You'd be using your network buffers as swap space. That's not a very efficient use of memory. If you have some LAN server with extra memory that you want to use as swap space for the other hosts, you're probably better off using a Network Block Device instead.
BTW, your Gbps ethernet probably has round-trip latency closer to 0.1 ms.
I've done the circuit-board swap on a hard drive. Not sure what you mean by identical... My replacement board came from the same model drive (Seagate ST11950N) but different firmware revision. The "donor" drive also had a significantly earlier serial number on it. I've heard that every drive is to some degree unique in its surface medium characteristics, and that the board is somehow "tuned" to the media during the initial burn-in and LLF at the factory.
Long story short: it worked fine for me. The data were 100% recovered and the drive continued to work perfectly until it was decommissioned due to obsolescence. (Anybody need a 7200 RPM 1.6 GB drive from 1993?)
A friend of mine has a few defective drives but alas he won't put up the $50 to hunt down an "identical" parts donor, so I haven't yet had the opportunity to try the board-swap operation on a newer drive. Mostly the damn things are just too reliable.
Yes, of course. That's what Kibobytes are for.
On HappyNet everybody has Kibobytes, Mebibytes, Gigglebytes, etc, each with its corresponding pack of BonusBytes[tm]. Hard driver makers, being boring nerdy engineer types, are still using those crappy little metric decimal kilobytes etc. Microsoft on the other hand are very innovative and have already moved to HappyNet, which is why they report everything in the Larger, Friendlier, Happier Gigglebytes.
Speakeasy's name servers were returning NXDOMAIN instead of sitefinder by the 17th. Maybe earlier but that was when I first checked. No discussion announcement as far as I know, they just did the right thing quietly and with impressive alacrity.
Linux desperately NEEDS more people looking at it from a marketing perspective
Well, Linux doesn't need to worry about *marketing* in the strict sense right now so much as making it more accessible to Average Users.
Why? Who is Linux, and why does Linux NEED anything like that? When you say "Linux NEEDS" do you mean "Red Hat NEEDS"? Or maybe, "I NEED Linux to do that thing"? Are you talking about Linux the kernel, or Linux the generic abstract non-specified free OS distribution? If you mean the latter, perhaps you should concentrate some particular distribution and try to convince its maintainers to make it more to your specifications. Or you could make a new distribution, since you have such great ideas on what Linux NEEDS, and own the market with it. Seems like a sure money-maker -- go for it! Get rich! What have you got to lose? Unless of course there's no money in it, which might explain why nobody's out there doing it already....
I'm utterly serious here, not trying to blow anyone off with a canned "Fix it yourself" response. If there really is a market for the Linux you imagine, and you seem very confident that there is, and that it's huge, and apparently nobody is exploiting that market... well... the world has basically handed you a fortune on a silver platter and all you have to do is grab it!
From my perspective, Linux and the distributions I use work fine just the way they are, and I see no reason to reorient them to satisfy some "market" out there. If I thought there were some obvious way Linux could work better for me, I'd change it. If I thought there were a huge market ready to buy into Linux(/GNU/etc) in a big way if only it were changed in some obvious and simple way... well, I sure wouldn't be wasting my time posting to Slashdot about it.
Maybe you feel there's a simple and obvious change that would make Linux(/etc) much better, but you're not quite competent to figure out exactly what it is? That's a common feeling. I think a lot of people feel that way about a lot of things. That doesn't necessarily mean there really is such an easy fix or that nobody competent has ever bothered to think about the problem. Most likely it's a much harder problem than it seems, and a lot of people have thought long and hard about it but come up with nothing yet that would satisfy you. What the world needs is more people coming up with good ideas, not more people exhorting others to hurry up and come up with those ideas for them.
It's not that they don't like the software, it's that they perceive the author to be an asshole.
People who make dumb choices like that aren't very relevant. You'd have to be an idiot to avoid good code just because you don't like the guy who wrote it. Now, avoiding code because it's not good -- no comments or explanations; single-character variable names, cryptic functions names, and no white space; crude and outdated designs that will never get new features -- that makes sense. It's also a good reason not to like the author. But if the code were good, it would be dumb not to use it in spite of its author being an asshole.
This phrase really grates on my nerves. Game machines, PVRs, PDAs, etc, are not powered by Linux. They're powered by electricity. They run Linux among other software.
And while we're on the subject of word misusage -- Slashdot is for nerds, not geeks. Geeks are sideshow performers that do entertain people by doing disgusting things. Technophilic misfits are nerds. It's right there on the Slashdot logo ("News for Nerds"), but people persistently refer to nerds as "geeks" in the postings.
The really bizarre thing about SCO v IBM is that SCO/Caldera seem to be claiming that the Unix copyright would have been worth at least $3 billion if only those meddling kids at IBM hadn't underhandedly turned Linux overnight into a Unix-killer. Why on earth should anybody think Unix is worth $3 billion or more? How much did Caldera pay for it? How much did (old)SCO before them pay? Novell? I'll bet each time it was sold for 1/10 as much as the previous time. Caldera couldn't possibly have paid more than a few million for it; they simply don't have the cash reserves. Unix an antique, a historical curiosity; no one would buy it except for sentimental reasons or for the prestige of saying, "Remember Unix? Well, we own that now!" And the price tag reflects that. Seems like a no-brainer.
You mean all the comments that call Darl &co a gang of homicidal maniacs attempting to hold the computer industry hostage with lies and fraud while they get rich on a sleazy stock market scam? Is that the kind of credibility you're talking about? Which dictionary do you use?
In general that's true, because they're human. Most humans (technical or not) are lousy administrators.
Everyone throws a tantrum occasionally. You have to pay big bucks to make sure they only do it on their own time. Even then it's a good idea to distribute the responsibility a bit, so when one administrator throws a tantrum on company time there's still another one to keep things running smoothly.
You see the big public tantrums mostly in volunteer efforts where nobody is paying anything and there is no backup, because it's really not all that important. What do you expect for nothing?
1. Possible infinite loops. MTA#1 connects to MTA#2 via SMTP; MTA#2 calls back MTA#1 to check for an open relay; MTA#1 calls back to MTA#2 to check for an open relay; ad infinitum. Easy enough to code around, but it means yet more complexity and obscure failure modes for already-bloated MTA code.
2. Most large organizations use different MTAs for sending and receiving, so connecting back to a sending MTA won't necessarily get you anything.
3. Very little spam is coming through open relays any more. Most of it is coming directly from throw-away accounts and hijacked PCs. Look in the headers of the next few spams you get and see how many Received lines there are -- it's probably just some residential ADSL in Mexico City connecting straight to your ISP's MTA.
For that matter, do you think access fees cover the cost of the backbone? If the entire Internet were paid for by access fees, everyone's connection would easily cost double or triple what it does now.
I don't get it. If access fees don't cover the cost of the backbone, then what does? Obviously the costs are being covered somehow.
Unfortunately a DDoS big enough to knock some guy and his puny ADSL all the way off the net is not necessarily big enough even to get fbi.gov's attention. The FBI probably has a lot of bandwidth.
No! False negatives are worse than false positives, assuming the blocking system is implemented correctly (with SMTP 550 rejection). False negatives mean more spam coming in. More spam coming in means a greater chance of the recipient accidentally deleting or ignoring non-spam. SMTP 550 rejection gives instant feedback to the sending MTA, which means guaranteed feedback to a legitimate sender. If the 550 text is sufficiently informative the rejection can be worked around quickly and easily, but at the very least the sender knows right away that their message was not delivered -- they get a delivery failure message from their MTA next time they check their mail -- and they can immediately try some alternate means of contact, e.g. telephone.
If your mailbox is full of spam and you overlook a legitimate email as a result, nobody ever knows what happened. The sender has no idea you never saw his message, and doesn't know to try an alternate means of contacting you.
This (and the bandwidth savings) is why I greatly prefer full SMTP blocking over sort-and-file spam filters. YMMV.
Speakeasy's name servers were returning NXDOMAIN instead of sitefinder by the 17th. Maybe earlier but that was when I first checked. No discussion announcement as far as I know, they just did the right thing quietly and with impressive alacrity.
Same here. I had one seller decide that my transaction with him was an "opt-in" for his monthly advert spam, but he LARTed easily enough. I've never gotten any random MakePenisFast or WindowsInfectionDuJour spam on my Ebay-only address.
Well, he didn't write one big check for $25 billion. He's been steadily giving it away for quite a while now. His % ownership of MSFT has gone way, way down in the last 10 years. Why do you think this would make huge headlines? Maybe you have to flip past the front page to read about it. Is Salon bribed or paid for? How about the BBC? What kind of proof would satisfy you?
Ha, got you beat there. We had one on our AT (286) in 1986. Mouse Systems optical, with the metal grid pad. We had lots of funky hardware, including a variety of digitizer pads and video frame grabbers. We made biomedical image processing software... our code was written in C, 8086 asm, and... wait for it... QuickBasic! Ah, memories.
Slashdot Mirror
They're malignant idiots. Spammers are stupid. Really. Yes, even stupid people can steal successfully... for a while.
BTW, your Gbps ethernet probably has round-trip latency closer to 0.1 ms.
Long story short: it worked fine for me. The data were 100% recovered and the drive continued to work perfectly until it was decommissioned due to obsolescence. (Anybody need a 7200 RPM 1.6 GB drive from 1993?)
A friend of mine has a few defective drives but alas he won't put up the $50 to hunt down an "identical" parts donor, so I haven't yet had the opportunity to try the board-swap operation on a newer drive. Mostly the damn things are just too reliable.
Hope this clears everything up for you.
... including 90% of the remaining 10% (unattributed corollary)
I'm utterly serious here, not trying to blow anyone off with a canned "Fix it yourself" response. If there really is a market for the Linux you imagine, and you seem very confident that there is, and that it's huge, and apparently nobody is exploiting that market... well... the world has basically handed you a fortune on a silver platter and all you have to do is grab it!
From my perspective, Linux and the distributions I use work fine just the way they are, and I see no reason to reorient them to satisfy some "market" out there. If I thought there were some obvious way Linux could work better for me, I'd change it. If I thought there were a huge market ready to buy into Linux(/GNU/etc) in a big way if only it were changed in some obvious and simple way... well, I sure wouldn't be wasting my time posting to Slashdot about it.
Maybe you feel there's a simple and obvious change that would make Linux(/etc) much better, but you're not quite competent to figure out exactly what it is? That's a common feeling. I think a lot of people feel that way about a lot of things. That doesn't necessarily mean there really is such an easy fix or that nobody competent has ever bothered to think about the problem. Most likely it's a much harder problem than it seems, and a lot of people have thought long and hard about it but come up with nothing yet that would satisfy you. What the world needs is more people coming up with good ideas, not more people exhorting others to hurry up and come up with those ideas for them.
This phrase really grates on my nerves. Game machines, PVRs, PDAs, etc, are not powered by Linux. They're powered by electricity. They run Linux among other software.
And while we're on the subject of word misusage -- Slashdot is for nerds, not geeks. Geeks are sideshow performers that do entertain people by doing disgusting things. Technophilic misfits are nerds. It's right there on the Slashdot logo ("News for Nerds"), but people persistently refer to nerds as "geeks" in the postings.
The really bizarre thing about SCO v IBM is that SCO/Caldera seem to be claiming that the Unix copyright would have been worth at least $3 billion if only those meddling kids at IBM hadn't underhandedly turned Linux overnight into a Unix-killer. Why on earth should anybody think Unix is worth $3 billion or more? How much did Caldera pay for it? How much did (old)SCO before them pay? Novell? I'll bet each time it was sold for 1/10 as much as the previous time. Caldera couldn't possibly have paid more than a few million for it; they simply don't have the cash reserves. Unix an antique, a historical curiosity; no one would buy it except for sentimental reasons or for the prestige of saying, "Remember Unix? Well, we own that now!" And the price tag reflects that. Seems like a no-brainer.
You mean all the comments that call Darl &co a gang of homicidal maniacs attempting to hold the computer industry hostage with lies and fraud while they get rich on a sleazy stock market scam? Is that the kind of credibility you're talking about? Which dictionary do you use?
Everyone throws a tantrum occasionally. You have to pay big bucks to make sure they only do it on their own time. Even then it's a good idea to distribute the responsibility a bit, so when one administrator throws a tantrum on company time there's still another one to keep things running smoothly.
You see the big public tantrums mostly in volunteer efforts where nobody is paying anything and there is no backup, because it's really not all that important. What do you expect for nothing?
2. Most large organizations use different MTAs for sending and receiving, so connecting back to a sending MTA won't necessarily get you anything.
3. Very little spam is coming through open relays any more. Most of it is coming directly from throw-away accounts and hijacked PCs. Look in the headers of the next few spams you get and see how many Received lines there are -- it's probably just some residential ADSL in Mexico City connecting straight to your ISP's MTA.
Unfortunately a DDoS big enough to knock some guy and his puny ADSL all the way off the net is not necessarily big enough even to get fbi.gov's attention. The FBI probably has a lot of bandwidth.
If your mailbox is full of spam and you overlook a legitimate email as a result, nobody ever knows what happened. The sender has no idea you never saw his message, and doesn't know to try an alternate means of contacting you.
This (and the bandwidth savings) is why I greatly prefer full SMTP blocking over sort-and-file spam filters. YMMV.
Speakeasy's name servers were returning NXDOMAIN instead of sitefinder by the 17th. Maybe earlier but that was when I first checked. No discussion announcement as far as I know, they just did the right thing quietly and with impressive alacrity.
horde = crowd
You're welcome.
Same here. I had one seller decide that my transaction with him was an "opt-in" for his monthly advert spam, but he LARTed easily enough. I've never gotten any random MakePenisFast or WindowsInfectionDuJour spam on my Ebay-only address.
Well, he didn't write one big check for $25 billion. He's been steadily giving it away for quite a while now. His % ownership of MSFT has gone way, way down in the last 10 years. Why do you think this would make huge headlines? Maybe you have to flip past the front page to read about it. Is Salon bribed or paid for? How about the BBC? What kind of proof would satisfy you?
The rest of that post is right on target though.
Actually if you enter QUIT at any prompt you get disconnected.
Ha, got you beat there. We had one on our AT (286) in 1986. Mouse Systems optical, with the metal grid pad. We had lots of funky hardware, including a variety of digitizer pads and video frame grabbers. We made biomedical image processing software... our code was written in C, 8086 asm, and... wait for it... QuickBasic! Ah, memories.