If one rogue nation chooses to defy UN law, there's not too much they can do...
That doesn't mean that nothing can be done... I and everyone else with a firewall in front of a mailserver can blackhole those nations that choose to tolerate spam.
I can't wait for IPv6... It should be even easier to throw away traffic from entire nations than it is now.
Which is why there were stories a couple years ago of car thieves with IR equipped PalmOS devices that would brute-force those systems.
The newer garage door systems use some sort of bidirectional handshake system. The car keyless entry folks could do the same if they really cared. Then it would be virtually impossible to fake it (even a replay attack would be no good). Such systems would fail-closed, of course. This is why they hide a button somewhere under the dash that lets you open the door with the key, setting off the alarm and activating the ignition interlock, then put the key in the ignition, turn it on, then find and hit the hidden button. That deactivates the alarm and the interlock and you're good to go.
For instance, how easy is it to find a driver for video card abc, built in audo chipset def, a DVD player compatible with DVD ghi, etc?
Well, specifically, of those examples, only the sound card requires a specific kernel driver. The video card will require an XFree86 driver. SCSI and ATAPI optical drives don't require specific drivers.
The hardware support situation is not different than Windows except for one thing.
If the piece of hardware in question is either reasonably generic or reasonably old, the OS (in this case of *nix, it's the kernel for most devices and XFree86 for video cards) has support for it built-in. For brand new or bizarro devices, you must count on the manufacturer to supply a driver.
That's where the one difference comes in: Most manufacturers of hardware devices that require their own drivers tend to be Windows-only devices because of the Windows monopoly. In some cases, the hardware has been reverse-engineered by tireless geeks, and in some other cases the OS has grown shims that actually simulate the Windows kernel environment to make Windows drivers actually work (there are NDIS shim systems for network cards). In other cases, the hardware simply winds up staying unsupported.
Of course, the exact same situation plays out with the Mac. As soon as *nix gets a reasonable desktop installed base, you'll start seeing "Works with Linux" stickers on more and more stuff.
Is there an 18.5 hour difference between setting up XP Home and a Linux desktop distro?
Setup alone doth not TCO make. But setting up an XP machine, in my experience, does take significantly longer than a *nix (I actually use FreeBSD) machine, given that the Windows installer nowadays takes longer than most *nix installers, and *then* you have to download and install all of the service packs and patches. Then you have to buy and install anti-virus and anti-malware utilities and maintain them... This all presumes you already have a firewall in place. Windows machines that are installed without being behind a firewall often are 0wn3d before the patch download and installation can be completed.
But if you really want to win the TCO game, buy a mac. Virtually all of the TCO is the purchase price, compared to (other) *nix or Windows.
Using a raw partition typically means bypassing the filesystem code in the OS. Since most databases simply consist of a small number of large files that are randomly accessed by the database system, the overhead of the filesystem is unnecessary. Not having a filesystem between your database and the disk also means faster crash recovery - there's no need to run FSCK on the (largely irrelevant) filesystem AND run a database consistency check - you can jump right to the latter.
You're right about this being for dedicated postgres boxes, but then dedicated database machines are exactly what you find in large enterprises. The "dot com" I work for has a big iron Sun running Oracle and nothing else, and a large number of smaller machines that do the "everything else". I think you'll find that fairly typical.
Make no mistake: I am both a PostgreSQL user and a major advocate for it where I work. At the same time, I would make one little comment on this:
Most of the arguable weaknesses in Postgres are in the more esoteric high-end feature space
Some of those esoteric features are things like clustering/failover, which in my view aren't really so esoteric. Yes, I do know that there is third party support for it, but it isn't free.
Raw partition support would also be a good checkbox in the 'enterprise ready' table.
Hands up all those surprised to hear this coming out of someone associated with Microsoft. Perhaps the phrase "Applications barrier to entry" rings a bell?
- Run it through hymn and dedrm and suddenly there are no more limitations at all.
But of course that would also require adding to the physical CD section...
- Careful use of either a sharpie or your PC's shift key and suddenly there are no more limitations at all.
And as long as we're categorizing plusses and minuses, you left out the CD booklet/liner notes as a plus for the physical version. I don't know if I'd suggest that it's worth $3.50 + shipping, but it is at least a + compared to the iTMS.
The idea of running alternative firmware in a netgear or linksys type box is intriguing to me for one reason in particular: The idea that one could very easily add IPv6 support (via 6to4). So long as there is an IPv6 firewall that's reasonably configurable, this would make it a lot easier for me to support my folks. Instead of having to perform rather unnatural acts with the router to get to the correct machine from the outside, I could just ssh/rdp/vnc directly.
The one downside is that if your outside IPv4 address changes, then you must propagate a new prefix to the inside nodes, but even that doesn't sound like a real show stopper.
However, that doesn't change the fact that the mass market seems to have selected against them,
You're changing your criteria (probably because you want to define away their success). You said "Fast," "Good," and making money. They are doing all three. QED.
The problem of swap containing sensitive data from running programs was addressed some time ago by OpenBSD. They generate a random key at boot time and use it to encrypt reads and writes to swap. By definition, you are not interested in the contents of swap the next time you boot up, so you can start with a brand new key. Not only is swap space secure against fishing expeditions like in TFA, but it's also secure against someone getting read privileges on the raw disk (unless they also get permissions on kernel memory and can go look up the key).
Not only do you need to be able to see the receiver for line-of-site, but you actually need to have a sort of cigar or sausage-shape area, with the diameter at the center being measured in meters, with no obstructions. That is, merely being able to spot the receiver through the gap between a couple of tree branches won't do.
It's not DOS. It's XP. Watch sometimes as it goes through all of the module loads just after you boot from the CD. One of the things it loads is the NT kernel. During one of the reboots between the time it writes the embryonic installation FS and it starts the GUI install procedure, it upgrades the FAT filesystem to NTFS (if you told it to). If you install with FAT and don't upgrade at install time, there is a filesystem migrator which will reboot and do the same thing - upgrade the FAT filesystem to NTFS.
Here's the problem I have with mail.app's spam filtering:
I have several macs, and an IMAP server. The simple fact is that Mail.app doesn't share the filtering database. So the training winds up being sort of haphazard.
I suppose I should designate a particular machine to be the spam filtering IMAP client and have the rest of them not participate, but then I can't train on those subservient machines.
It'd be much better if multiple Mail.app IMAP clients could store their database on the server and share it.
Is it just me, or can't you just burn the cd, then rip it into mp3?
Sure. But when you do so, you lose quality. For one thing, some of the songs at the itms come from 24 bit masters. So immediately when you turn them into 16 bit CD audio tracks, you throw away a lot of potential dynamic range. That's beside the fact that decoding the AAC and then re-encoding it to either AAC or MP3 will lower the audio quality, AND you'll lose all the meta data.
The real problem with the story told in the article is that there was no firewall between the system and the Internet. It is simply no longer acceptable to connect Windows machines directly to the internet without a firewall.
That statement doesn't really change the conclusions in the article very much, but in the past I've reinstalled friends' windows machines and downloaded and installed all the updates without any trouble at all -- because I did it behind a firewall.
If you wouldn't leave your car parked unlocked with all the windows down in the middle of [bad part of town], then don't connect Windows machines up to the Internet without a firewall. The end.
I love my '96 Passatt TDI. No matter how or where I drive, with or without AC, I get 35 mpg. It never wavers at all. And it's got surprising pickup for something officially rated at 90 hp. Of course, the backside of that pickup is that you need to shift before the end of the intersection.:-)
It's exempt from emissions testing too, which is a big plus.
My next car will probably be a Beetle TDI as soon as they have factory installed XM radios.
Of all the people I know with PowerBooks, absolutely none of them have ever used a PC Card in one.
Then you don't know me. I have a Linksys WPC54G in my TiBook. This gives me much better reception than the built-in antenna and an upgrade to 802.11g using Apple's own AirportExtreme driver kext.
The law you're looking for is the AHRA - the American Home Recording Act. It was the step 1 in the two step shuffle of which the DMCA was step 2. The AHRA is what mandates serial copy management for non-computer-peripheral digital multimedia devices. It is why standalone CD burners must use "audio" CD-Rs which differ only in having an extra bit turned on that indicates you paid a tax on them.
The one thing the AHRA gave back was it formalized the legality of format shifting.
But as is the case with ordinary copyright law, it does not apply where the use of the media is covered by a more restrictive contract. In this case, the iTunes license agreement that everyone clicked 'agree' on before iTunes would start up and let them start buying tracks.
That's what's different about CDs versus m4p files. CDs do not come in shrink-wrap licenses. At least, not yet.
In my neighborhood, there is a light that has a hundred yards back from the light that causes it to go yellow. There's no connection with your speed, of course. They just don't want anyone to NOT have to stop. For those who want to know, it's at Newhall and Monroe just outside Santa Clara. I'm sure it's not unique. I'd be overjoyed if someone sabotaged it someday.
I would take Pruneridge as an alternate route, but there is a nasty radar trap at the I-880 undercrossing. The revenuers make a mint off it.
All of this could be avoided if they would conduct honest traffic surveys and not set speed limits lower than the 85th percentile, as proper traffic engineering practice dictates. But though speed limits were originally designed as a means to control unreasonable law enforcement (before speed limits, the cop could decide anything over 15 mph was not "reasonable or prudent"), they clearly have strayed from that purpose.
Slashdot Mirror
That doesn't mean that nothing can be done... I and everyone else with a firewall in front of a mailserver can blackhole those nations that choose to tolerate spam.
I can't wait for IPv6... It should be even easier to throw away traffic from entire nations than it is now.
Which is why there were stories a couple years ago of car thieves with IR equipped PalmOS devices that would brute-force those systems.
The newer garage door systems use some sort of bidirectional handshake system. The car keyless entry folks could do the same if they really cared. Then it would be virtually impossible to fake it (even a replay attack would be no good). Such systems would fail-closed, of course. This is why they hide a button somewhere under the dash that lets you open the door with the key, setting off the alarm and activating the ignition interlock, then put the key in the ignition, turn it on, then find and hit the hidden button. That deactivates the alarm and the interlock and you're good to go.
Well, specifically, of those examples, only the sound card requires a specific kernel driver. The video card will require an XFree86 driver. SCSI and ATAPI optical drives don't require specific drivers.
The hardware support situation is not different than Windows except for one thing.
If the piece of hardware in question is either reasonably generic or reasonably old, the OS (in this case of *nix, it's the kernel for most devices and XFree86 for video cards) has support for it built-in. For brand new or bizarro devices, you must count on the manufacturer to supply a driver.
That's where the one difference comes in: Most manufacturers of hardware devices that require their own drivers tend to be Windows-only devices because of the Windows monopoly. In some cases, the hardware has been reverse-engineered by tireless geeks, and in some other cases the OS has grown shims that actually simulate the Windows kernel environment to make Windows drivers actually work (there are NDIS shim systems for network cards). In other cases, the hardware simply winds up staying unsupported.
Of course, the exact same situation plays out with the Mac. As soon as *nix gets a reasonable desktop installed base, you'll start seeing "Works with Linux" stickers on more and more stuff.
Setup alone doth not TCO make. But setting up an XP machine, in my experience, does take significantly longer than a *nix (I actually use FreeBSD) machine, given that the Windows installer nowadays takes longer than most *nix installers, and *then* you have to download and install all of the service packs and patches. Then you have to buy and install anti-virus and anti-malware utilities and maintain them... This all presumes you already have a firewall in place. Windows machines that are installed without being behind a firewall often are 0wn3d before the patch download and installation can be completed.
But if you really want to win the TCO game, buy a mac. Virtually all of the TCO is the purchase price, compared to (other) *nix or Windows.
Using a raw partition typically means bypassing the filesystem code in the OS. Since most databases simply consist of a small number of large files that are randomly accessed by the database system, the overhead of the filesystem is unnecessary. Not having a filesystem between your database and the disk also means faster crash recovery - there's no need to run FSCK on the (largely irrelevant) filesystem AND run a database consistency check - you can jump right to the latter.
You're right about this being for dedicated postgres boxes, but then dedicated database machines are exactly what you find in large enterprises. The "dot com" I work for has a big iron Sun running Oracle and nothing else, and a large number of smaller machines that do the "everything else". I think you'll find that fairly typical.
Most of the arguable weaknesses in Postgres are in the more esoteric high-end feature space
Some of those esoteric features are things like clustering/failover, which in my view aren't really so esoteric. Yes, I do know that there is third party support for it, but it isn't free.
Raw partition support would also be a good checkbox in the 'enterprise ready' table.
Hands up all those surprised to hear this coming out of someone associated with Microsoft. Perhaps the phrase "Applications barrier to entry" rings a bell?
Under the iTunes section you left out...
- Run it through hymn and dedrm and suddenly there are no more limitations at all.
But of course that would also require adding to the physical CD section...
- Careful use of either a sharpie or your PC's shift key and suddenly there are no more limitations at all.
And as long as we're categorizing plusses and minuses, you left out the CD booklet/liner notes as a plus for the physical version. I don't know if I'd suggest that it's worth $3.50 + shipping, but it is at least a + compared to the iTMS.
The idea of running alternative firmware in a netgear or linksys type box is intriguing to me for one reason in particular: The idea that one could very easily add IPv6 support (via 6to4). So long as there is an IPv6 firewall that's reasonably configurable, this would make it a lot easier for me to support my folks. Instead of having to perform rather unnatural acts with the router to get to the correct machine from the outside, I could just ssh/rdp/vnc directly.
The one downside is that if your outside IPv4 address changes, then you must propagate a new prefix to the inside nodes, but even that doesn't sound like a real show stopper.
Has anyone set up something like this?
You're changing your criteria (probably because you want to define away their success). You said "Fast," "Good," and making money. They are doing all three. QED.
Been there, done that.
The problem of swap containing sensitive data from running programs was addressed some time ago by OpenBSD. They generate a random key at boot time and use it to encrypt reads and writes to swap. By definition, you are not interested in the contents of swap the next time you boot up, so you can start with a brand new key. Not only is swap space secure against fishing expeditions like in TFA, but it's also secure against someone getting read privileges on the raw disk (unless they also get permissions on kernel memory and can go look up the key).
Too bad more systems don't embrace the idea.
It's also why IBM died. Oh wait.
lose money, bad.
This lesson brought to you by the Kenny Blankenship and Vic Romano school of economics.
Did anyone else read that at first as "...Microsoft's shared 'all your code are belong to us' source license"?
Seems more correct, somehow.
Not only do you need to be able to see the receiver for line-of-site, but you actually need to have a sort of cigar or sausage-shape area, with the diameter at the center being measured in meters, with no obstructions. That is, merely being able to spot the receiver through the gap between a couple of tree branches won't do.
It's not DOS. It's XP. Watch sometimes as it goes through all of the module loads just after you boot from the CD. One of the things it loads is the NT kernel. During one of the reboots between the time it writes the embryonic installation FS and it starts the GUI install procedure, it upgrades the FAT filesystem to NTFS (if you told it to). If you install with FAT and don't upgrade at install time, there is a filesystem migrator which will reboot and do the same thing - upgrade the FAT filesystem to NTFS.
Here's the problem I have with mail.app's spam filtering:
I have several macs, and an IMAP server. The simple fact is that Mail.app doesn't share the filtering database. So the training winds up being sort of haphazard.
I suppose I should designate a particular machine to be the spam filtering IMAP client and have the rest of them not participate, but then I can't train on those subservient machines.
It'd be much better if multiple Mail.app IMAP clients could store their database on the server and share it.
Sure. But when you do so, you lose quality. For one thing, some of the songs at the itms come from 24 bit masters. So immediately when you turn them into 16 bit CD audio tracks, you throw away a lot of potential dynamic range. That's beside the fact that decoding the AAC and then re-encoding it to either AAC or MP3 will lower the audio quality, AND you'll lose all the meta data.
The real problem with the story told in the article is that there was no firewall between the system and the Internet. It is simply no longer acceptable to connect Windows machines directly to the internet without a firewall.
That statement doesn't really change the conclusions in the article very much, but in the past I've reinstalled friends' windows machines and downloaded and installed all the updates without any trouble at all -- because I did it behind a firewall.
If you wouldn't leave your car parked unlocked with all the windows down in the middle of [bad part of town], then don't connect Windows machines up to the Internet without a firewall. The end.
I love my '96 Passatt TDI. No matter how or where I drive, with or without AC, I get 35 mpg. It never wavers at all. And it's got surprising pickup for something officially rated at 90 hp. Of course, the backside of that pickup is that you need to shift before the end of the intersection. :-)
It's exempt from emissions testing too, which is a big plus.
My next car will probably be a Beetle TDI as soon as they have factory installed XM radios.
Then you don't know me. I have a Linksys WPC54G in my TiBook. This gives me much better reception than the built-in antenna and an upgrade to 802.11g using Apple's own AirportExtreme driver kext.
The law you're looking for is the AHRA - the American Home Recording Act. It was the step 1 in the two step shuffle of which the DMCA was step 2. The AHRA is what mandates serial copy management for non-computer-peripheral digital multimedia devices. It is why standalone CD burners must use "audio" CD-Rs which differ only in having an extra bit turned on that indicates you paid a tax on them.
The one thing the AHRA gave back was it formalized the legality of format shifting.
But as is the case with ordinary copyright law, it does not apply where the use of the media is covered by a more restrictive contract. In this case, the iTunes license agreement that everyone clicked 'agree' on before iTunes would start up and let them start buying tracks.
That's what's different about CDs versus m4p files. CDs do not come in shrink-wrap licenses. At least, not yet.
In my neighborhood, there is a light that has a hundred yards back from the light that causes it to go yellow. There's no connection with your speed, of course. They just don't want anyone to NOT have to stop. For those who want to know, it's at Newhall and Monroe just outside Santa Clara. I'm sure it's not unique. I'd be overjoyed if someone sabotaged it someday.
I would take Pruneridge as an alternate route, but there is a nasty radar trap at the I-880 undercrossing. The revenuers make a mint off it.
All of this could be avoided if they would conduct honest traffic surveys and not set speed limits lower than the 85th percentile, as proper traffic engineering practice dictates. But though speed limits were originally designed as a means to control unreasonable law enforcement (before speed limits, the cop could decide anything over 15 mph was not "reasonable or prudent"), they clearly have strayed from that purpose.