Wireless-enabled equipment required. Existing customers not upgrading their package can buy a Virgin Media Hub for £60 or a Virgin Media SuperHub for £75. Equipment remains property of Virgin Media.
Ah yes virgin the kings of charging the customer upfront for something and then keeping it as their property anyway. ASSHOLES.
Though it's only recently that virgin media have supplied routers at all, it used to be that they only supplied a cable modem and if you wanted to hook up more than one machine you had to supply your own router (and IIRC having a router was tolerated but not supported).
I think your 80% figure is wildly optimistic. Between websites that don't offer v6, non-v6 capable routers and non v6 capable end systems (IIRC XP doesn't support IPv6 unless manually enabled on the command line) I would expect far more than 20% of traffic to remain v4 in a scenario where the ISP offeres IPV6+Natted IPv4.
The ISP could try to push things along with new routers and tutorials but unless they offer some largecarrot and/or used some large stick (for example blocking IPv4 acess to youtube) I doubt many users would take them up on the offer unless they had no choice.
APNIC has now run out of IPv4 addresses, so I imagine that some services in the Asia-Pacific region will start to be v6-only in the near futures
Maybe some minor ones but I would expect any service of importance to get a v4 address. When the choice is to give a home luser paying $xx per month the IP or give a server whose owner is paying $xxx or possiblly $xxxx per month the IP who do you think will get it?
Also of note is that unlike most RIRs apnic allows IPs to be transferred freely so it's probablly easier to just buy a block of IPs (rather than buying a block of IPs with strings attached such as having to use a particlar ISP).
Hmm, in my experiance it's usual to get them from the ISP and it may or may not be free but I was under the impression that the router generally became the property of the user.
I know when ADSL first came out BT owned the modem (this was before routers became common) but IIRC at some point they decided it was too much hassle and gave the modems to the users.
If I don't post it back when I cancel the contract they'll charge me for it.
You can, and it is called NAT46. The problem is that it is not stateless
The problem is that addresses for the v4 side of the mapping have to be taken from a limited pool (most likely some subset of NET10) and they have to be shared between the NAT46 box and the DNS server. This raises two issues.
1: not everyone uses their ISPs DNS. 2: even if a user is using their ISPs DNS there is no gaurantee they will be using the most local one
Furthermore some ISPs already have heavy pressure on NET10 (or have run out of NET10 addresses completely) for other uses. Adding mapping addresses as yet another load on net10 is probablly something they want to avoid.
All in all it's a massive headache for an ISP to solve what is most likely a non-problem. IPv6 will likely get used for peer-peer stuff and some client-server connections but all the important services are likely to remain available on v4 for a long time.
Such mapping could be done at the home router level but since they seem to be about the last things to get upgraded in any way I wouldn't hold out much hope.
They don't have much choice. It is pretty clear that v4 addresses will run out for at least some ISPs before we reach a situation where a v6 only connection is acceptable for clients. Therefore it is pretty much inevitable that ISPs will have to deploy some sort of ISP level NAT regardless of whether they also deploy IPv6.
Hosting companies such as Go Daddy charge per IP address. And given that a lot of deployed web browsers still require a distinct IPv4 address for each distinct site [wikipedia.org], SSL site operators have to pay up.
The thing is until v4 only clients become negligable (not likely to happen any time soon) SSL site operators will have to pay up for an IPv4 address. This applies REGARDLESS of whether they also deploy IPv6.
IIRC the reason hosting providers charge per IP is more to do with the paperwork and because they can than because of a shortage of IPs.
If I was running an ISP i'd be trying to MAXIMISE the number of v4 addresses I could justify right now. The more addresses an ISP gets before the pool runs dry the more addresses they have to re-allocate in the post-exhaustion world.
Doubtless ISPs will offer a real, honest-to-FSM IPv4 address, but they won't offer it to domestic subscribers.
Maybe i'm an optimist but I'd hope that ISPs would make public v4 addresses available for a small extra charge. It would be a new revenue stream for them and they really only need to get the lusers behind NAT to free up more than enough IPs for their other uses.
While kind of true this is rather misleading because not all network stacks have all the layers from the OSI model. In particular TCP/IP doesn't have a "presentation" (OSI layer 5) or "session" (OSI level 6) layer. So in reality the web server app is only two layers above IP.
They are supposed to be completely independent.
While that is a nice idea it's a bit of a fairy tale. In reality certain things have to pass between layers. In most protocol stacks one of those things is network addresses. IPv6 uses a completely new addressing system that is totally seperate from the IPv4 one (there are mappings that are used in certain contexts to present an IPv4 address as if it was an IPv6 address but when actually on the network the two systems are completely seperate). So for a website to receive IPv6 connections all of the following must be in place.
The machine running the webserver must have v6 connectivity. The web server software must support IPv6 and if IP based virtual hosting is in use (as it will be for most SSL websites due to issues with SSL and name based vitual hosting) it must be configured with which v6 IPs to map to which site. The admins must create DNS records to tell the client the sites v6 IP addresses.
The thing is websites are actually one of the services with the least motivation to offer v6. They will need to offer v4 as long as there are some v4 only clients (read: a long time) and the web works just fine through NAT.
Well there was the HP omnigo 700LX which apparently came out in 1996 and ran DOS (apparently it was even capable of running windows 3.0 if a big enough flash card was placed in the PCMCIA slot). It didn't have the phone built in but had a slot in the back in which a nokia 2110 was piggybacked. Does that count as a smartphone?
A big problem for all the early smartphones was that mobile data at the time was both circuit switched (so you had to dial-up to an ISP before doing anything) and very slow. Then GRPS came along but carriers priced it so high that hardly anyone could justify using it. It was many years before prices on mobile data came down to a level where it was reasonable for ordinary folks to use it.
China needs to take chunks of its manufacturing and relocate it to the USA
But why would a Chinese firm want to set up manufacturing in a country that was both far from home and expensive to operate in? Unless there is either a huge currency change, huge domestic changes (either in the US, china or both) in wages and other costs of doing buisness or huge tarrifs blocking easy sale of Chinese products in the USA I just don't see the attraction to a Chinese firm of operating in the USA.
This way all applications that can print can save to PDF-files. For efficient then implementing it in every program.
Which is fine if you don't care about things like exporting the table of contents, hyperlinks and such like. Going through a printer driver tends to destroy that sort of stuff.
I found a tool called PDF-T-MAKER that semed to work for producing decent PDF but was a massive pain to set up. I can see how a quality PDF exported built into word would be a substantial threat to adobes buisness as producer of the PDF production tool for those who want more than a basic PDF printer driver can offer.
These days I've just given up on word for complex documents and use LaTeX instead.
Using NAT at the ISP level is basicly evil and should not be considered
Unfortunately the fact that most people have adopted a "don't do anything until we have to" approach means that some form of NAT (possiblly with protocol translation) at the ISP level is pretty much inevitable.
The facts: * IPv4 addresses have basically run out. From here on the only way you will get one in asia is to buy it off someone else or be in a special class of user. The same will happen in the rest of the world in the next year or so. * Many customers are still using V4 only equipment (desktop PC operating systems have supported v6 for years but other stuff hasn't) and will be reluctant to replace it * while some ISPs are growing quickly and will need to find a soloution right now others are not really growing much and have no pressing problem to solve. * it is impractical to host multiple SSL websites with different owners on one IP while windows XP is still in widespread use.
To me these facts imply. * new customers on many ISPs will not be able to have public v4 addresses * eventually existing customers will also lose their public v4 addresses to free them up for more lucrative uses. * those customers will still want to be able to use their existing equipment. At the most they can probablly be convinced to replace their broadband router but probablly not replace or significantly reconfigure (IIRC XP requires use of a command line to activate IPv6) the stuff behind it.
Realistically this results in three options ISPs must choose to tide them over between the time when they can't give every customer a public v4 address and the time that all websites offer IPv6 and all customer equipment supports ipv6 seamlessly. Afaict there are three soloutions in the running none of them pretty and all involving some form of NAT.
1: NAT464 (translate requests from v4 to v6 at the customers premisis and then back from v6 to v4 at the ISP) This option does have the advantage that it provides the most seamless transition to pure v6 but gatewaying v4 clients to v6 servers is about the ugliest type of NAT arround because of the need for statefull DNS translation. As such I would expect few ISPs to go down this route. New customer premisis equpment would also be needed. 2: NAT444: do v4 nat twice, once at the customers premisis and once at the ISP) This avoids the need for new customer premis equipment but double NAT can cause problems and there is also the issue of what address space to use (most ISPs going down this road seem to be using 10.0.0.0/8) and in particular the fact that afaict linux NAT and other common implementation are not designed to cope with both sides of a NAT using the same IP space so there is potential for conflicts. For some huge ISPs there are also problems with running out of space in 10.0.0.0/8 (comcast has this problem). Still i'd expect this to be the option most ISPs choose. 3: DS-LITE: The customer premisis equipment sends v4 packets with private addresses down a special tunnel to the ISP's NAT gateway. The NAT at the ISP looks at which tunnel a packet came down as well as it's private address so there is no need for NAT at the customer premisis (and hence no need for double NAT). IIRC comcast created this soloution because they couldn't easilly deploy soloution 2 due to 10.0.0.0/8 exhaustion in their internal networks. As with soloution 1 new customer premisis equipment will be needed but unlike soloution 1 there is no need to mess with DNS.
Whichever way an new or existing ISP choses to go they will need a pool of v4 IPs for the requests they translate (either from private v4 or from v6) to public v4 and that is why APNIC have set asside these addresses for new ISPs to avoid new ISPs being locked out of the market completely.
He said he pulled the key from the ROM - by definition that can't be updated.
As someone who does embedded stuff the term "ROM" is often used very losely to include things like EPROM, EEPROM, flash (of the type that acts as a rom) etc.
Depends on the type of rom and the design of the rest of the device.
Afaict most embedded systems at this level (the really small stuff uses on-chip flash) use a paralell flash chip connected direct to the main processor bus. To dump this you attatch a JTAG cable and use software on the PC to take control of the processors busses and read out the chip.
I beleive some systems use a serial flash chip that is copied into ram. I would expect this could be read out in a similar manner by using JTAG to control the interface pins.
Software for accessing flash chips in this manner is usually readilly available because it is what is used to program the chips during development and possiblly during production as well.
Some systems attempt to protect their firmware against such readout attempts but router vendors tend not to bother.
Desoldering the flash chip and reading it out directly would be another option but afaict that is usually a last resort.
Unless an ISP wants every PS3/xBox/Skype user/PC-Gamer to be mass calling their tech support and asking why their internet is broken, they will continue to hand out public IPs.
Sure they will until they are on the brink of running out and the RIRs refuse to supply them with more.
After that they will have to decide how to allocate the limited pool of IPs they have and given that they are buisnesses I would expect them to allocate them to where they will generate the most profit. To start with I would expect them to apply the "squeaky wheel" tactic and give back public IPs to those who bitch too much and are too smart to have the blame deflected. Eventually it may reach the point where if you want a public IP you have to pay for the privilage. Providers of internet based services will have to learn to adapt to a world where some proportion of their users are behind ISP level NAT.
I wouldn't expect them to make an instant cutover because there is no point, as you say it would put a huge load on tech support for no immediate gain. Instead I would expect IPs to be recovered a few hundred at a time as they are needed for more lucrative services.
Computers generally either work or don't work, and rarely do they half work, or generally slow down.
ROFLMAO
I've seen loads of cases where computers sort of work but are either much slower than expected and/or crash frequently. Sometimes it's a software problem (e.g. the bonzai buddy you mention but also good old viruses or even just bad but "legit" software), sometimes it's a hardware problem (e.g. overheating processor, dying hard drive or faulty ram).
If software is changing
Unless you live in a world without virus scanners and with no patches software does change. The basic software may stay at the same headline versions but patches, virus scanner updates and whatever other crap IT dreams up tends to increase the load on machines over time.
That is less then 60,000 possible connections can be made by a router with a single public IP address.
That depends on how clever the NAT is. Technically each server you talk to doesn't know what ports you are using to talk to each other servers. So there is nothing stopping a nat using an internet side port to talk to multiple servers at the same time. Such a scheme will completely any protocol that tries to do "nat traversal" but it should keep the basics working at very high user:IP ratios.
Still I would expect IPv6 to seep in if only to try and reduce the load on the big nats.
IPs aren't UUIDs as such but they aren't really that heirechical either. They are handed out by the RIRs in various size blocks and each of those blocks (and sometimes even sub-blocks of it) ends up in the global routing table. Very small providers will take a portion of one of their providers blocks but most bigger organisations will have their own block(s).
Running out of space in routing tables is a potential issue but at least so far the vendors have been able to keep up with routing table growth.
When does IPv4 not just run out, but get painfully expensive to acquire?
Indeed, at least in the west most home lusers still have public V4 IPs. I would expect ISPs to gradually reclaim those IPs for more lucrative customers and so it will be a while (possiblly a decade) before the shortage really bites on western ISPs.
It is over in the east that things are REALLY going to get hairy with so many new users coming online that I would expect IP values to dramatically rise. ISP level nat will help to an extent but there are limits on the ratios that can practially be used. I would expect them to try importing IPs but I don't know whether the IANA and the RIRs will let them get away with it.
By train: 1h to London St Pancras, 30m check in time, 2h 15m to Paris Gare du Nord, total time, 3h 45m
They say on a standard ticket you are required to check in at least 30 minuites before departure, they don't say what happens if you miss that time. Given it will take some time to walk accross the station and trains in the UK are frequently late i'd say you should be planning to arrive in london arround an hour before you are due out again. Also IIRC heathrow is one of the worst airpots for check in times. Still you are right if you actually want to end up in the center of paris then the train is probablly about the same as flying and maybe quicker.
OTOH If you don't actually want to end up in the center of paris and/or you don't live close to london on a train line that happens to end up at kings cross/st pancras it's probablly quicker to fly.
Another big difference is that small scale electricty generation has high running costs and swithing between two sources of electricty can be done very quickly after the initial source fails. This means that even if you find the centralised provider too unreliable on their own it STILL makes sense to use them most of the time and only use your local generation as a backup.
With computing if you need the resources locally some of them time (say because your internet connection is unreliable) it makes sense to have them local all the time rather than try to implement some kind of automatic switch when your internet connection goes down.
Are you seriously saying you would run a server farm under oil without the cooperation of whoever made the servers?! I'd be far more comfortable swapping the cooling components for watercooling ones than immersing unknown materials in oil.
Slashdot Mirror
Wireless-enabled equipment required. Existing customers not upgrading their package can buy a Virgin Media Hub for £60 or a Virgin Media SuperHub for £75. Equipment remains property of Virgin Media.
Ah yes virgin the kings of charging the customer upfront for something and then keeping it as their property anyway. ASSHOLES.
Though it's only recently that virgin media have supplied routers at all, it used to be that they only supplied a cable modem and if you wanted to hook up more than one machine you had to supply your own router (and IIRC having a router was tolerated but not supported).
I think your 80% figure is wildly optimistic. Between websites that don't offer v6, non-v6 capable routers and non v6 capable end systems (IIRC XP doesn't support IPv6 unless manually enabled on the command line) I would expect far more than 20% of traffic to remain v4 in a scenario where the ISP offeres IPV6+Natted IPv4.
The ISP could try to push things along with new routers and tutorials but unless they offer some largecarrot and/or used some large stick (for example blocking IPv4 acess to youtube) I doubt many users would take them up on the offer unless they had no choice.
APNIC has now run out of IPv4 addresses, so I imagine that some services in the Asia-Pacific region will start to be v6-only in the near futures
Maybe some minor ones but I would expect any service of importance to get a v4 address. When the choice is to give a home luser paying $xx per month the IP or give a server whose owner is paying $xxx or possiblly $xxxx per month the IP who do you think will get it?
Also of note is that unlike most RIRs apnic allows IPs to be transferred freely so it's probablly easier to just buy a block of IPs (rather than buying a block of IPs with strings attached such as having to use a particlar ISP).
In the UK it's unusual to buy/own a router.
Hmm, in my experiance it's usual to get them from the ISP and it may or may not be free but I was under the impression that the router generally became the property of the user.
I know when ADSL first came out BT owned the modem (this was before routers became common) but IIRC at some point they decided it was too much hassle and gave the modems to the users.
If I don't post it back when I cancel the contract they'll charge me for it.
Out of interest which ISP is this.
You can, and it is called NAT46. The problem is that it is not stateless
The problem is that addresses for the v4 side of the mapping have to be taken from a limited pool (most likely some subset of NET10) and they have to be shared between the NAT46 box and the DNS server. This raises two issues.
1: not everyone uses their ISPs DNS.
2: even if a user is using their ISPs DNS there is no gaurantee they will be using the most local one
Furthermore some ISPs already have heavy pressure on NET10 (or have run out of NET10 addresses completely) for other uses. Adding mapping addresses as yet another load on net10 is probablly something they want to avoid.
All in all it's a massive headache for an ISP to solve what is most likely a non-problem. IPv6 will likely get used for peer-peer stuff and some client-server connections but all the important services are likely to remain available on v4 for a long time.
Such mapping could be done at the home router level but since they seem to be about the last things to get upgraded in any way I wouldn't hold out much hope.
They don't have much choice. It is pretty clear that v4 addresses will run out for at least some ISPs before we reach a situation where a v6 only connection is acceptable for clients. Therefore it is pretty much inevitable that ISPs will have to deploy some sort of ISP level NAT regardless of whether they also deploy IPv6.
Hosting companies such as Go Daddy charge per IP address. And given that a lot of deployed web browsers still require a distinct IPv4 address for each distinct site [wikipedia.org], SSL site operators have to pay up.
The thing is until v4 only clients become negligable (not likely to happen any time soon) SSL site operators will have to pay up for an IPv4 address. This applies REGARDLESS of whether they also deploy IPv6.
IIRC the reason hosting providers charge per IP is more to do with the paperwork and because they can than because of a shortage of IPs.
If I was running an ISP i'd be trying to MAXIMISE the number of v4 addresses I could justify right now. The more addresses an ISP gets before the pool runs dry the more addresses they have to re-allocate in the post-exhaustion world.
Doubtless ISPs will offer a real, honest-to-FSM IPv4 address, but they won't offer it to domestic subscribers.
Maybe i'm an optimist but I'd hope that ISPs would make public v4 addresses available for a small extra charge. It would be a new revenue stream for them and they really only need to get the lusers behind NAT to free up more than enough IPs for their other uses.
Bunk! Websites are layer 7. IPv6 is layer 3.
While kind of true this is rather misleading because not all network stacks have all the layers from the OSI model. In particular TCP/IP doesn't have a "presentation" (OSI layer 5) or "session" (OSI level 6) layer. So in reality the web server app is only two layers above IP.
They are supposed to be completely independent.
While that is a nice idea it's a bit of a fairy tale. In reality certain things have to pass between layers. In most protocol stacks one of those things is network addresses. IPv6 uses a completely new addressing system that is totally seperate from the IPv4 one (there are mappings that are used in certain contexts to present an IPv4 address as if it was an IPv6 address but when actually on the network the two systems are completely seperate). So for a website to receive IPv6 connections all of the following must be in place.
The machine running the webserver must have v6 connectivity.
The web server software must support IPv6 and if IP based virtual hosting is in use (as it will be for most SSL websites due to issues with SSL and name based vitual hosting) it must be configured with which v6 IPs to map to which site.
The admins must create DNS records to tell the client the sites v6 IP addresses.
The thing is websites are actually one of the services with the least motivation to offer v6. They will need to offer v4 as long as there are some v4 only clients (read: a long time) and the web works just fine through NAT.
Well there was the HP omnigo 700LX which apparently came out in 1996 and ran DOS (apparently it was even capable of running windows 3.0 if a big enough flash card was placed in the PCMCIA slot). It didn't have the phone built in but had a slot in the back in which a nokia 2110 was piggybacked. Does that count as a smartphone?
A big problem for all the early smartphones was that mobile data at the time was both circuit switched (so you had to dial-up to an ISP before doing anything) and very slow. Then GRPS came along but carriers priced it so high that hardly anyone could justify using it. It was many years before prices on mobile data came down to a level where it was reasonable for ordinary folks to use it.
On the other hand IIRC there have been underground coal mine fires that have rendered the towns on top unsafe for human habitation.
China needs to take chunks of its manufacturing and relocate it to the USA
But why would a Chinese firm want to set up manufacturing in a country that was both far from home and expensive to operate in? Unless there is either a huge currency change, huge domestic changes (either in the US, china or both) in wages and other costs of doing buisness or huge tarrifs blocking easy sale of Chinese products in the USA I just don't see the attraction to a Chinese firm of operating in the USA.
This way all applications that can print can save to PDF-files. For efficient then implementing it in every program.
Which is fine if you don't care about things like exporting the table of contents, hyperlinks and such like. Going through a printer driver tends to destroy that sort of stuff.
I found a tool called PDF-T-MAKER that semed to work for producing decent PDF but was a massive pain to set up. I can see how a quality PDF exported built into word would be a substantial threat to adobes buisness as producer of the PDF production tool for those who want more than a basic PDF printer driver can offer.
These days I've just given up on word for complex documents and use LaTeX instead.
Using NAT at the ISP level is basicly evil and should not be considered
Unfortunately the fact that most people have adopted a "don't do anything until we have to" approach means that some form of NAT (possiblly with protocol translation) at the ISP level is pretty much inevitable.
The facts:
* IPv4 addresses have basically run out. From here on the only way you will get one in asia is to buy it off someone else or be in a special class of user. The same will happen in the rest of the world in the next year or so.
* Many customers are still using V4 only equipment (desktop PC operating systems have supported v6 for years but other stuff hasn't) and will be reluctant to replace it
* while some ISPs are growing quickly and will need to find a soloution right now others are not really growing much and have no pressing problem to solve.
* it is impractical to host multiple SSL websites with different owners on one IP while windows XP is still in widespread use.
To me these facts imply.
* new customers on many ISPs will not be able to have public v4 addresses
* eventually existing customers will also lose their public v4 addresses to free them up for more lucrative uses.
* those customers will still want to be able to use their existing equipment. At the most they can probablly be convinced to replace their broadband router but probablly not replace or significantly reconfigure (IIRC XP requires use of a command line to activate IPv6) the stuff behind it.
Realistically this results in three options ISPs must choose to tide them over between the time when they can't give every customer a public v4 address and the time that all websites offer IPv6 and all customer equipment supports ipv6 seamlessly. Afaict there are three soloutions in the running none of them pretty and all involving some form of NAT.
1: NAT464 (translate requests from v4 to v6 at the customers premisis and then back from v6 to v4 at the ISP) This option does have the advantage that it provides the most seamless transition to pure v6 but gatewaying v4 clients to v6 servers is about the ugliest type of NAT arround because of the need for statefull DNS translation. As such I would expect few ISPs to go down this route. New customer premisis equpment would also be needed.
2: NAT444: do v4 nat twice, once at the customers premisis and once at the ISP) This avoids the need for new customer premis equipment but double NAT can cause problems and there is also the issue of what address space to use (most ISPs going down this road seem to be using 10.0.0.0/8) and in particular the fact that afaict linux NAT and other common implementation are not designed to cope with both sides of a NAT using the same IP space so there is potential for conflicts. For some huge ISPs there are also problems with running out of space in 10.0.0.0/8 (comcast has this problem). Still i'd expect this to be the option most ISPs choose.
3: DS-LITE: The customer premisis equipment sends v4 packets with private addresses down a special tunnel to the ISP's NAT gateway. The NAT at the ISP looks at which tunnel a packet came down as well as it's private address so there is no need for NAT at the customer premisis (and hence no need for double NAT). IIRC comcast created this soloution because they couldn't easilly deploy soloution 2 due to 10.0.0.0/8 exhaustion in their internal networks. As with soloution 1 new customer premisis equipment will be needed but unlike soloution 1 there is no need to mess with DNS.
Whichever way an new or existing ISP choses to go they will need a pool of v4 IPs for the requests they translate (either from private v4 or from v6) to public v4 and that is why APNIC have set asside these addresses for new ISPs to avoid new ISPs being locked out of the market completely.
He said he pulled the key from the ROM - by definition that can't be updated.
As someone who does embedded stuff the term "ROM" is often used very losely to include things like EPROM, EEPROM, flash (of the type that acts as a rom) etc.
Depends on the type of rom and the design of the rest of the device.
Afaict most embedded systems at this level (the really small stuff uses on-chip flash) use a paralell flash chip connected direct to the main processor bus. To dump this you attatch a JTAG cable and use software on the PC to take control of the processors busses and read out the chip.
I beleive some systems use a serial flash chip that is copied into ram. I would expect this could be read out in a similar manner by using JTAG to control the interface pins.
Software for accessing flash chips in this manner is usually readilly available because it is what is used to program the chips during development and possiblly during production as well.
Some systems attempt to protect their firmware against such readout attempts but router vendors tend not to bother.
Desoldering the flash chip and reading it out directly would be another option but afaict that is usually a last resort.
Unless an ISP wants every PS3/xBox/Skype user/PC-Gamer to be mass calling their tech support and asking why their internet is broken, they will continue to hand out public IPs.
Sure they will until they are on the brink of running out and the RIRs refuse to supply them with more.
After that they will have to decide how to allocate the limited pool of IPs they have and given that they are buisnesses I would expect them to allocate them to where they will generate the most profit. To start with I would expect them to apply the "squeaky wheel" tactic and give back public IPs to those who bitch too much and are too smart to have the blame deflected. Eventually it may reach the point where if you want a public IP you have to pay for the privilage. Providers of internet based services will have to learn to adapt to a world where some proportion of their users are behind ISP level NAT.
I wouldn't expect them to make an instant cutover because there is no point, as you say it would put a huge load on tech support for no immediate gain. Instead I would expect IPs to be recovered a few hundred at a time as they are needed for more lucrative services.
Computers generally either work or don't work, and rarely do they half work, or generally slow down.
ROFLMAO
I've seen loads of cases where computers sort of work but are either much slower than expected and/or crash frequently. Sometimes it's a software problem (e.g. the bonzai buddy you mention but also good old viruses or even just bad but "legit" software), sometimes it's a hardware problem (e.g. overheating processor, dying hard drive or faulty ram).
If software is changing
Unless you live in a world without virus scanners and with no patches software does change. The basic software may stay at the same headline versions but patches, virus scanner updates and whatever other crap IT dreams up tends to increase the load on machines over time.
That is less then 60,000 possible connections can be made by a router with a single public IP address.
That depends on how clever the NAT is. Technically each server you talk to doesn't know what ports you are using to talk to each other servers. So there is nothing stopping a nat using an internet side port to talk to multiple servers at the same time. Such a scheme will completely any protocol that tries to do "nat traversal" but it should keep the basics working at very high user:IP ratios.
Still I would expect IPv6 to seep in if only to try and reduce the load on the big nats.
IPs aren't UUIDs as such but they aren't really that heirechical either. They are handed out by the RIRs in various size blocks and each of those blocks (and sometimes even sub-blocks of it) ends up in the global routing table. Very small providers will take a portion of one of their providers blocks but most bigger organisations will have their own block(s).
Running out of space in routing tables is a potential issue but at least so far the vendors have been able to keep up with routing table growth.
When does IPv4 not just run out, but get painfully expensive to acquire?
Indeed, at least in the west most home lusers still have public V4 IPs. I would expect ISPs to gradually reclaim those IPs for more lucrative customers and so it will be a while (possiblly a decade) before the shortage really bites on western ISPs.
It is over in the east that things are REALLY going to get hairy with so many new users coming online that I would expect IP values to dramatically rise. ISP level nat will help to an extent but there are limits on the ratios that can practially be used. I would expect them to try importing IPs but I don't know whether the IANA and the RIRs will let them get away with it.
It was officially PS2 only but there is now an unofficial PC version.
By train: 1h to London St Pancras, 30m check in time, 2h 15m to Paris Gare du Nord, total time, 3h 45m
They say on a standard ticket you are required to check in at least 30 minuites before departure, they don't say what happens if you miss that time. Given it will take some time to walk accross the station and trains in the UK are frequently late i'd say you should be planning to arrive in london arround an hour before you are due out again. Also IIRC heathrow is one of the worst airpots for check in times. Still you are right if you actually want to end up in the center of paris then the train is probablly about the same as flying and maybe quicker.
OTOH If you don't actually want to end up in the center of paris and/or you don't live close to london on a train line that happens to end up at kings cross/st pancras it's probablly quicker to fly.
Another big difference is that small scale electricty generation has high running costs and swithing between two sources of electricty can be done very quickly after the initial source fails. This means that even if you find the centralised provider too unreliable on their own it STILL makes sense to use them most of the time and only use your local generation as a backup.
With computing if you need the resources locally some of them time (say because your internet connection is unreliable) it makes sense to have them local all the time rather than try to implement some kind of automatic switch when your internet connection goes down.
Are you seriously saying you would run a server farm under oil without the cooperation of whoever made the servers?! I'd be far more comfortable swapping the cooling components for watercooling ones than immersing unknown materials in oil.