The RealTek DMA system is a joke. Tiny failures aren't tolerated, nor do the two popular drivers handle buffer overfills -- a bad network cable is enough to destabilize a system.
Still, it was a hard lock, no screen blanking. (Or are you sure it hasn't locked up while the console screen blanker was active?)
Grabbing a DEC Tulip-based card will increase performance and stability both, and you can get one for $5-15 if you have a decent shop near you.
I'm afraid you won't get much more than anecdotal evidence on this question. Here's mine.
We had a bad network adapter which would fail when other DMA devices were busy. This meant that whenever disk I/O was heavy, using the network adapter was likely to cause a complete system lockup. This took a while to diagnose as the problems took upward of two weeks to reproduce.
Despite the equivalent of having the power cable yanked randomly a dozen times when the machine was at its busiest, we never had a single problem with Reiser. The file which was being written to existed as the old version, and there wasn't even a lengthy fsck. Integrity was 100%.
The best way to handle these people...
on
The Economics of Spam
·
· Score: 5, Insightful
Absolutely the best way to handle these people is to consume their resources. The most easily diminished is time.
Visit a spammer's website and gather some contact information, then fire off an email. Don't be shy about including your phone number, suggesting you might be interested in mass mailing.
A couple minutes with pen and paper and you can probably come up with enough questions to keep them busy for an hour, asking about the effectiveness of their marketing technique, options, haggling on payment, so on and so on. If this type thinks there's any chance of completing the transaction, they will stay on the line for a long time. Never tell them off, leave them constantly wondering if you're another perspective client.
It's not dull. You learn quite a few things about the type of person who will do something like this. It's an insight into a pretty twisted world, and it's several million spams they won't get out.
I'd run the folding project, but the source is closed. I've got a lot of machines with spare cycles, but I sure don't want to grab some bit of code and run it 24/7 if I'm not free to peek under the hood in exchange.
We get a lot of long recorded calls in the area which start playback as soon as the line is picked up. This means that Privacy Manager thinks it's got a live human on the calling end, records part of the message as the Privacy Manager name identification and rings on through with that.
I thought this was particularly choice spin, leaving reporters without any catchy quotes:
When contacted for a response, Microsoft's PR firm responded, "Thank you for calling us with your story about the MS Reader. At this time, we are choosing not to participate in this particular
opportunity, but wish you well."
This is very cool; but does it scale to multiple CPU systems? More and more, SMP, split-bus and multi-core architectures are going to be taking over. If this holds up in those environments, Linux may actually have a leg up on some of the dedicated task heavyweights.
It's not the distro that's failing, it's either the sysadmin or the network. (or perhaps the hardware)
If he's purely talking about web hosting, the question doesn't seem to make much sense.
For a dedicated server, I know that I'd prefer Debian, however. Because it's better? No.
I've got several years wrapped up in Debian and FreeBSD. They're what I know, and what I trust myself most to secure and, more importantly, to administer remotely without botching and needing to go back to a tape or a CD reinstaller and a load of rsyncing. I think in this context, the question is quite valid.
I've been using Debian at home for 4 years now, and I think their security team does an excellent job! Especially wrt not upgrading a package version unless it is absolutely necessary. (Then they dropped the ball with SSH a few weeks ago.;)
Did I miss something? They backported changes on one problem, avoided a second problem by not having upgraded to the version with a new vulnerability like everyone else in favor of backporting a patch, then leapt to the newest version when the last hole appeared which was declared as majorly bad but (at that point) undisclosed.
What is up with all the negative articles on MS Hardware?
Take a look at your mouse! Have you ever used an MS Joystick?
Sure, lots of you have complaints with MS's OS's. Some of you hate the whole "Major Corportation" thing.
But, damnit, MS MAKES GOOD HARDWARE!
I've said many a time that I look forward to the day when Microsoft is fondly remembered as "that rather nice mouse manufacturer."
In the few areas where we're suffering (compile times, render times, etc.), a faster CPU doesn't do much good because it spends most of its time starved on the bus.
More focus on taking care of this problem is what's needed. Past that, I don't want a faster CPU. I want more level 1 cache.
Or you could patch your HTTPS server. Or, if you're not using HTTPS, you could turn it off and/or block port 443. Wouldn't either of those be easier?
The list above isn't for this attack only. This is for all ongoing bits of bad or not-yet-explained heavy traffic (frequent probes, etc) as reported by others.
You can get a current list of the top C networks which are participating in attacks of various sorts from dshield.org. Depending on your application, it may be advantageous to just add a cron job which grabs this and feeds it to your firewall rules, hosts.deny or access control lists.
"Intel anticipates that PC vendors will bundle the media adapters with multimedia PCs in order to allow consumers to deliver music and video to their entertainment centers from their PCs."
Ok, aside from it not using the PCs CPU horsepower, how is this altogether different from a really long set of A/V cables? (or a 900mhz broadcaster?)
It's more like having two devices which can use DVDs off the same shelf. The key thing here is that multiple people can use it at the same time, and you've got local control in both places.
Is this really necessary? Personally I'd think a much better approach would be to simply set up test accounts (not with.gov, but I mean on AOL, local ISPs, etc) and reference the email on a couple of webpages, and perhaps in a usenet posting. They will, without any doubt, very quickly get every spam that everyone else gets
I have four different primary email accounts between work accounts, and personal and general 'net use. All four get a lot of spam, but there's very little overlap in the spam they get.
If there are thousands of spammers with thousands of different lists, it would take more than what you mention in order to get a sample of every spam.
These guys give you a nice, professional front end for selling tunes, they create sampler discs, host music samples, and what's more - they'll even help you set yourselves up to take credit cards for your merchandise (and/or ticks) at your gigs.
They're 100% RIAA-free so far as I know, and there's a LOT of good stuff there.
the amount of mass we would have to move to be detectable at a significant distance would be prohibitive
Well, perhaps a very finely-tuned vibration of a specific type could show up in a Fourier spread, even if the power were incredibly low. But even so.....
It goes both ways though. If you can move something large back and forth quickly to transmit via gravity, then you also endure resistance (energy loss) remotely.
This would mean faster-than-light energy transmission. Again, though... at very low magnitude.
People keep responding to the transfer speed as if you have to hover within 30 feet of the computer and wait almost a minute between shots.
The thing ships with a 16mb memory stick, and can take larger sticks. It also has USB and a cradle for faster transfers. 47 seconds is also for the largest format picture. It can also send video at several frames per second, or a VGA resolution snapshot in under two seconds.
Backing up, the point of Bluetooth isn't Raw Speed. The point of including Bluetooth in a device like this is automation: As soon as you come near the proper PC, this and the PC will detect each other and begin the exchange. You might not have taken the camera out of your pocket or done more than set it down on walking in the door before it finishes the transfer.
If you need the pictures more quickly, simply set it in the USB cradle, or pop out the memory stick and stick it in one of those PC drive bay memory stick adapters.
Later on, you'll be able to configure your Bluetooth-enabled cell phone as a conduit, so pictures can automatically ride a secure tunnel back to your machine wherever you are, giving you an effectively infinite amount of space for your pictures. That's what Bluetooth is for.
The goal: when a key is held down, auto-repeat at say 5 chars/sec.
The "correct" solution: start a 200 msec timer that triggers an interrupt, allowing the CPU to do other things in the meantime.
The MSDOS solution: stay in an infinite loop until 200 msec is up.
The above, about MS-DOS, is not entirely correct. The reason your laptop probably got hot is that original versions of MS-DOS didn't have a way to tell the BIOS that it was idle, and so it would sit in a loop waiting for anything to happen. I believe BIOS calls for CPU idle weren't added until after Toshiba introduced the first laptop with software speed switching and realized they could extend battery life by throttling back to the lower clockrate. (Again - by memory - I believe this was before the x86 series had support for throwing away cycles in a power-efficient way). The looping definitely wasn't happening in order to handle keyboard repeating, however...
MS-DOS has always relied on the keyboard to generate repeating itself. Try this (under DOS, not Windows) - set the keyboard repeat to be very fast. Now unplug and reinsert the keyboard so it loses power. Your keyboard will reset, and the repeat rate will be slow again. Similarly, many BIOSes let you set the keyboard repeat at boot time, and that setting is preserved by the keyboard itself.
Interrupt 9 is fired once for the key down, once for each repeat, and once for key up. MS-DOS services this interrupt directly.
Modern Windows versions only watch the key down and key up. I don't think MS-DOS even looked at the key up signals for anything but the modifier keys. This is why you can hold down control, unplug the keyboard, and release and still have control active until you tap it again with any of these OSes, but you can only punch an alphanumeric key, remove, release and reinsert, and see it repeat under modern Windows.
If you dive into the free shell accounts they're offering, you might want to spend a little time here. It's the master documentation site for all your OpenVMS needs.
This seems to be the best guide for a user who's never even looked at VMS before.
Transcoding to.Ogg results in quite a bit of degredation. This happens when you move from one lossy format to another, -usually-.
I don't know a whole lot about MP3 encoding, but I understand it relies on Fourier series parameters to reconstruct audio, and the patents apply to the methods used to compress those parameters. A transcoder which decoded down to the series of parameters, then compressed the series differently might not involve those same patents.
If the Ogg Vorbis folks could implement something which relies on the same types of data as MP3, but which was executed in such a way that it transcodes losslessly, they could include this in the ogg standard, perhaps calling it "Degraded Mode.Ogg" or similar. Being able to quickly and losslessly convert existing material to the resulting umbrella standard would do wonders for the adoption rate of true.Ogg for new files.
Slashdot Mirror
The RealTek DMA system is a joke. Tiny failures aren't tolerated, nor do the two popular drivers handle buffer overfills -- a bad network cable is enough to destabilize a system.
Still, it was a hard lock, no screen blanking. (Or are you sure it hasn't locked up while the console screen blanker was active?)
Grabbing a DEC Tulip-based card will increase performance and stability both, and you can get one for $5-15 if you have a decent shop near you.
We had a bad network adapter which would fail when other DMA devices were busy. This meant that whenever disk I/O was heavy, using the network adapter was likely to cause a complete system lockup. This took a while to diagnose as the problems took upward of two weeks to reproduce.
Despite the equivalent of having the power cable yanked randomly a dozen times when the machine was at its busiest, we never had a single problem with Reiser. The file which was being written to existed as the old version, and there wasn't even a lengthy fsck. Integrity was 100%.
Visit a spammer's website and gather some contact information, then fire off an email. Don't be shy about including your phone number, suggesting you might be interested in mass mailing.
A couple minutes with pen and paper and you can probably come up with enough questions to keep them busy for an hour, asking about the effectiveness of their marketing technique, options, haggling on payment, so on and so on. If this type thinks there's any chance of completing the transaction, they will stay on the line for a long time. Never tell them off, leave them constantly wondering if you're another perspective client.
It's not dull. You learn quite a few things about the type of person who will do something like this. It's an insight into a pretty twisted world, and it's several million spams they won't get out.
I'd run the folding project, but the source is closed. I've got a lot of machines with spare cycles, but I sure don't want to grab some bit of code and run it 24/7 if I'm not free to peek under the hood in exchange.
We get a lot of long recorded calls in the area which start playback as soon as the line is picked up. This means that Privacy Manager thinks it's got a live human on the calling end, records part of the message as the Privacy Manager name identification and rings on through with that.
PR is media hacking.
This is very cool; but does it scale to multiple CPU systems? More and more, SMP, split-bus and multi-core architectures are going to be taking over. If this holds up in those environments, Linux may actually have a leg up on some of the dedicated task heavyweights.
If he's purely talking about web hosting, the question doesn't seem to make much sense.
For a dedicated server, I know that I'd prefer Debian, however. Because it's better? No.
I've got several years wrapped up in Debian and FreeBSD. They're what I know, and what I trust myself most to secure and, more importantly, to administer remotely without botching and needing to go back to a tape or a CD reinstaller and a load of rsyncing. I think in this context, the question is quite valid.
Did I miss something? They backported changes on one problem, avoided a second problem by not having upgraded to the version with a new vulnerability like everyone else in favor of backporting a patch, then leapt to the newest version when the last hole appeared which was declared as majorly bad but (at that point) undisclosed.
It seemed like the perfect set of maneuvers.
In the few areas where we're suffering (compile times, render times, etc.), a faster CPU doesn't do much good because it spends most of its time starved on the bus.
More focus on taking care of this problem is what's needed. Past that, I don't want a faster CPU. I want more level 1 cache.
The list above isn't for this attack only. This is for all ongoing bits of bad or not-yet-explained heavy traffic (frequent probes, etc) as reported by others.
You can get a current list of the top C networks which are participating in attacks of various sorts from dshield.org. Depending on your application, it may be advantageous to just add a cron job which grabs this and feeds it to your firewall rules, hosts.deny or access control lists.
Then go check out cdbaby.com or one of the other marvelous independent artist distribution sites!
Still, the DRM's as much of a bitch as ever.
I have four different primary email accounts between work accounts, and personal and general 'net use. All four get a lot of spam, but there's very little overlap in the spam they get.
If there are thousands of spammers with thousands of different lists, it would take more than what you mention in order to get a sample of every spam.
These guys give you a nice, professional front end for selling tunes, they create sampler discs, host music samples, and what's more - they'll even help you set yourselves up to take credit cards for your merchandise (and/or ticks) at your gigs.
They're 100% RIAA-free so far as I know, and there's a LOT of good stuff there.
On the down side, they're making me go broke.
It goes both ways though. If you can move something large back and forth quickly to transmit via gravity, then you also endure resistance (energy loss) remotely.
This would mean faster-than-light energy transmission. Again, though... at very low magnitude.
The thing ships with a 16mb memory stick, and can take larger sticks. It also has USB and a cradle for faster transfers. 47 seconds is also for the largest format picture. It can also send video at several frames per second, or a VGA resolution snapshot in under two seconds.
Backing up, the point of Bluetooth isn't Raw Speed. The point of including Bluetooth in a device like this is automation: As soon as you come near the proper PC, this and the PC will detect each other and begin the exchange. You might not have taken the camera out of your pocket or done more than set it down on walking in the door before it finishes the transfer.
If you need the pictures more quickly, simply set it in the USB cradle, or pop out the memory stick and stick it in one of those PC drive bay memory stick adapters.
Later on, you'll be able to configure your Bluetooth-enabled cell phone as a conduit, so pictures can automatically ride a secure tunnel back to your machine wherever you are, giving you an effectively infinite amount of space for your pictures. That's what Bluetooth is for.
More details here for Japanese speakers.
The above, about MS-DOS, is not entirely correct. The reason your laptop probably got hot is that original versions of MS-DOS didn't have a way to tell the BIOS that it was idle, and so it would sit in a loop waiting for anything to happen. I believe BIOS calls for CPU idle weren't added until after Toshiba introduced the first laptop with software speed switching and realized they could extend battery life by throttling back to the lower clockrate. (Again - by memory - I believe this was before the x86 series had support for throwing away cycles in a power-efficient way). The looping definitely wasn't happening in order to handle keyboard repeating, however...
MS-DOS has always relied on the keyboard to generate repeating itself. Try this (under DOS, not Windows) - set the keyboard repeat to be very fast. Now unplug and reinsert the keyboard so it loses power. Your keyboard will reset, and the repeat rate will be slow again. Similarly, many BIOSes let you set the keyboard repeat at boot time, and that setting is preserved by the keyboard itself.
Interrupt 9 is fired once for the key down, once for each repeat, and once for key up. MS-DOS services this interrupt directly.
Modern Windows versions only watch the key down and key up. I don't think MS-DOS even looked at the key up signals for anything but the modifier keys. This is why you can hold down control, unplug the keyboard, and release and still have control active until you tap it again with any of these OSes, but you can only punch an alphanumeric key, remove, release and reinsert, and see it repeat under modern Windows.
This seems to be the best guide for a user who's never even looked at VMS before.
And no - this isn't tipping anyone off there... slashdot.org seems to be blocked too.
To be fair, I don't remember people being this kind to MS when it shipped mere drivers which only worked with the Microsoft Mouse.
I don't know a whole lot about MP3 encoding, but I understand it relies on Fourier series parameters to reconstruct audio, and the patents apply to the methods used to compress those parameters. A transcoder which decoded down to the series of parameters, then compressed the series differently might not involve those same patents.
If the Ogg Vorbis folks could implement something which relies on the same types of data as MP3, but which was executed in such a way that it transcodes losslessly, they could include this in the ogg standard, perhaps calling it "Degraded Mode .Ogg" or similar. Being able to quickly and losslessly convert existing material to the resulting umbrella standard would do wonders for the adoption rate of true .Ogg for new files.