. Colon. No, not "dot". Yeah, I know they use a lot of dots, but this one's a colon. No, you have to press shift first - that's a semicolon.
Scenario Two:
Me: Pictures of the kids are online at www.strauser.com. That's right! Your own last name, right there on the Internet! No, they can't get your social security number. No, seriously. I bought it. The government didn't just assign it to us. No, you don't have to change your email address. Come on, ma, it's supposed to be cute, and no one's gonna steal your identity just because I registered our last name. Ma? Hello?
You know, now that I think of it, he may have a point.
Because it's not as embarrassing when you're naked?
Sorry, bro, but that was too easy to pass up.
When you first load an application, the VM system (not swap - VM!) loads the parts of that program needed to get it up and running. Then, after the application opens, parses, and loads its data files, you're ready to begin working.
When you revisit an application that's been idle for a while, the VM system (not swap - VM!) loads the parts of that program needed to get it "active" again, typically just the GUI code until you start doing stuff. You're ready to begin working.
The main difference between the two scenarios is that in the first case, you're guaranteed to have to reload the data from disk. In the latter, there's a chance that it'll still be in RAM. Even if it's not, though, you only have to load the preprocessed data (and only enough of it to begin your operations on that data).
You guys seem to think that your applications and data live on fiber channel SANs, while your swap space consists of a floppy striped with a Zip drive. In reality, the only difference is that your swapped out applications are pre-initialized and in the same state you left them. I don't understand why you insist that's a bad thing.
Having said that, this case closer J. Random Troll getting a patent for "barter - in space!" and then suing eBay because they have customers with satellite linkups. It's still a stupid patent lawsuit, but not for the reason you gave.
That's unusual. All of the rich people I've known would have answered: "used".
He wants his spaces to resonate. Maybe he needs an interior decorator.
Fair enough. Just don't come crying when a new Costco bulldozes your house via imminent domain. After all, it's the law - right?
Nope. It was a private server for citizens of Denmark.
D-Link settled with an NTP admin for an undisclosed amount after abusing his server.
Longer seconds. The change in length of a day is extremely gradual ("glacial" is fast by comparison), but as seconds are defined in terms of physical constants, a changing second means that our physics have ripped and we're fixin' to die.
Oh, about 10. But how many weird things do you know that not many others would value?
Some people are really, really into keeping time. It's a hobby for them. This book is for that sort of person. Besides, although my company didn't need to hire a person to do nothing but NTP, they certainly needed at least one person on staff with that skillset (hint: Active Directory, Kerberos, "clockskew") to keep everything else working. How fortunate for me that my boss needs the skills that I picked up out of personal curiosity!
2. Get D-Link to use you as the non-configurable time source for a line of disposable networking gear.
3. Profit!
Congrats to PHK for finding the elusive middle step!
Ah, I think I see the problem. You're using the Microsoft "VM == swap" definition instead of the standard OS theory definition. There's actually quite a huge difference between the two concepts.
That would seem to support my position: why allocate the memory, deallocate it at program exit, then reallocate it when loading the program and data file again?
Seems simpler and more reliable to me. I've been known to be a little crazy from time to time though.
I'm lost at you connection between desktops and Oracleland. I'll take you at your word and trust that you're an expert DBA, but that doesn't make you an expert at OS design.
Ain't no way. A drive with a sustained 30MB/s transfer rate would take 60 seconds to load that image into RAM, ignoring all the malloc()s and other processing that would be required. With VM, however, revisiting that image only requires pulling enough of it from swap back into RAM to start whatever process you're wanting to do on it.
What third-rate OS are you using that doesn't support proper virtual memory? On a modern system, the primary difference between an idle application and one that's not loaded at all is that the idle app's data (but not the app itself) might be moved from RAM to swap. Should you need to access that data, paging it back in is almost certainly going to be faster than starting the app from scratch (thereby having to re-initialize everything, load the data file from disk, parse it into its in-memory representation, etc.).
I'm not really clear on what supposed benefits I should be getting from constantly closing and later restarting my applications. Next time I feel I'm being too productive, I'll give it a shot.
Next time my mom knows or cares which version of which web browser she's using, I'll take a look at that.
In the mean time, I switched my in-laws from a WebTV to a little quiet PC running Kubuntu. It's easy, it Just Works, and they like it. The only time anything will ever be installed on that PC is if I do it, so talk of the relative difficulty of situations that will never come up is fairly pointless.
They point to pool.ntp.org, which is designed expressly for this purpose.
As another example, many Linux distributions point to a download site for Microsoft msttcorefonts. Do you think they have permission?
They universally point to SourceForge, which was specifically designed for this exact purpose.
Any other examples?
Please provide a source that doesn't include the words "marijuana" or "cannabis" in the title or domain name. Every single one of the first ten Google hits for "hemp faq" has at least one of them, which doesn't help a lot on the credibility-o-meter.
Seriously, is anyone not pro-marijuana in favor of industrial hemp farming? I'm genuinely curious.
How much money can we save by forgetting that the middle east exists? Even if E85 were no cheaper per mile than gasoline, the savings from not counting on a bunch of psychopaths to support our economy should be impressive.