After playing with the Winamp plugin, winamps memory usage ballooned to 60megs and growing. Not sure if thats the fault of Quicktime API or the plugin, but its not ideal.
I think apple just wants people to keep using its iTunes software. I personally like Winamp a heck of a lot more, but I put up with iTunes because i like the music store. (Only iTunes is capable of playing protected AAC files)
This ofcourse makes people much more willing to go buy iPod's which is apple's real revenue stream.
If people can use playfair to convert to non-protected AAC which can play in a number of players, they lose their iPod lock which is their main revenue stream.
If you are using MD5 encryption that is built into BGP4 that won't help you. That merely will stop route poisoning and such. As long as your underlying TCP/IP implementation isn't authenticating source/destination you can still get hosed.
Wishful thinking. Mexico isn't graduating millions of highly skilled university graduates every year.
And actually, china is starting to design their own chips, trying to focus most of their O/S engineers on a version of FOSS etc to really become independant of US companies.
Please. I worked tech support for 2.5 years at my university.
I'll spend 5-10 minutes trying to help someone who just randomly comes up and says 'Hey, I remember you from the help desk. I have this....' Or some friend of a friend. 'Hey, this is my buddy, his computer is...' But thats it. I hardly know the person, and I don't have time. Between my own computer issues and those I was dealing with at work, I want some time not devoted to dealing with how buggy people can make their systems.
If its a close friend, of course its not a problem. But apparently just because you don't get asked frequently, doesn't mean others don't. Don't let that stop you from making sweeping generalizations though.
The above poster is exactly right. The privacy issue is that they have records of what you bought, not that they sent you a notice. However, since they obvious already have this data, which you probably agrred to let them track anyway, I think its a nice gesture to go ahead and notify people of their elevated risk.
I am well aware of cache and branch mispredict penalties.
However, look at some data for non-SSE apps that have been patched to support SSE. The difference is incredible. We're talking >30% for heavily optimized applications.
Samsung has this GPS feature and it is set to turn on only when calling 911 by default. It can be set to always on however.
As long as that is the default setting which I was happy to see that it was, I see nothing to worry about. (unless ofcourse it doesnt work as advertised).
"You should also use a receive only sniffer cable on this box to protect integrity... a receive only box has a zero chance of infection as it's physically impossible."
Am I the only one that balks at this statement? Maybe I am missing something but it does seem that even with rx-only you could be infected, just not by any connection oriented protocols? (Or maybe even still if some really strange bug crops up).
I don't know for sure, but I'd imagine that there would be some slight overhead involved, and from the viewpoint of an attacker you might be able to just look at the global BGP tables and make a judgement call about the size of the subnet and just randomize the targets.
I believe its actually a generally syn/sec threshold. Atleast with PIX's (what i am familiar with).
When you go above it, it initially ages out unanswered syn's, then starts setting more aggressive timers for aging out the syn setup. What happens is that eventually it will stop accepting new syn's until the old ones age out. How large that table is depends on the PIX. This is all on a global level.
I was trying to dig up the info on cisco's site to be 100% sure but at a cursory glance but wasn't successful.
Slashdot Mirror
Not all.
I run Gentoo, upgraded to 2.6 and it booted just fine.
What do you know? A new update is already available.
.9.0, .9.1 and .9.2!
Can't they *fix that* already? It's been in
SafeSex!
Yay Nullsoft.
http://www.nullsoft.com/free/safesex/
Unfortunately after a few hours of usage winamp has now grown to over 60megs of memory =x
After playing with the Winamp plugin, winamps memory usage ballooned to 60megs and growing. Not sure if thats the fault of Quicktime API or the plugin, but its not ideal.
Interesting, thanks. I'll have to play with this.
I think apple just wants people to keep using its iTunes software. I personally like Winamp a heck of a lot more, but I put up with iTunes because i like the music store. (Only iTunes is capable of playing protected AAC files)
This ofcourse makes people much more willing to go buy iPod's which is apple's real revenue stream.
If people can use playfair to convert to non-protected AAC which can play in a number of players, they lose their iPod lock which is their main revenue stream.
I see the reasoning for this, but is it practical?
What happens if the card has an embedded controller that uses an architecture that isn't easily compiled by gcc or something?
What do you do then?
Obviously is keep it precompiled.
This, while good intentioned, I can see breaking a lot of things which aren't really purposefully breaking a lot of things without easy solutions.
Thanks, learn something new everyday =)
Anyway, happen to know if most tcp/ip implementations support this nowadays?
If you are using MD5 encryption that is built into BGP4 that won't help you. That merely will stop route poisoning and such. As long as your underlying TCP/IP implementation isn't authenticating source/destination you can still get hosed.
Duck and cover.
Wishful thinking. Mexico isn't graduating millions of highly skilled university graduates every year.
And actually, china is starting to design their own chips, trying to focus most of their O/S engineers on a version of FOSS etc to really become independant of US companies.
Just give it time until they catch up.
The point of the article is not that the center isn't needed. It's that something so horribly stupid can occur there in a lvl 4 facility.
Simply saying "Well we had back up generators, but they didn't work. Sorry." Does not cut it.
Please. I worked tech support for 2.5 years at my university.
I'll spend 5-10 minutes trying to help someone who just randomly comes up and says 'Hey, I remember you from the help desk. I have this....' Or some friend of a friend. 'Hey, this is my buddy, his computer is...' But thats it. I hardly know the person, and I don't have time. Between my own computer issues and those I was dealing with at work, I want some time not devoted to dealing with how buggy people can make their systems.
If its a close friend, of course its not a problem. But apparently just because you don't get asked frequently, doesn't mean others don't. Don't let that stop you from making sweeping generalizations though.
The above poster is exactly right. The privacy issue is that they have records of what you bought, not that they sent you a notice. However, since they obvious already have this data, which you probably agrred to let them track anyway, I think its a nice gesture to go ahead and notify people of their elevated risk.
I am well aware of cache and branch mispredict penalties.
However, look at some data for non-SSE apps that have been patched to support SSE. The difference is incredible. We're talking >30% for heavily optimized applications.
If you were to use SSE2 you would see an incredible performance boost.
It's the weakening batteries and changing position of the sun that makes it fail.
"After all this time, the Progressive article is now online"
Not for long.
Not triangulation. GPS "location based services".
It transmits your GPS coordinates to 911.
Samsung has this GPS feature and it is set to turn on only when calling 911 by default. It can be set to always on however.
As long as that is the default setting which I was happy to see that it was, I see nothing to worry about. (unless ofcourse it doesnt work as advertised).
If you saw Shanghai these days you might be suprised.
"You should also use a receive only sniffer cable on this box to protect integrity... a receive only box has a zero chance of infection as it's physically impossible."
Am I the only one that balks at this statement? Maybe I am missing something but it does seem that even with rx-only you could be infected, just not by any connection oriented protocols? (Or maybe even still if some really strange bug crops up).
Or am I just missing something...
I don't know for sure, but I'd imagine that there would be some slight overhead involved, and from the viewpoint of an attacker you might be able to just look at the global BGP tables and make a judgement call about the size of the subnet and just randomize the targets.
I believe its actually a generally syn/sec threshold. Atleast with PIX's (what i am familiar with).
When you go above it, it initially ages out unanswered syn's, then starts setting more aggressive timers for aging out the syn setup. What happens is that eventually it will stop accepting new syn's until the old ones age out. How large that table is depends on the PIX. This is all on a global level.
I was trying to dig up the info on cisco's site to be 100% sure but at a cursory glance but wasn't successful.