So SCO employees are responsible for upper management decisions now? Much the same as citizens are responsible for government decisions, yeah?
So I guess if SCO employees should be ashamed of working for SCO due to actions completely out of their control, citizens of the USA should feel equally responsible and ashamed for actions such as...well where do we start? Watergate? Iran-Contra? (....Lewinski?):)
I feel that what you've said of SCO employees is absolutely and completely unreasonable.
No arguments from me in that regard - 64 bits of addressing if you want autoconf seems quite ridiculous.
I just quietly accept things like that safe in the knowledge that the people that write these standards such as IPV6 know what they're talking about. Just like the 802.11 guys knew what they were talking about when they developed WEP...
Hang on - I know exactly why they made it 64 bits...
You're actually pretty close to the mark saying that in one 64bit subnet you could virtually fit every single device know to man. That's the POINT. Autoconf doesn't just throw you a random IP and hope for the best - it assigns you an IP which is a function of the subnet and your EUI-64 address. (EUI 64 address is a function of your MAC address).
Firstly - it's quite possible to have the 128 bits of addressing become exhausted. When you consider that every organization will be allocated 64 bits of address space MINIMUM and that many may get 80 bits of addressing then you will hopefully realise that although IPV6 has a massive amount of room to move it's not the bottomless pit that people seem to think it is.
Secondly - IPV6 is a lot more than just IPV4 with 128 bits of addressing. IPV6 has support for autoconfiguration just for starters. This means that people will very rarely (in some cases, never) need to deal with an IPV6 address.
Wow! If your manager is smart now while he's in his 40's can you imagine how impressive he must have been back in his 20's!
Seriously though - I completely agree with you - my 'mentor' at work is a bloody bright guy and he'd be in his 40's too. Maybe the stereotype of old=dumb is created because more young people out of college/university have a more learning oriented outlook on life while more experienced guys might have a more 'get the job done' attitude??
1. 2900's are not available...2950's are available so the 3500 is not the logical replacement of the 2900. Regardless of this, the 3500 is also discontinued (see 3550). Then there is the 3750. Simple:)
2. Putting a whole bunch of untrusted groups on the same VLAN thinking that it's safe because it's a switched network is a typical amateur mistake to make. I'd have expected that a senior network person for a large auto group would be on top of these problems but hey, you're probably just a little kid saying you're a big-shot to add emphasis to his/her argument. Why is it a big mistake? well...it's all about ARP cache poisoning and the ability for any machine on the vlan to mascarade as another person's gateway.
Considering that large network providers have a handful (if not 1) ASN then I'd say that EBGP isn't the only routing protocol on backbone links.
EBGP gets you from one AS to another...something else needs to get you through a providers network, this is done with an interior routing protocol, IBGP, OSPF, etc.
Your point seems fair though - I'm also under the understanding that layer 2 and 3 multipath load balancing is fairly widely used.
Routing protocols aren't overly complicated due to the requirement for robustness as well as making a route selection as fast as possible so I suppose that mistakes will happen from time to time. Having said that, it's cool to think that research such as this could lead to a new generation of routing protocol.
I don't think the author of the article was right to make the alleged issues with routing algorithms look like "selfishness" amongst (presumably) network administrators.
The point that the author briefly touched on regarding a routing algorithm to take into regard the effect of adding extra load onto a particular network path was interesting but I don't really see the point. For example, OSPF (Dijkstra algorithm) will know that a 10Mb link at 50% capacity is a more attractive option compared to a 2Mb link at 0% capacity (or a 100Mb link at 98% capacity). Calculating what would happen to the network after your, say, 1Mb of capacity requirements are loaded onto a paticular path seems largely useless as the most attractive route is already going to be selected... Feel free to enlighten me on this.
Are you quite sure that issues such as a whole network of routers constantly jumping from one route to another (and back again) is something that you'll often find in the real world these days?
Further, while you'd like to see more randomness in routing, I'd like to see more work put into correct path selection.
In the days of EFTPOS and credit cards, is another form of cash-alternative that desirable?
The way I see it, the only difference between EFTPOS and a cash-card is that with a cash-card you don't have to have it linked to an actual bank account.
I'm sure a whole lot of people lead a fairly paperless life these days thanks to current plastic alternatives.
Further, I'd imagine that a POS that is not currently EFTPOS enabled is unlikely to adapt this new technology so I don't see any advantage.
I disagree..com is not an international TLD, it's a TLD for the USA. What's the native language of the USA? Well some messed up hybrid of English.
So I don't think that a.com or.net root should need to work with anything other than it's current character set, maybe this could be addressed further down the line if something was to change.
Having said that. Regional roots should be able to use a local character set if they so wish...maybe this is done already (??).
If your monsterous city registers under the regional NIC that it was supposed to and you still have trouble then I'd say it's a bit rude.
5 minutes of research would sort out the fundamental issue of whether regional DNS zones are implimenting local character sets... I'd also imagine that the DNS RFC would need to have taken such requirements into consideration...which I don't think is currently done (??).
So yeah, DNS isn't my strong point but I still think I should be able to disagree with you over expecting TLD's of the USA to support every other countries character set:)
The article outlines that the current revision of the standard is based on 10/100 Mbps Ethernet (IEEE 802.3 and 802.3u). If a 'Magic' Network can operate at 10Mbps then I'd suggest that the bandwidth of an IEEE 802.11b network would be acceptable...just.
Of course bandwidth isn't the only factor to consider.
I would have assumed that if MediaForce really wanted results they'll need to do more than make the copyrighty infringer move to a new ISP.
What have the achieved by making the ISP close the users account? Inconvenience? Not a whole lot more.
Now, if MediaForce forgot about bothering ISP's and went after the actual people who infringed on the copyright then they might be actually doing something which will bring them closer to their Ultimate Goal.
Since yesterday I'd forgotten all about how disgusting I am and now you have to just drag up the past!
I think I'm going to turn off email alerts for when people reply to my posts...
I hope you have it turned on though - otherwise you'll never read this...because what sort of lamer replies to a 3 day old post? If you don't turn have it on then this post was a waste of time - in which case......I think I'm going to have a lay down.
Yeah - I was really really impressed by it. It was quite natural to use.
BUT
I found that even as I was scrolling through at close to maximum speed it still took longer to enter text than using grafitti or maybe even a virtual keyboard.
By far the best portable data entry system I've ever seen was an optical keyboard that projected onto a desk or whatever...that was cool. Obviously if you're on the bus you'd have to revert back to grafitti though.
Slashdot Mirror
MS are taking legal action against spammers, that didn't go down to badly with slashdotters.
It'll be interesting to see ifslashdotters feel the same about this particular case.
I disagree, I think having more Linux exposure will make the decision to use it later in life a lot easier to make.
You don't need a club for that, you just need to prove that it works.
A list of email addresses to make available to spammers so that they can do the right thing.
Sounds like a great idea...but something makes me feel there could be a fundamental problem with giving a list like that to spammers...hmmm....
So SCO employees are responsible for upper management decisions now? Much the same as citizens are responsible for government decisions, yeah?
:)
So I guess if SCO employees should be ashamed of working for SCO due to actions completely out of their control, citizens of the USA should feel equally responsible and ashamed for actions such as...well where do we start? Watergate? Iran-Contra? (....Lewinski?)
I feel that what you've said of SCO employees is absolutely and completely unreasonable.
He's kinda temporarily profiting anyway.
:)
As part of the dismissal he had to pay $12K but he only pays $5K now and $2.3K next year and $2.3K the year after etc. etc.
So he's ahead for the next couple of years at least
I'd like to avoid people who just don't smell the coffee :)
By far one of the more entertaining Trolls I've read.
oh my god I'm one of "them".
Didn't RTFA, instead I just got the gist by reading the comments of people that had not RTFA.
Change the subject to "Progress" and it's all better.
This coming from the same government that just put forward several million dollars towards a Linux supercomputer.
Due to service gaps isolating cities from the uber-network this will just not work.
It could be a good idea for 'last mile' though. Telcos and ISPs could charge less because they don't need to pay for the last mile infrastructure.
No arguments from me in that regard - 64 bits of addressing if you want autoconf seems quite ridiculous.
I just quietly accept things like that safe in the knowledge that the people that write these standards such as IPV6 know what they're talking about. Just like the 802.11 guys knew what they were talking about when they developed WEP...
Hang on - I know exactly why they made it 64 bits...
You're actually pretty close to the mark saying that in one 64bit subnet you could virtually fit every single device know to man. That's the POINT. Autoconf doesn't just throw you a random IP and hope for the best - it assigns you an IP which is a function of the subnet and your EUI-64 address. (EUI 64 address is a function of your MAC address).
See...it all makes sense.
I disagree entirely with most of these posts.
Firstly - it's quite possible to have the 128 bits of addressing become exhausted. When you consider that every organization will be allocated 64 bits of address space MINIMUM and that many may get 80 bits of addressing then you will hopefully realise that although IPV6 has a massive amount of room to move it's not the bottomless pit that people seem to think it is.
Secondly - IPV6 is a lot more than just IPV4 with 128 bits of addressing. IPV6 has support for autoconfiguration just for starters. This means that people will very rarely (in some cases, never) need to deal with an IPV6 address.
Your thoughts?
a is 5GHz, which is not available to be used in Australia due to it being used for Aero navigation.
b and g are 2.4 GHz whish is fine to use in almost all countries.
It's the transmitting frequency that is key in making b and g work with the same gear.
intentional blink182 reference?
Wow! If your manager is smart now while he's in his 40's can you imagine how impressive he must have been back in his 20's!
Seriously though - I completely agree with you - my 'mentor' at work is a bloody bright guy and he'd be in his 40's too. Maybe the stereotype of old=dumb is created because more young people out of college/university have a more learning oriented outlook on life while more experienced guys might have a more 'get the job done' attitude??
A couple of things:
:)
1. 2900's are not available...2950's are available so the 3500 is not the logical replacement of the 2900. Regardless of this, the 3500 is also discontinued (see 3550). Then there is the 3750. Simple
2. Putting a whole bunch of untrusted groups on the same VLAN thinking that it's safe because it's a switched network is a typical amateur mistake to make. I'd have expected that a senior network person for a large auto group would be on top of these problems but hey, you're probably just a little kid saying you're a big-shot to add emphasis to his/her argument. Why is it a big mistake? well...it's all about ARP cache poisoning and the ability for any machine on the vlan to mascarade as another person's gateway.
Considering that large network providers have a handful (if not 1) ASN then I'd say that EBGP isn't the only routing protocol on backbone links.
EBGP gets you from one AS to another...something else needs to get you through a providers network, this is done with an interior routing protocol, IBGP, OSPF, etc.
Your point seems fair though - I'm also under the understanding that layer 2 and 3 multipath load balancing is fairly widely used.
Routing protocols aren't overly complicated due to the requirement for robustness as well as making a route selection as fast as possible so I suppose that mistakes will happen from time to time. Having said that, it's cool to think that research such as this could lead to a new generation of routing protocol.
I don't think the author of the article was right to make the alleged issues with routing algorithms look like "selfishness" amongst (presumably) network administrators.
The point that the author briefly touched on regarding a routing algorithm to take into regard the effect of adding extra load onto a particular network path was interesting but I don't really see the point. For example, OSPF (Dijkstra algorithm) will know that a 10Mb link at 50% capacity is a more attractive option compared to a 2Mb link at 0% capacity (or a 100Mb link at 98% capacity). Calculating what would happen to the network after your, say, 1Mb of capacity requirements are loaded onto a paticular path seems largely useless as the most attractive route is already going to be selected... Feel free to enlighten me on this.
Are you quite sure that issues such as a whole network of routers constantly jumping from one route to another (and back again) is something that you'll often find in the real world these days?
Further, while you'd like to see more randomness in routing, I'd like to see more work put into correct path selection.
In the days of EFTPOS and credit cards, is another form of cash-alternative that desirable?
The way I see it, the only difference between EFTPOS and a cash-card is that with a cash-card you don't have to have it linked to an actual bank account.
I'm sure a whole lot of people lead a fairly paperless life these days thanks to current plastic alternatives.
Further, I'd imagine that a POS that is not currently EFTPOS enabled is unlikely to adapt this new technology so I don't see any advantage.
I disagree. .com is not an international TLD, it's a TLD for the USA. What's the native language of the USA? Well some messed up hybrid of English.
.com or .net root should need to work with anything other than it's current character set, maybe this could be addressed further down the line if something was to change.
:)
So I don't think that a
Having said that. Regional roots should be able to use a local character set if they so wish...maybe this is done already (??).
If your monsterous city registers under the regional NIC that it was supposed to and you still have trouble then I'd say it's a bit rude.
5 minutes of research would sort out the fundamental issue of whether regional DNS zones are implimenting local character sets... I'd also imagine that the DNS RFC would need to have taken such requirements into consideration...which I don't think is currently done (??).
So yeah, DNS isn't my strong point but I still think I should be able to disagree with you over expecting TLD's of the USA to support every other countries character set
The article outlines that the current revision of the standard is based on 10/100 Mbps Ethernet (IEEE 802.3 and 802.3u). If a 'Magic' Network can operate at 10Mbps then I'd suggest that the bandwidth of an IEEE 802.11b network would be acceptable...just.
Of course bandwidth isn't the only factor to consider.
I would have assumed that if MediaForce really wanted results they'll need to do more than make the copyrighty infringer move to a new ISP.
What have the achieved by making the ISP close the users account? Inconvenience? Not a whole lot more.
Now, if MediaForce forgot about bothering ISP's and went after the actual people who infringed on the copyright then they might be actually doing something which will bring them closer to their Ultimate Goal.
Why did you have to reply!
...I think I'm going to have a lay down.
Since yesterday I'd forgotten all about how disgusting I am and now you have to just drag up the past!
I think I'm going to turn off email alerts for when people reply to my posts...
I hope you have it turned on though - otherwise you'll never read this...because what sort of lamer replies to a 3 day old post? If you don't turn have it on then this post was a waste of time - in which case...
.005256 minutes down/year ~= (HALF A MILLISECOND/YEAR!!!!)
...I'm so disgusted in myself, I hate pedantic pricks that correct people on simple and obvious mistakes.
Half a milliminute
Yeah - I was really really impressed by it. It was quite natural to use.
BUT
I found that even as I was scrolling through at close to maximum speed it still took longer to enter text than using grafitti or maybe even a virtual keyboard.
By far the best portable data entry system I've ever seen was an optical keyboard that projected onto a desk or whatever...that was cool. Obviously if you're on the bus you'd have to revert back to grafitti though.