You are forgetting that ST Voyager moved a long distance, many quadrants, in just an instant after leaving deep space nine. This is the foundation of the show. In ST, they do not understand everything yet...
You can express IPv6 adresses with quad-dotted notation if you wish, going way over the 255 limit. The truth is that it is the underlying number of bytes in an IP packet header that matters.
IPv6 addresses range is 0.0.0.0 to 4294967295.4294967295.4294967295.4294967295
From TFA: "A bank could ask customers to sign up for a program that would scan their PC for signs of infection during online sessions"
hello ? privacy issues anybody ?
So basically organizations that do business with consumers would be allowed to scan the consumer PC. Great idea...
Next step, you have to allow the government, banks, Ebay, Paypal and what not to scan your PC otherwise they will refuse to do business with you. Since they may not have a linux or other OS scanners, you would be required to use Windows of course.
Note: a 22,000 miles tower would actually be needed to achieve what I describe, 22 miles wouldn't do it.
Then again, you could always use a geostationary satellite with a large mass and attach a rope to it instead of building a tower and make other satellites to be launched climb up that rope instead.
Build up the mass of the pulling satellite slowly, in many operations ($$$). Take pieces of that rope gradually to the pulling satellite and have space welders assemble it in the way back to Earth.
The nice thing is there is no "you need to support the weight of the rope" non-sense because with centripetal forces, it all equates out and the rope doesn't have to support its own weight. Note to myself: I need to QA this with regards to what happens in the the close to Earth part of the rope, forces might not be applied in a linear fashion after all but still...
Does that remind you some movie were they had a space elevator ?;-)
Not true if you have a 36,000 km tall tower -> geostationary orbit, taller than that you actually need to slow the object down so it would stay in orbit.
2. Because it would require stateful inspection of connections to know which outbound packets are in reply to incoming connections that have been allowed, and would hence put a substantially higher load on my firewall, requiring a more expensive firewall in order to be sure it could cope with all the connections.
Not true, all reply packets related to incoming connections originate from port 80 or 443 on your server and this is how you usually block outgoing traffic from your server when you are concerned with firewall overhead; you only allow packets coming from port 80 and 443 on your server to make it to the Internet. No stateful inspection is needed. This is also how things were done with ipfwadm, before iptables and ipchains. You had no other choices with ipfwadm since it did not support any kind of state related filtering as far as I can remember.
It that's true, then that solves the question with regards to why it would cause a slowdown. I could image that parsing a whole Slashdot page with 500 comments in it to find phone numbers and inserting links would cause some kind of additional slowdown;-)
Well this sure sounds like when they need to give somebody access to *some* data, they just give her/him a username/password which then grants her/him access to the whole database.
ACLs ? group based authorization ? For example, very few people should be allowed to view credit card numbers, a representative should only be allowed to view his own customers data, etc.
Kind of like: You are the new guy who is managing our blog ? Here is the root password on all our systems, thanks to yp, they are the same on all machines. Have fun in your new job.
My problem is that I have SSH at second highest priority, first priority is VOIP packets. I use rsync over SSH so I need to use the --bwlimit option in rsync in order to not starve other services while still maintaining good interactivity when typing commands on a remote system.
To be on the safe side, I use 85% but mileage may vary...
Anyway, the poster could have solved his problem easily by using the --bwlimit option instead of interrupting the transfer to please his kids. I routinely transfer GB of data with that option without any impact on the network.
What tool has he been using to come to his conclusions ?
I found that the bottleneck always seemed to be between the DSL modem or the cable modem and the provider router by doing something as simple as pinging and watching the dropped packets and high latency results between the modem and the provider router when you try to use a bandwidth rate that is above what the link can support.
The link degrades pretty quickly in those situations. Furthermore, pinging a distant far away router 10 hops away when the congestion occurs reveals no additional degradation compared to pinging the provider router which seem to indicate that the problem is mostly between the cable modem and the provider router, in the DOCSIS link.
It is different when you are on a commercial truly bi-directional fiber link or T1/T3 but DSLs and cable modems aren't typically bi-directional, they even have different upload and download speeds.
I might be wrong, but intuitively, I think his findings need more reviewing. The guy is talking about a cable modem and Comcast. Not exactly how a commercial fiber link or T1/T3 would work.
I mean there is tons of articles on the Internet on how to do traffic shaping in order to use VOIP on consumer grade links. Most of them refer to the limitations of cable modems and DSLs. Most of the articles warn of what happens, especially when you saturate the upload bandwidth typically limited to less than 1Mb/s just like this guy was doing...
This is not new at all, I have been observing that since the late 90s when DSLs and cable modems first came out.
I have never been using Facebook although I have heard a lot about it. Obviously some people, mostly mainstream from what I can understand, seem to enjoy it a lot.
What would be the percentage of "Facebook penetration" amongst the/. users ?
> Start a test framework yourself, of bugs you cover, > and run it against the code day by day.
I second that. It is amazing sometimes how you have to take the decision and just bill it on normal bug fix time or whatever task is assigned to you. Often, managers won't care about it as long as the average time to fix a bug or the time spent on your regular tasks remain the same.
It is harder to do it this way because you then have do it gradually. We even have redesigned application cores without letting the end customer know ! If we had told him, he would have been disturbed and worrying about it. For him, all we did was change requests and bugfixes. In this case, in was more complex this way because the old design had to be compatible with the new design for quite a while.
In the end, we managed to implement changes more easily, we could afford to charge the customer less while making better margins.;-)
I would say the most important thing to save money is a good design and programming guidelines for the application.
Test suites are obviously a good idea but it may end up as a tedious task if the application as poor or no real starting design and if everybody just did things their ways.
I have seen many applications that had grown larger than expected that typically cost 10 times more than it should to maintain, bug fix and implement change requests.
It is sometimes possible to fit existing code in a new clean design and that saves a lot of money if the application is going to be used for a long time.
Let's hope your application has proper design !;-)
Slashdot Mirror
You are forgetting that ST Voyager moved a long distance, many quadrants, in just an instant after leaving deep space nine. This is the foundation of the show. In ST, they do not understand everything yet...
While at it, why not XML based IP packets with XML parsers in the routers to route every packet ?
FYI: That's exactly what "speaking in order of magnitude" means.
http://en.wikipedia.org/wiki/Order_of_magnitude
http://www2.pvc.maricopa.edu/tutor/chem/chem151/metric/magnitude.html
Well, speaking in order of magnitude, we are almost half way there with IPv6:
"Two approximate calculations give the number of atoms in the observable universe to be close to 10^80."
http://en.wikipedia.org/wiki/Observable_universe
The number of IPv6 addresses is:
2^128 or 3.4×10^38
> but the million-year ping times would be a bit annoying anyway.
Nah... subspace channels are much faster than the speed of light. I have seen it being used in many situations...
You can express IPv6 adresses with quad-dotted notation if you wish, going way over the 255 limit. The truth is that it is the underlying number of bytes in an IP packet header that matters.
IPv6 addresses range is 0.0.0.0 to 4294967295.4294967295.4294967295.4294967295
2^128 or 3.4×10^38 addresses.
From TFA:
"A bank could ask customers to sign up for a program that would scan their PC for signs of infection during online sessions"
hello ? privacy issues anybody ?
So basically organizations that do business with consumers would be allowed to scan the consumer PC. Great idea...
Next step, you have to allow the government, banks, Ebay, Paypal and what not to scan your PC otherwise they will refuse to do business with you. Since they may not have a linux or other OS scanners, you would be required to use Windows of course.
This guys is a genuis !
By the guy, I meant the blog owner. I thought it was obvious... ;-)
Anybody with some type of aviation or military experience would have although ;-)
The guy has been lucky so far, wait until the mob gets at him ;-)
Note: a 22,000 miles tower would actually be needed to achieve what I describe, 22 miles wouldn't do it.
Then again, you could always use a geostationary satellite with a large mass and attach a rope to it instead of building a tower and make other satellites to be launched climb up that rope instead.
Build up the mass of the pulling satellite slowly, in many operations ($$$). Take pieces of that rope gradually to the pulling satellite and have space welders assemble it in the way back to Earth.
The nice thing is there is no "you need to support the weight of the rope" non-sense because with centripetal forces, it all equates out and the rope doesn't have to support its own weight. Note to myself: I need to QA this with regards to what happens in the the close to Earth part of the rope, forces might not be applied in a linear fashion after all but still...
Does that remind you some movie were they had a space elevator ? ;-)
Not true if you have a 36,000 km tall tower -> geostationary orbit, taller than that you actually need to slow the object down so it would stay in orbit.
2. Because it would require stateful inspection of connections to know which outbound packets are in reply to incoming connections that have been allowed, and would hence put a substantially higher load on my firewall, requiring a more expensive firewall in order to be sure it could cope with all the connections.
Not true, all reply packets related to incoming connections originate from port 80 or 443 on your server and this is how you usually block outgoing traffic from your server when you are concerned with firewall overhead; you only allow packets coming from port 80 and 443 on your server to make it to the Internet. No stateful inspection is needed. This is also how things were done with ipfwadm, before iptables and ipchains. You had no other choices with ipfwadm since it did not support any kind of state related filtering as far as I can remember.
It that's true, then that solves the question with regards to why it would cause a slowdown. I could image that parsing a whole Slashdot page with 500 comments in it to find phone numbers and inserting links would cause some kind of additional slowdown ;-)
Well this sure sounds like when they need to give somebody access to *some* data, they just give her/him a username/password which then grants her/him access to the whole database.
ACLs ? group based authorization ? For example, very few people should be allowed to view credit card numbers, a representative should only be allowed to view his own customers data, etc.
Kind of like: You are the new guy who is managing our blog ? Here is the root password on all our systems, thanks to yp, they are the same on all machines. Have fun in your new job.
Congratulations, you have understood all that needed to be understood ;-)
Is that why my wife is always running around ?
Exactly what I was trying to say ;-)
My problem is that I have SSH at second highest priority, first priority is VOIP packets. I use rsync over SSH so I need to use the --bwlimit option in rsync in order to not starve other services while still maintaining good interactivity when typing commands on a remote system.
To be on the safe side, I use 85% but mileage may vary...
Anyway, the poster could have solved his problem easily by using the --bwlimit option instead of interrupting the transfer to please his kids. I routinely transfer GB of data with that option without any impact on the network.
What tool has he been using to come to his conclusions ?
I found that the bottleneck always seemed to be between the DSL modem or the cable modem and the provider router by doing something as simple as pinging and watching the dropped packets and high latency results between the modem and the provider router when you try to use a bandwidth rate that is above what the link can support.
The link degrades pretty quickly in those situations. Furthermore, pinging a distant far away router 10 hops away when the congestion occurs reveals no additional degradation compared to pinging the provider router which seem to indicate that the problem is mostly between the cable modem and the provider router, in the DOCSIS link.
It is different when you are on a commercial truly bi-directional fiber link or T1/T3 but DSLs and cable modems aren't typically bi-directional, they even have different upload and download speeds.
I might be wrong, but intuitively, I think his findings need more reviewing. The guy is talking about a cable modem and Comcast. Not exactly how a commercial fiber link or T1/T3 would work.
I mean there is tons of articles on the Internet on how to do traffic shaping in order to use VOIP on consumer grade links. Most of them refer to the limitations of cable modems and DSLs. Most of the articles warn of what happens, especially when you saturate the upload bandwidth typically limited to less than 1Mb/s just like this guy was doing...
This is not new at all, I have been observing that since the late 90s when DSLs and cable modems first came out.
Is Facebook a viable long term business model ?
I have never been using Facebook although I have heard a lot about it. Obviously some people, mostly mainstream from what I can understand, seem to enjoy it a lot.
What would be the percentage of "Facebook penetration" amongst the /. users ?
A better definition could be:
"A user saturating its broadband connection by transferring 20GB files and not taking care of using the --bwlimit (limit bandwidth) option with rsync"
I have been using it for ages to prevent this very problem.
Other types of traffic shaping can be done, with Linux tc as an example, but it is always best to do it at the application level when possible.
Hey, don't forget Jughead and Veronica !
Poor Betty, I am not aware of her getting anything...
> My opinion is to break it down into small incremental changes....
I agree, I just posted a reply that is along the same line:
http://slashdot.org/comments.pl?sid=1932550&cid=34742086
> Start a test framework yourself, of bugs you cover,
> and run it against the code day by day.
I second that. It is amazing sometimes how you have to take the decision and just bill it on normal bug fix time or whatever task is assigned to you. Often, managers won't care about it as long as the average time to fix a bug or the time spent on your regular tasks remain the same.
It is harder to do it this way because you then have do it gradually. We even have redesigned application cores without letting the end customer know ! If we had told him, he would have been disturbed and worrying about it. For him, all we did was change requests and bugfixes. In this case, in was more complex this way because the old design had to be compatible with the new design for quite a while.
In the end, we managed to implement changes more easily, we could afford to charge the customer less while making better margins. ;-)
I would say the most important thing to save money is a good design and programming guidelines for the application.
Test suites are obviously a good idea but it may end up as a tedious task if the application as poor or no real starting design and if everybody just did things their ways.
I have seen many applications that had grown larger than expected that typically cost 10 times more than it should to maintain, bug fix and implement change requests.
It is sometimes possible to fit existing code in a new clean design and that saves a lot of money if the application is going to be used for a long time.
Let's hope your application has proper design ! ;-)