BTW, DSLAM means DSL Access Multiplexor. These are where all the DSL lines terminate...
At least they don't insisit on incorrectly calling it a "modem", like the devices at the customer's endpoint which are usually ATM bridges or often even full routers.
What I'm a lot less clear on is whether "line sharing" will still be OK - right now, for instance, my Speakeasy service is operating split on my Verizon line, via equipment co-located at my CO. Will that state of affairs continue, or will Speakeasy have to lease a wire from Verizon in it's entirety? The other variable will be what happens when Verizon gets FIOS deployed - will there still be a place for the CLECs at that point?
Well, it's in the telcos best interest if they continue to allow that... Right now I have SBC for local service and line-share Speakeasy DSL on that.
SBC may decide they want to prohibit line-sharing with CLECs, either directly or by making it too expensive for them. In that case, I'll just drop the SBC service, let Speakeasy have the line entirely to themselves, and try out that Speakeasy VoIP servce they've been advertising.
Before the telcos rejoice too much about a possible broadband monopoly, they would do well to remember that they now have competition for POTS as well. Between VoIP and cell providers it's not too difficult to cut them out of the loop entirely if they get too out of line.
Yes, you can probably tell I don't do much PC gaming these days;) With the exception of Doom 3, which I played some but didn't really enjoy very much, the last 4 major titles I've played were all console based -- Halo, KOTOR, Halo 2, and KOTOR 2. Those obviously use D3D...
There really hasn't been anything compelling lately to convince me to dual boot. Hell, last RTS I played was probably Starcraft; hooray for bnetd. I won't touch MMORPGs with a 10 foot pole, most are filled with lamers and if I want that I can get on IRC and not have to spend hours leveling.
I know the Unreal engine at least was designed so that it could be applied to more than an FPS. I'm pretty sure at least one RPG used it, but I can't remember off the top of my head which.
Thanks for the correction on Half Life, I knew it was one of the Quakes but my memory is a little fuzzy.
Retrofitting engines like Quake 3 with D3D seems like an awful waste of effort. Never had a problem with Quake 3 myself, but I guess some really cheap graphics cards don't have good opengl drivers.
The key itself hasn't been comprimised as far as I know, but ISTR at least one instance where Verisign mistakenly issued a certificate to someone claiming to be "Microsoft Corp".
That's one of the reasons for the various "Root CA Updates" showing up in windows update a while back.
I would laugh. Seriously. I don't condone the idea, but if the whole damn Internet got shut down for a few days, it would serve them right for pushing through something so monumentally idiotic.
So will they stop reinforcing Microsoft's lock-in by providing good drivers for systems other than Windows? I think NVidia provides good drivers for Linux and FreeBSD these days, but AFAIK they are the only one.
That would certainly be nice. It took forever to get nvidia drivers on FreeBSD, and they're still not available for amd64. Releases are few and far between, so it tends to lag behind the linux driver.
I'm not complaining too much since now that they're here, they work well on i386 at least. Early versions were buggy and crashed a lot, but later ones are much improved.
Of course in several cases the linux/freebsd nvidia drivers actually behave MUCH better than their windows counterparts. For example, on FreeBSD, I can swap pretty much any nvidia card for my current one and it will just work. Don't have to touch the configuration at all. On Windows it will usually go to VGA mode and demand that you reinstall the driver, even though it's supposedly a universal download.
It's also quite nice to be able to activate TwinView without a minimum of TWO reboots.
Except the big ones. How many games actually bother to write their own engine these days? Many license it from someone else, and the big ones I can think of are Unreal, Quake2/3, Source (HL2), and Doom 3.
Guess what, two of those are primarily written and optimized for OpenGL. I'm pretty sure that the Doom 3 engine doesn't support Direct3D at all. Even the Unreal engine, though it uses DirectX by default, has a pluggable architecture so that it can use OpenGL or even its own software renderer.
IIRC, Half-life used a heavily modified Quake 2 engine that MAY have supported Direct3D (or maybe it just used DirectX for sound, I can't remember). But with that much effort put into it, it might have been easier for Valve to just write their own, which they did for HL2.
A 30 fps synthesized CG such as a game does not have free motion-blur - currently it is only achievable using multisampling, which has no benefit at all over just increasing the frame rate.
The only benefit I can think of is if your framerate is higher than the refresh rate (or data bandwidth in the case of DVI/LCD) of your display device. In that case, it might be advantageous to blend consecutive frames together so that the extra rendering isn't wasted.
Go play some fast-action FPS at 80 fps, then get back down to 30. Tell me then which is easier on the eyes.
The main reason 80fps looks better is because even if sampled (more or less) 30 times a second, those samples will have more realistic motion blur. A video card rendering at 30fps has virtually no motion blur so it just looks artificial.
That's one of the reasons that movies can look quite realistic even at the low rate of 24fps, but realtime computer graphics at that frame rate would look like crap.
Apparently the reason we don't have to worry about it with regular microwaves is either the different frequency or the shielding around it.
It's both the shielding and the fact that it's not directed so it will fall off pretty quickly radiating in all or most directions (depending on how much of the shielding has been removed).
Also, I'm guessing the antenna on the roof is a parabolic dish? The focusing effect would definitely boost the concentration of the power (and make it fall off less quickly), and the water molecules in people's eyes don't react too well to being heated...
MUMPS is much older than Apache, PHP, Python, MySQL, or Linux. In the last 30 years it has probably changed less than any of the above did in their last minor point release.
PAL sucks because in order to air content that was originally 24fps progressive, the audio has to be sped up to keep it in sync and everything is too high pitched.
Our phone system's voice mail processor used to be on an OS/2 Warp 3.x box (with the GUI disabled). Thing was stable as hell, ran for years without being touched.
When we "upgraded" the phone system, it got replaced with one that runs on NT. It came preloaded with an 'at' job to reboot it nightly...
My laptop is a crappy old Compaq -- I definitely don't have enough RAM.
FreeBSD helps because the OS itself is fairly minimalistic and has a small memory footprint. It also doesn't run a lot of extraneous services by default (yeah, I could disable them on some other OS but it's easier to just add what you need).
The other part that's useful is that the VM is tunable so I can tweak it for what I want. For example, by default FreeBSD will preemptively swap a copy of idle pages sometimes. That makes sense for servers and desktops where the drive is constantly running. If the memory is used again that's fine, it's still there. If pressure on the VM increases and it needs more space, those pages can be quickly discarded from memory since they're already on the disk.
In my situation, however, that's not very desirable. It's only a simple sysctl to change the behavior to favor swapping as little as possible.
This is exactly what I do with my current laptop. 1GB flash card with the OS and most commonly used apps. The hard drive has swap (almost never touched since I'm running FreeBSD), apps that are not used very often, and large documents that won't fit on the flash. Volatile stuff like/var runs out of a ramdisk.
Works great; I have ataidle spin down the drive as soon as it boots up, and set the timeout low so that if it has to spin up for something it doesn't run very long. The thing is near silent and doesn't get as hot as it used to. I can check email, browse with firefox, and ssh to remote systems without the drive ever spinning up.
No, the solution is that your company should have an external authenticated mail relay that is included in the SPF record.
Authenticated is the key word here. Anybody who's roaming uses the company's relay. Hell, use it internally too and you don't have to change any settings while away. I've yet to come across a mail client that doesn't support SMTP AUTH, and many allow you to "use the same password that I do for checking mail" for convenience.
The mail relay should run on the submission port (587), or better yet over SSL (port 465). This gets around the port 25 blocks and transparent redirects that many brain-dead ISPs and hotels have.
This is nothing more than growing pains, and in the long run, capitalism SHOULD (unless someone screws it up) raise everyone's standards up. Even Ethiopia, so that everyone can have 5 weeks vacation, and 3 hour lunches.
That's a nice ideal, but the question is: Does the planet have enough resources for everyone to have that high of a standard of living? Don't forget there are billions of people living at a much lower standard than Americans or Eurpoeans both are used to.
Not saying that it's right or that I have an answer for it, but sometimes the universe doesn't allow for a perfect solution. It's only going to get worse as the population increases, too. If there was massive investment in space exploration and enough people were willing to relocate, maybe, but that's a longshot at best.
Depends how you set it up. On my Postfix server, it checks SPF before the message is received, but only rejects the message if the domain has an explicit "-all" in its SPF record. You'd be surprised how many domains are actually publishing SPF v1 (i.e. NOT Sender-ID) records.
Other methods involve using SPF as an additional spam score in systems like SpamAssassin. Those of course don't check until after the message is received, but can apply fuzzier maching than simple accept/reject choices.
Slashdot Mirror
BTW, DSLAM means DSL Access Multiplexor. These are where all the DSL lines terminate...
At least they don't insisit on incorrectly calling it a "modem", like the devices at the customer's endpoint which are usually ATM bridges or often even full routers.
What I'm a lot less clear on is whether "line sharing" will still be OK - right now, for instance, my Speakeasy service is operating split on my Verizon line, via equipment co-located at my CO. Will that state of affairs continue, or will Speakeasy have to lease a wire from Verizon in it's entirety? The other variable will be what happens when Verizon gets FIOS deployed - will there still be a place for the CLECs at that point?
Well, it's in the telcos best interest if they continue to allow that... Right now I have SBC for local service and line-share Speakeasy DSL on that.
SBC may decide they want to prohibit line-sharing with CLECs, either directly or by making it too expensive for them. In that case, I'll just drop the SBC service, let Speakeasy have the line entirely to themselves, and try out that Speakeasy VoIP servce they've been advertising.
Before the telcos rejoice too much about a possible broadband monopoly, they would do well to remember that they now have competition for POTS as well. Between VoIP and cell providers it's not too difficult to cut them out of the loop entirely if they get too out of line.
Yes, you can probably tell I don't do much PC gaming these days ;) With the exception of Doom 3, which I played some but didn't really enjoy very much, the last 4 major titles I've played were all console based -- Halo, KOTOR, Halo 2, and KOTOR 2. Those obviously use D3D...
There really hasn't been anything compelling lately to convince me to dual boot. Hell, last RTS I played was probably Starcraft; hooray for bnetd. I won't touch MMORPGs with a 10 foot pole, most are filled with lamers and if I want that I can get on IRC and not have to spend hours leveling.
I know the Unreal engine at least was designed so that it could be applied to more than an FPS. I'm pretty sure at least one RPG used it, but I can't remember off the top of my head which.
Thanks for the correction on Half Life, I knew it was one of the Quakes but my memory is a little fuzzy.
Retrofitting engines like Quake 3 with D3D seems like an awful waste of effort. Never had a problem with Quake 3 myself, but I guess some really cheap graphics cards don't have good opengl drivers.
The key itself hasn't been comprimised as far as I know, but ISTR at least one instance where Verisign mistakenly issued a certificate to someone claiming to be "Microsoft Corp".
That's one of the reasons for the various "Root CA Updates" showing up in windows update a while back.
That's why I built my own router.
http://www.soekris.com
Well, that's not the real reason but it's a nice fringe benefit.
DSL modem could still be bugged, but that's why I encrypt everything.
Just because you're paranoid doesn't mean they're not out to get you.
(only half-joking)
I would laugh. Seriously. I don't condone the idea, but if the whole damn Internet got shut down for a few days, it would serve them right for pushing through something so monumentally idiotic.
So will they stop reinforcing Microsoft's lock-in by providing good drivers for systems other than Windows? I think NVidia provides good drivers for Linux and FreeBSD these days, but AFAIK they are the only one.
That would certainly be nice. It took forever to get nvidia drivers on FreeBSD, and they're still not available for amd64. Releases are few and far between, so it tends to lag behind the linux driver.
I'm not complaining too much since now that they're here, they work well on i386 at least. Early versions were buggy and crashed a lot, but later ones are much improved.
Of course in several cases the linux/freebsd nvidia drivers actually behave MUCH better than their windows counterparts. For example, on FreeBSD, I can swap pretty much any nvidia card for my current one and it will just work. Don't have to touch the configuration at all. On Windows it will usually go to VGA mode and demand that you reinstall the driver, even though it's supposedly a universal download.
It's also quite nice to be able to activate TwinView without a minimum of TWO reboots.
Except the big ones. How many games actually bother to write their own engine these days? Many license it from someone else, and the big ones I can think of are Unreal, Quake2/3, Source (HL2), and Doom 3.
Guess what, two of those are primarily written and optimized for OpenGL. I'm pretty sure that the Doom 3 engine doesn't support Direct3D at all. Even the Unreal engine, though it uses DirectX by default, has a pluggable architecture so that it can use OpenGL or even its own software renderer.
IIRC, Half-life used a heavily modified Quake 2 engine that MAY have supported Direct3D (or maybe it just used DirectX for sound, I can't remember). But with that much effort put into it, it might have been easier for Valve to just write their own, which they did for HL2.
A 30 fps synthesized CG such as a game does not have free motion-blur - currently it is only achievable using multisampling, which has no benefit at all over just increasing the frame rate.
The only benefit I can think of is if your framerate is higher than the refresh rate (or data bandwidth in the case of DVI/LCD) of your display device. In that case, it might be advantageous to blend consecutive frames together so that the extra rendering isn't wasted.
Go play some fast-action FPS at 80 fps, then get back down to 30. Tell me then which is easier on the eyes.
The main reason 80fps looks better is because even if sampled (more or less) 30 times a second, those samples will have more realistic motion blur. A video card rendering at 30fps has virtually no motion blur so it just looks artificial.
That's one of the reasons that movies can look quite realistic even at the low rate of 24fps, but realtime computer graphics at that frame rate would look like crap.
Apparently the reason we don't have to worry about it with regular microwaves is either the different frequency or the shielding around it.
It's both the shielding and the fact that it's not directed so it will fall off pretty quickly radiating in all or most directions (depending on how much of the shielding has been removed).
Also, I'm guessing the antenna on the roof is a parabolic dish? The focusing effect would definitely boost the concentration of the power (and make it fall off less quickly), and the water molecules in people's eyes don't react too well to being heated...
LOL, tested and stable?
MUMPS is much older than Apache, PHP, Python, MySQL, or Linux. In the last 30 years it has probably changed less than any of the above did in their last minor point release.
PAL sucks because in order to air content that was originally 24fps progressive, the audio has to be sped up to keep it in sync and everything is too high pitched.
Our phone system's voice mail processor used to be on an OS/2 Warp 3.x box (with the GUI disabled). Thing was stable as hell, ran for years without being touched.
When we "upgraded" the phone system, it got replaced with one that runs on NT. It came preloaded with an 'at' job to reboot it nightly...
Not that obscure. As soon as I saw the comment title I was waiting for someone to say that :)
Could they be the miners?
Sure they're like 3 years old.
Miners, not minors!
And besides, Joshua wasn't really evil, just a little confused.
Was that English or Bad English?
My laptop is a crappy old Compaq -- I definitely don't have enough RAM.
FreeBSD helps because the OS itself is fairly minimalistic and has a small memory footprint. It also doesn't run a lot of extraneous services by default (yeah, I could disable them on some other OS but it's easier to just add what you need).
The other part that's useful is that the VM is tunable so I can tweak it for what I want. For example, by default FreeBSD will preemptively swap a copy of idle pages sometimes. That makes sense for servers and desktops where the drive is constantly running. If the memory is used again that's fine, it's still there. If pressure on the VM increases and it needs more space, those pages can be quickly discarded from memory since they're already on the disk.
In my situation, however, that's not very desirable. It's only a simple sysctl to change the behavior to favor swapping as little as possible.
This is exactly what I do with my current laptop. 1GB flash card with the OS and most commonly used apps. The hard drive has swap (almost never touched since I'm running FreeBSD), apps that are not used very often, and large documents that won't fit on the flash. Volatile stuff like /var runs out of a ramdisk.
Works great; I have ataidle spin down the drive as soon as it boots up, and set the timeout low so that if it has to spin up for something it doesn't run very long. The thing is near silent and doesn't get as hot as it used to. I can check email, browse with firefox, and ssh to remote systems without the drive ever spinning up.
No, the solution is that your company should have an external authenticated mail relay that is included in the SPF record.
Authenticated is the key word here. Anybody who's roaming uses the company's relay. Hell, use it internally too and you don't have to change any settings while away. I've yet to come across a mail client that doesn't support SMTP AUTH, and many allow you to "use the same password that I do for checking mail" for convenience.
The mail relay should run on the submission port (587), or better yet over SSL (port 465). This gets around the port 25 blocks and transparent redirects that many brain-dead ISPs and hotels have.
This is nothing more than growing pains, and in the long run, capitalism SHOULD (unless someone screws it up) raise everyone's standards up. Even Ethiopia, so that everyone can have 5 weeks vacation, and 3 hour lunches.
That's a nice ideal, but the question is: Does the planet have enough resources for everyone to have that high of a standard of living? Don't forget there are billions of people living at a much lower standard than Americans or Eurpoeans both are used to.
Not saying that it's right or that I have an answer for it, but sometimes the universe doesn't allow for a perfect solution. It's only going to get worse as the population increases, too. If there was massive investment in space exploration and enough people were willing to relocate, maybe, but that's a longshot at best.
And I hate to reply to myself, but Sender-ID systems probably do have to receive the whole message, since Sender-ID checks the From: field also...
Depends how you set it up. On my Postfix server, it checks SPF before the message is received, but only rejects the message if the domain has an explicit "-all" in its SPF record. You'd be surprised how many domains are actually publishing SPF v1 (i.e. NOT Sender-ID) records.
Other methods involve using SPF as an additional spam score in systems like SpamAssassin. Those of course don't check until after the message is received, but can apply fuzzier maching than simple accept/reject choices.
You can still experiment with your own mail servers. It just means you get to experiment with your own DNS too in order to make it work.