I can't remember but if this was ever true, but it isn't now.
Extreme Networks and Cisco both have long haul GBICs available, and have done for a little while.
I don't have info on the Cisco models available right now (Ciscos website is BIG!), but Extreme have a 1000BaseLX-70, that will do 70 km on singlemode dark fibre. On top if this they do what I could best call a gigabit fibre line driver in the form of a "SummitGbX"[tm]. They claim it will definately do 80 km, and possibly up to 100 km, I have heard one claim that these units managed to reach 120 km. Basically, you hook these up to your 1000BaseSX interfaces at each end, and it does some wibbly-bits to bridge the 80 km or whatever length of fibre:)
I would love to verify all this independently:) but I never even dared ask the prices:/
Of course, all this doesn't really help your cause much... you might be able to find a friendly telco that will blow fibre in for you. Apart from that, your realistic options might not include wireless. At that range, you will have to go for some sort of microwave transmission, and even then you will probably have problems due to the Fresnel effect, which bends and scatters photons at the earths surface. The maths escapes me at this moment, but to reach 45 miles in one hop you would probably need to have each end nearly 200 foot up in the air in order to clear inconvenient obstacles in between, like buildings and trees (how careless to put them there!). To do it in several smaller hops might be easier, but then you have to rent or buy locations to put your repeater stations on. Another possible thing to do would be to link all the schools locally to a central point using some easily available method. Microwave links might be suitable here, as well as optical wireless links, T1 leased lines, or whatever you chose. The most usefull central point would be a telco CO, which means that you will only have to rent backhaul bandwidth on the telco's network without having to pay for an expensive tail to anywhere else. Almost all CO's will be served by fibre now, which makes renting a fast connection very much cheaper and easier to provision. As you well know, it is that last mile that makes it expensive.
I wish you good luck:) maybe you'll let us know how it turns out?
btw, I don't have any connection with Cisco or Extreme, I just use their kit in my job
Irix 6.5 comes with the kernel headers that you need for free.
maybe you could contact your local SGI office to see if that machine is entitled to run Irix 6.5, and if it is, where to get a media set from.
Once you have that, then you can use GCC to compile whatever you need.Alternatively, there is a good selection of freeware for SGIs at http://freeware.sgi.com
That includes a quite large amount of GPL software in binary form for you to download and install
> On of the biggest problems with EFNet, has been that there is no 'ownership' of channels - or the
> fact that no ircops participate in such matters.
This has always been the case, it always worked before, but now I suppose there is a better class of asshole wondering the internet.
and what do you think IRCops should do?
should they do what they are meant to, and look after servers?
or should they pander to thousands of whining users that accidently lost ops in a channel, and to hell with looking after a high load server
channel ownership, in my experience, only works in a small network where there is some feasibility of controlling the channels... you can't do that on efnet...
so instead of having users fighting amongst themselves, the users would fight amongst themselves AND the IRCops, wasting their time and taking it away from their real purpose of looking after the servers... if you don't look after the servers then you will never have an IRC network for the channels to exist in..
I agree with this, I went to university here in england, to start with I was going for the bit of paper...
I realised soon that going to university for a bit of paper was completely the wrong thing, on my course myself and about a tenth of the rest of us (there were about 120 in my year doing CS) were quite obviously more knowledgable than most of the people who were supposed to be teaching us things.
I remember getting the most deathly stare for correcting one of the lecturer's a little more loudly than I intended... it was meant to be "sotto vocce" but it came out just a little bit louder:)
so why go to university/college?
all the stuff I learnt there had nothing to do with a degree or a bit of paper, all the usefull stuff was about how to look after myself and how to live.
all the cool people i met and the cool things I did...
that is why you should go to university...
but having said that i left university, simply because i didn't think that the degree i was going to end up with would be worthwhile, the course was abysmal, the resources nearly non-existant. I had a good time though:) even if i didn't get that piece of paper.
mind you, you don't get bits of paper to say I can do what I can do... so it all seems a bit pointless really
water has an enormous specific heat capacity, I don't know what it is, but I am sure I could find out if i were enthused enough.
a specific heat capacity of something is the number of joules of energy that it takes to raise a kilogram of that by one degree Celcius.
so, let's say a cup of coffee has 300 grams of water in it... 300 grams of water that have a LOT more energy in them than 300 grams of air, which would occupy a vastely greater volume.. and so have even less energy that can be transferred to something else
> Actually reworking SMT is not all that hard. I used to assemble PCMCIA video cards
> (204 pin TQFP, some SOJ, a handful of 1206 resistors and capacitors, etc.) with
> nothing more than a pair of tweezers and a fine pitch soldering iron.
Well, I have two points to make..
1) you were presumably trained or taught yourself, you must have made a mess of things a lot of times before you learnt...
2) you were presumably paid to have a skill to do this
you said yourself you made mistakes, how many people who read these are going to be able to achieve this, the only reason it isn't hard to you is because you have practise. The real big danger here is someone is going to see this, and try it... and then turn their Palm into a paperweight
you make it sound simple, and I admit it is. but whilst what to do is simple enough, the actual doing of it isn't - it needs practise, skill and a steady hand. (or i suppose a powdered solder/flux mix you can paint on, and a soldering iron with a wide tip to heat a side at once:)
Back to the main reason for this post though..
you say it is easy, when for 99% of people it isn't, and I had to point out the real world here:)
oh, and clipping all the pins IS a cheating method, but hey, it is how i would have done it, then your can remove the old solder and remains of legs at you leisure.
And yes, whilst soldering a new chip in is technically quite simple (hint: stick the chip down to hold it in place, but don't use massively strong glue in case you want to reposition it:). that is a fantastically huge number of pins and I would never be able to keep my hand steady enough to do it without making a mess of the whole thing:(
all through this, you have to make sure you only apply as much heat as you need, because too much heat will make all the tracks on the board curl and peel off. I watched this happen once, it was a horrible sinking feeling in the pit of my stomach.... AND the damn solder still didn't melt
ok, time to stop waffling and go back to working:)
> I was just suddenly filled with excitement that I had an excuse to see the inside
> of my new toy. It was at this point that I realized Wow, I bet most people
> wouldn't be doing this.
hang on, you need an excuse?:)
I never needed an excuse...
though I did occasionally need a quick hand to put all the screws back in before my parents got back, but that was when i was much younger.
This looked fascinating to me until I realised that you need to rework surface mounted components!
I don't know how many people who read/. have tried this, but I have a few years ago. It is possibly one of the most incredibly difficult things you can attempt! I could move that little resistor they have marked, but to succesfully remove an entire DRAM chip and resolder a new one would require some extraordinary skill!
Even when I was at my most practised and my hands seemed rock steady to me I would NEVER have attempted this!
even now i could remove the chip, but to put in place a new one would be unthinkable
Please no-one try this unless you are completely sure you can pull it off, I have visions of hundreds of wrecked palms from people try to do this.
If you DO want to try it still.. good luck:)
but don't come crying if your Palm gets amnesia;)
Any machine is only as secure as the administrator makes it.
imho there are a lot of things in the article that aren't the result of Debian being "bad" they are the result of incompetence or of overlooking obvious details.
ok, so maybe we should do what this poster suggests, but if you can't handle doing basic security-wise configuration, you shouldn't be setting up a machine that needs it. Sometimes you just have to take responsibility for things, you can't just say "suchandsuch should have done thingummybob".
Demon still use SMTP mail delivery as the main method...
and don't forget that Demon Internet (now part of Thus) also has a static IP for each dialup customer that has a homepage.. ie, not only is there an IP for luser.demon.co.uk but if they upload web pages, an IP is assigned by demon to www.luser.demon.co.uk !
I wonder if demon will change all their dialup users homepages to name based virtual hosting and give back however many thousands of IPs this will free up
I don't know about anywhere else, but at the ISP I work at, we only give a user a static IP if they can justify it, then we can justify it to RIPE
fair's fair, right?
It doesn't really matter to us, or anyone else.. it would work the same and still give an encrypted connection..
but, people would notice, customers _always_ somehow manage to notice the least little thing, all they see is the box that comes up, they wouldn't look at the little padlock that says the connection is encrypted.. and of course as someone else posted, without the certificate checking you can mount a "man-in-the-middle" attack.
also, the thought occurs that when you generate the CSR (certificate signing request) it uses the server key, the server key has the servers name in it doesn't it? so the servers _name_ will always be in the certificate, and fail to match with the address that someone typed in to access one of these virtually hosted SSL sites.
surely a browser would alert someone to this discrepency?
This will probably immediately send J. Random Bloggs into a fit of panic that it isn't a secure connection, regardless of whether it really is or not.
This would make you and your company look _bad_
I can't actually test whether this happens or not, not tending to have a spare SSL certificate lying around to play with.. can anyone confirm/disprove this?
> By the way is this really a problem? Wouldn't
> you want a dedicated server anyway to host your
> site if it has to hanlde confidencial data? I
> don't like the idea of a Web-shop running on a
> 100+ site web-hotel, storing my credit card
> number - SSL or no SSL
well, you can either secure the system hosting this "web-hotel" completely and partition all the users, or you can arrange to store the confidential details on a specially designed and secured repository outside of that server.. both of these and just theories of course..
or, you can pay us a lot of money and we (and most of the other ISPs in the world) will build, install and host a dedicated server for you
This is a good point, but you have to choose to do something that is feasible, sensible and cost effective. You have to make a decision about which trade-off's to accept. (just like anything else really)
they may be issued on a name basis, but the problem here is that SSL is a transport, you have to negotiate the SSL link complete with the certificate before you get to talk to the actual web server... at this point the server doesn't know which web site you are looking for, and therefor has no way to know which certificate it should send.
Where I work right now, SSL is one of the biggest problems, we have 5 servers here running host-header based virtual hosting, but we have had to set aside relatively large chunks of our IP space to cater for the customers who want SSL.
To top this off, the SSL-hosting IPs can only do one thing each, and cannot be accelerated by our caching system... a single SSL site on one server generates 3 times as much traffic as the whole of the other sites on that server, because the normal sites can be accelerated, SSL can't.
So... how do we fix the SSL https issue?
I would love to do name based SSL hosting.. but I can't see how
> Perhaps all telcos should provide open source, impartial network monitoring software in order
> to self-regulate the quality of service. In Britain we have a telco watchdog that would
> stomp all over this thing.
ok, so why hasn't it stomped over BT and their DREADFULL handling of DSL, imho OFTEL are a lame duck and have precious little bite against any telco
Re:was I the only one to have a problem with Yast2
on
SuSE 7.0
·
· Score: 1
or if you had read the documentation, you would know to type "manual" at the LILO: prompt to get the choice of YaST 1 or 2:)
or alternatively, as another poster suggests, to use the second CD to boot from.
the reason it comes with that huge manual is to tell you all these things before you start and finish and go "DOH!":)
Slashdot Mirror
I can't remember but if this was ever true, but it isn't now.
Extreme Networks and Cisco both have long haul GBICs available, and have done for a little while.
I don't have info on the Cisco models available right now (Ciscos website is BIG!), but Extreme have a 1000BaseLX-70, that will do 70 km on singlemode dark fibre. On top if this they do what I could best call a gigabit fibre line driver in the form of a "SummitGbX"[tm]. They claim it will definately do 80 km, and possibly up to 100 km, I have heard one claim that these units managed to reach 120 km. Basically, you hook these up to your 1000BaseSX interfaces at each end, and it does some wibbly-bits to bridge the 80 km or whatever length of fibre :)
I would love to verify all this independently :) but I never even dared ask the prices :/
Of course, all this doesn't really help your cause much... you might be able to find a friendly telco that will blow fibre in for you. Apart from that, your realistic options might not include wireless.
At that range, you will have to go for some sort of microwave transmission, and even then you will probably have problems due to the Fresnel effect, which bends and scatters photons at the earths surface. The maths escapes me at this moment, but to reach 45 miles in one hop you would probably need to have each end nearly 200 foot up in the air in order to clear inconvenient obstacles in between, like buildings and trees (how careless to put them there!). To do it in several smaller hops might be easier, but then you have to rent or buy locations to put your repeater stations on.
Another possible thing to do would be to link all the schools locally to a central point using some easily available method. Microwave links might be suitable here, as well as optical wireless links, T1 leased lines, or whatever you chose. The most usefull central point would be a telco CO, which means that you will only have to rent backhaul bandwidth on the telco's network without having to pay for an expensive tail to anywhere else. Almost all CO's will be served by fibre now, which makes renting a fast connection very much cheaper and easier to provision. As you well know, it is that last mile that makes it expensive.
I wish you good luck :) maybe you'll let us know how it turns out?
btw, I don't have any connection with Cisco or Extreme, I just use their kit in my job
What version of Irix are you running?
Irix 6.5 comes with the kernel headers that you need for free.
maybe you could contact your local SGI office to see if that machine is entitled to run Irix 6.5, and if it is, where to get a media set from.
Once you have that, then you can use GCC to compile whatever you need.Alternatively, there is a good selection of freeware for SGIs at http://freeware.sgi.com
That includes a quite large amount of GPL software in binary form for you to download and install
This sounds like packet radio to me, Aloha protocol anybody? (a derivative of X.25).
:)
Of course, packet radio is only 9600 bps
Wouldn't it be cool if you could do this with broadband wireless though?
> On of the biggest problems with EFNet, has been that there is no 'ownership' of channels - or the
... you can't do that on efnet...
... if you don't look after the servers then you will never have an IRC network for the channels to exist in ..
> fact that no ircops participate in such matters.
This has always been the case, it always worked before, but now I suppose there is a better class of asshole wondering the internet.
and what do you think IRCops should do?
should they do what they are meant to, and look after servers?
or should they pander to thousands of whining users that accidently lost ops in a channel, and to hell with looking after a high load server
channel ownership, in my experience, only works in a small network where there is some feasibility of controlling the channels
so instead of having users fighting amongst themselves, the users would fight amongst themselves AND the IRCops, wasting their time and taking it away from their real purpose of looking after the servers
well, there is my 2p
I agree with this, I went to university here in england, to start with I was going for the bit of paper...
... it was meant to be "sotto vocce" but it came out just a little bit louder :)
...
:) even if i didn't get that piece of paper.
... so it all seems a bit pointless really
I realised soon that going to university for a bit of paper was completely the wrong thing, on my course myself and about a tenth of the rest of us (there were about 120 in my year doing CS) were quite obviously more knowledgable than most of the people who were supposed to be teaching us things.
I remember getting the most deathly stare for correcting one of the lecturer's a little more loudly than I intended
so why go to university/college?
all the stuff I learnt there had nothing to do with a degree or a bit of paper, all the usefull stuff was about how to look after myself and how to live.
all the cool people i met and the cool things I did
that is why you should go to university...
but having said that i left university, simply because i didn't think that the degree i was going to end up with would be worthwhile, the course was abysmal, the resources nearly non-existant. I had a good time though
mind you, you don't get bits of paper to say I can do what I can do
but this was the air that was at 210 F...
... 300 grams of water that have a LOT more energy in them than 300 grams of air, which would occupy a vastely greater volume.. and so have even less energy that can be transferred to something else
water has an enormous specific heat capacity, I don't know what it is, but I am sure I could find out if i were enthused enough.
a specific heat capacity of something is the number of joules of energy that it takes to raise a kilogram of that by one degree Celcius.
so, let's say a cup of coffee has 300 grams of water in it
it is the energy that burns, not the temperature
> Actually reworking SMT is not all that hard. I used to assemble PCMCIA video cards
..
...
... and then turn their Palm into a paperweight
:)
..
:)
> (204 pin TQFP, some SOJ, a handful of 1206 resistors and capacitors, etc.) with
> nothing more than a pair of tweezers and a fine pitch soldering iron.
Well, I have two points to make
1) you were presumably trained or taught yourself, you must have made a mess of things a lot of times before you learnt
2) you were presumably paid to have a skill to do this
you said yourself you made mistakes, how many people who read these are going to be able to achieve this, the only reason it isn't hard to you is because you have practise. The real big danger here is someone is going to see this, and try it
you make it sound simple, and I admit it is. but whilst what to do is simple enough, the actual doing of it isn't - it needs practise, skill and a steady hand. (or i suppose a powdered solder/flux mix you can paint on, and a soldering iron with a wide tip to heat a side at once
Back to the main reason for this post though
you say it is easy, when for 99% of people it isn't, and I had to point out the real world here
Well, I actually taught myself how to do this :)
:). that is a fantastically huge number of pins and I would never be able to keep my hand steady enough to do it without making a mess of the whole thing :(
:)
oh, and clipping all the pins IS a cheating method, but hey, it is how i would have done it, then your can remove the old solder and remains of legs at you leisure.
And yes, whilst soldering a new chip in is technically quite simple (hint: stick the chip down to hold it in place, but don't use massively strong glue in case you want to reposition it
all through this, you have to make sure you only apply as much heat as you need, because too much heat will make all the tracks on the board curl and peel off. I watched this happen once, it was a horrible sinking feeling in the pit of my stomach.... AND the damn solder still didn't melt
ok, time to stop waffling and go back to working
> I was just suddenly filled with excitement that I had an excuse to see the inside
:)
:)
> of my new toy. It was at this point that I realized Wow, I bet most people
> wouldn't be doing this.
hang on, you need an excuse?
I never needed an excuse...
though I did occasionally need a quick hand to put all the screws back in before my parents got back, but that was when i was much younger.
I guess niether of us are "most people"
This looked fascinating to me until I realised that you need to rework surface mounted components!
/. have tried this, but I have a few years ago. It is possibly one of the most incredibly difficult things you can attempt! I could move that little resistor they have marked, but to succesfully remove an entire DRAM chip and resolder a new one would require some extraordinary skill!
.. good luck :)
;)
I don't know how many people who read
Even when I was at my most practised and my hands seemed rock steady to me I would NEVER have attempted this!
even now i could remove the chip, but to put in place a new one would be unthinkable
Please no-one try this unless you are completely sure you can pull it off, I have visions of hundreds of wrecked palms from people try to do this.
If you DO want to try it still
but don't come crying if your Palm gets amnesia
Any machine is only as secure as the administrator makes it.
imho there are a lot of things in the article that aren't the result of Debian being "bad" they are the result of incompetence or of overlooking obvious details.
ok, so maybe we should do what this poster suggests, but if you can't handle doing basic security-wise configuration, you shouldn't be setting up a machine that needs it. Sometimes you just have to take responsibility for things, you can't just say "suchandsuch should have done thingummybob".
Demon still use SMTP mail delivery as the main method ... .. ie, not only is there an IP for luser.demon.co.uk but if they upload web pages, an IP is assigned by demon to www.luser.demon.co.uk !
and don't forget that Demon Internet (now part of Thus) also has a static IP for each dialup customer that has a homepage
I wonder if demon will change all their dialup users homepages to name based virtual hosting and give back however many thousands of IPs this will free up
I don't know about anywhere else, but at the ISP I work at, we only give a user a static IP if they can justify it, then we can justify it to RIPE
fair's fair, right?
It doesn't really matter to us, or anyone else .. it would work the same and still give an encrypted connection ..
.. and of course as someone else posted, without the certificate checking you can mount a "man-in-the-middle" attack.
.. can anyone confirm/disprove this?
but, people would notice, customers _always_ somehow manage to notice the least little thing, all they see is the box that comes up, they wouldn't look at the little padlock that says the connection is encrypted
also, the thought occurs that when you generate the CSR (certificate signing request) it uses the server key, the server key has the servers name in it doesn't it? so the servers _name_ will always be in the certificate, and fail to match with the address that someone typed in to access one of these virtually hosted SSL sites.
surely a browser would alert someone to this discrepency?
This will probably immediately send J. Random Bloggs into a fit of panic that it isn't a secure connection, regardless of whether it really is or not.
This would make you and your company look _bad_
I can't actually test whether this happens or not, not tending to have a spare SSL certificate lying around to play with
> By the way is this really a problem? Wouldn't
.. both of these and just theories of course ..
> you want a dedicated server anyway to host your
> site if it has to hanlde confidencial data? I
> don't like the idea of a Web-shop running on a
> 100+ site web-hotel, storing my credit card
> number - SSL or no SSL
well, you can either secure the system hosting this "web-hotel" completely and partition all the users, or you can arrange to store the confidential details on a specially designed and secured repository outside of that server
or, you can pay us a lot of money and we (and most of the other ISPs in the world) will build, install and host a dedicated server for you
This is a good point, but you have to choose to do something that is feasible, sensible and cost effective. You have to make a decision about which trade-off's to accept. (just like anything else really)
they may be issued on a name basis, but the problem here is that SSL is a transport, you have to negotiate the SSL link complete with the certificate before you get to talk to the actual web server ... at this point the server doesn't know which web site you are looking for, and therefor has no way to know which certificate it should send.
... a single SSL site on one server generates 3 times as much traffic as the whole of the other sites on that server, because the normal sites can be accelerated, SSL can't.
... how do we fix the SSL https issue?
.. but I can't see how
Where I work right now, SSL is one of the biggest problems, we have 5 servers here running host-header based virtual hosting, but we have had to set aside relatively large chunks of our IP space to cater for the customers who want SSL.
To top this off, the SSL-hosting IPs can only do one thing each, and cannot be accelerated by our caching system
So
I would love to do name based SSL hosting
> Perhaps all telcos should provide open source, impartial network monitoring software in order
> to self-regulate the quality of service. In Britain we have a telco watchdog that would
> stomp all over this thing.
ok, so why hasn't it stomped over BT and their DREADFULL handling of DSL, imho OFTEL are a lame duck and have precious little bite against any telco
or if you had read the documentation, you would know to type "manual" at the LILO: prompt to get the choice of YaST 1 or 2 :)
:)
or alternatively, as another poster suggests, to use the second CD to boot from.
the reason it comes with that huge manual is to tell you all these things before you start and finish and go "DOH!"