Using four unsigned longs is horrible please read out this address: 536937584.520688358.33825791.4270907832
The hex quads can at least be read out without error.
Quite frankly a 128 bit address is too large for almost anyone to remember you have to use names. You'll probably get a/64 that means that the first half of the address isn't controlled by you eg: 2001:470:1f09:12e6::/64. you can then assign any you like for the other half eg: 2001:470:1f09:12e6::7 even making everything as nice as I can your version would be 536937584.520688358.0.7, horrific.
The double colon is a neat feature eg: 2a00:1450:8006::68 but multiple sets are ambiguous, you'd have to start counting colons, and that's a shitty job.
Except it's not a jungle, it's more like the space between galaxies.
You throw a packet into the IPv4 address space and you'll probably hit a machine but it's a different scale in IPv6 space.
I have a little solar system, a few tiny points of light in a single/64. There's a (very) simple firewall across the whole/64 but I've yet to see a single packet that's even a slightly possible candidate for an unsolicited connection of any sort. The few packets the firewall has blocked are traceable to leftovers of old connections.
I'm not using Teredo, I've signed up of a 'Hurricane Electric' 6to4 tunnel (no charge) so I've got static IPv6 addresses but even just the Teredo section of the address space is vast and scanning it is basically impossible.
Stupid EULA terms, which are generally not read and not enforceable unless you're a huge target like Debian.
On the cheap (dumb) programming. The theory is you can write the BLOB using cheap programmers doing barely tested 'piece work' and fix the inevitable problems later. With full true hardware designs you have just one chance to get it right.
But the same mentality means that if a problem in the BLOB doesn't make Windows crash too much then it's just ignored and will never be fixed. This is ALWAYS seen as a buggy Linux driver.
If we were talking satellite data you'd be right, a few gigabits per second shared between the 20 million people in the footprint of one satellite is never going to be enough, and going much higher would be changing the laws of physics. (Shannon's law IIRC)
But enough 8kbit/s voice channels could fit. The problem is money.
That's the point; by a 'wavebox' I mean a machine that runs the wave locally so 90% (asstat) of the traffic doesn't even touch Google's servers. Google would be used mainly as the trusted third party to startup a wave.
That way when you lose access to the Google account it actually can be a few days before it gets to be a problem.
Though, that was probably the bit I wasn't thinking through. Expecting Google set up a communication link that they can't monitor... what was I thinking, I'm not usually that dumb!
Sure the internet is reliable in your basement and there's loads of open wifi's in the student block next door.
My DSL is pretty reliable too; no basements or students round here though so Wifi is in short supply. You could notice it too if you went further than the local Starbucks; not everywhere has internet connection in fact some of the best places to visit you don't get any internet connection for days.
Personally I thought the first GoogleOS netbook was going to be a Wavebox for just this reason.
But joined up thinking was never a corporate strong point.
Linux and BSD the OS's under Android and iPhone both have solid security tools. Linux's version was written by the NSA FFS. But once the machine leaves the hands of G&A the 'integrators' have full control over what goes in and what stays out. For Linux the major security enhancements can be turned off with a single switch & kernel recompile.
The only way either company could force the issue is to use legal means and renegotiate their agreements with these 'integrators' AKA 'Phone companies'.
We need everyone to verify their age via CC or SS to browse the web
The ISP ALREADY DOES THIS!
They even bill you every month so you won't forget you took responsibility for the connection.
What can they do if you abuse this trust and let someone else browse the web?
As for trusting some random website owner can read you mind and work out what you would be offended by and give a shit once they've done this... are you a complete idiot, a moron or just a dolt!?
Your message better be a sarcasm overload, because if you can't learn you have a severe retardation problem.
I've done a quick scan and the autogroups thing creates a new scheduling group whenever a user process sets or clears the current CTTY value for the process. This should give a process grouping almost identical to the session id (setsid(3)). You can see your current session ids with ps(1) in the SID column.
ps jaxk sid
I can't tell for sure without trying it but I think this will give a new process group for each application started from many GUI menu programs too because they actively detach you from the CTTY. Even if it doesn't there are many user programs that change the CTTY, the simplest is probably script(1).
But, this is definitely one of those changes that demands the question of why it wasn't though of before; why didn't I think of it, I know what session IDs are and I know about process scheduling groups. It's obvious, in hindsight, that they are a pretty good match for each other.
I dunno who you mean by "The GPL", might be the FSF but they're a charity and a stand in copyright holder.
There are quite a few more sites than the one you mention that release hardware designs (GPL and other free licenses) but they run into the problem that there are significant costs to duplication and distribution of hardware. So the closest you're going to get to your silly example is initially a single sponsor leading to a large group of people and companies. Say perhaps Google, they could sponsor an open phone design, they'd need a good name like "the robot phone" or perhaps something a little more human. They would make a lot of cold hard cash out of free software.
The universities were a good source, but now the better sources of paid for free software development are companies, like the afore mentioned Google, or IBM or other similar (mostly) service oriented companies. Companies that know their assets are their people and the skills they have. Companies that think of their software as an asset would never be a good source, even when the source becomes available (eg the copyright expires, or it's just released) it's likely to be severely over valued and mostly out of date.
As for the MIT license; it's a completely different fish from the GPL. The MIT license is moderately close in form to the BSD license but even with that there are significant differences. But you're right about one thing; the GPL is very well written. So well written in fact that most respectable lawyers will tell you to start running now. Begging for forgiveness all the way.
Um, Copyright law in general does apply without distribution, if you (eg Apple) break the GPL license it's automatically revoked in toto so copyright law then comes into play. Also the GPL's clauses that come into play only for distribution are rather unusual, most copyright licenses (including the Apple ones) come into play straight away and so if they attempt to be a restraint on distribution they immediately revoke the GPL license; at least that how it's intended.
BTW: The GPL does have clauses that come into play immediately, eg: everything to do with granting you the right to use the software. But as they're grants or very standard clauses they're rarely contested. It's just the right to distribution that get contested.
I'm pretty sure that taken alone Apple are in breach of the GPL (The DRM and related license restraints) but there's lots of wriggle room if they're legally acting as an Agent.
I suppose Apple have tried to set themselves up as an agent so that they aren't legally doing the distribution, they say they are just providing a distribution service in the same sort of way that DHL or UPS or a shop chain does.
I suspect that would fly very well with a judge; especially if they pull things from the store promptly.
The only issue I'd see from Apple's point of view is that they could still be held partially responsible if it could be shown that they should know that the license that the software is being distributed under (GPLv2) is in direct conflict with their agreement with the legal distributor or the customers. That could possibly be seen as a conspiracy to commit copyright infringement.
With "GNU Go" I understand there was such a 'direct conflict' because their agreement tried to add terms to the GPL. But they've changed things since then.
All the provisions of the GPL apply to distributing exact and modified copies of the software.
If you make changes to GPL software but do not distribute it there is nothing that says you must provide your changes to anyone. ONLY if you distribute the software must you make provision for distributing the source of the version you distribute.
I'm not sure where you've got the idea of a 'ridiculous amount of proxying' from. The best solution to having a host connecting to both IPv4 and IPv6 hosts is to use dual stacks; that way you don't get any fragile hacks with routers inspecting and modifying the contents of the connection that your idea (which after all is a form of NAT) would involve. BUT it's quite possible to use NAT to connect between domains that have to be either pure IPv4 or IPv6. It's even got a special name: IPv6 NAT-PT. It lets a pure IPv6 host connect to any IPv4 host through a 'NAT' gateway and allows a pure IPv4 host to connect to selected IPv6 hosts through a 'reverse' NAT.
Because an "interactive protocol" would not work. The core idea of IP is to allow global communications using only local decisions, for any machine connected to the internet you can route any packet with just a small list of "routes" which tell you which port to send the packet and the values in the packet header itself.
Adding some sort of negotiation phase would mean that this information would have to be saved, you have to record the fact that a successful "connection" had been made and what sort of connection it was and probably broadcast the fact that you've made a successful connection to anyone who's interested. But IP doesn't have connections... so it wouldn't be IP any more. You'd lose the massive advantages of IP, the local decisions and the tremendous speed that this allows.
This has been tried before, there are lots of connection oriented protocols around, they are very good for working out how much to charge people and once a connection is established you can easily do things like guaranteed bandwidth allocations. But it takes a long time to setup a connection, every node along the route has to agree to and record the connection (and possibly even talk to an authentication database to decide if the connection is allowed) which takes time, memory and other resources.
In fact it gets so complicated and expensive that most systems end up dumbing down the nodes in the middle so they're fast and light, so they just send packets without worrying where they come from, just pointing them to the next hop on their journey... IP gets reinvented and eventually the inner nodes actually run IP because the hardware for a given guaranteed bandwidth is cheaper; even if there's a need for "over provision".
Quite simply, many people skip all that hassle nowadays and just use IP from the start.
No quite true, perhaps for a non-rotating black hole but any 'normal' black hole is going to be rotating very very fast. This causes very strange effects in the immediate neighbourhood because of the serious space-time distortions. In addition matter near the black hole is under heavy stresses which can cause effects that resemble gamma ray lasers, gravitational shock waves and all sorts of other fun things.
Such things will mean that no, it's not a safe nor a stable orbit, but which direction the planet goes will be entirely up to chance (or rather chaos).
18,446,744,073,709,551,616 seems like an apparently inexhaustible number of addresses, but who knows what else we'll find to add connectivity to?
That's what the IPv6 people thought, so IPv6 has 340,282,366,920,938,463,463,374,607,431,768,211,456 addresses, to get to that limit before IP timeouts you need faster than light communications.
NAT in IPv6 is only of any use when connecting from an IPv6 only host to an IPv4 only host and in that case it's supposed to belong to the IPv4 network.
IPv4 addresses would have run out years ago if it hadn't been for more aggressive allocation policies ( you snooze you lose) and LOTS of NAT devices on the current internet.
There are currently well over 5 billion devices connected to the internet and it's still growing exponentially with all the smartphones being the most recent jump.
It's quite possible to run multiple pppoe connections across one DSL line each pppoe connection could then be assigned a single IP address. It's a bit of a pain to setup on windows though. Better for Windows (or DSL using pppoa) would be to make your default address a NAT address and give you two VPN/tunnel connections to a server on their network. Of course, this is still much more of a pain than just routing IP addresses as they do now. AND they would have to provide a reliable VPN endpoint server.
But, as a matter of fact, you've been assigned NINE IPs, not just five, which can all be used.
First there's the single that's normally assigned to the public IP of your router. If terminate the pppoe link on your grown up router your can use that as the public IP for your NAT machines. Then you just use tunnels, proxy ARP or pptp links to assign all the other eight; including the two that are officially broadcast and the one normally assigned to the inside port of your router.
Again Windows can get a little tetchy with some of the IPs (with some tunnel types) but Unix machines usually have no problems as they don't put pointopoint links into a different class.
But that doesn't mean that some packets aren't more important than others.
My LAN network is designed in the way that LAN networks should be; it's fast enough that it's never the bottleneck, the gigabit switch is very fast and rather expensive.
But the WAN links cannot be provisioned to that level, maybe someday for wired networks, but NEVER for wireless because it's a broadcast medium. So every packet heading for my ADSL line has a TOS/DiffServ value branded on it. Each known TOS value gets put into one of several priority queues (the rest are treated as bulk) each queue has a peak allowed bandwidth. ie. If something tries to flood the 'max-override' queue they get 'max override' packets but only a few percent of the available bandwidth. The result is that even when the link is being flooded with email connections (or other PtP style connections) the "remote desktop" connections still work perfectly.
The important thing here is control, the 'user' has the choice of which queue any packet goes into, the network delivers the packet depending on that choice, not port numbers nor IP addresses nor any other sort of 'inspection'. If the user sends everything out as 'max override' that's what they get, 15kbit/s of top priority data. OTOH the worst case of 'bulk' data might be just 10kbit/s but normally it'll be several Mb/s just like now. (Of course it should be round-robin'd between "users" not "connections" like it is now.)
The 3G networks are designed with just this sort of traffic sorting, the designers knew what they were doing... some of the operators, not so much. (at least that's what they would have us believe!)
I would make the same assumption about two year old pictures as the AC.
This despicable law is just a crutch for lazy investigators.
Quite simply the guy was not at his computer 24 hours a day and it would have been easy to install a key logger that records the first 200 keys after boot. Such a 'wiretapping' process is currently well covered by existing laws.
Point of order; you can't touch any part of a UK plug that becomes live until it's disconnected from the mains. Pulling it half way out does NOT expose anything you can feed power into or get a shock from.
You'd have to do something much more dangerous, like cutting open the cable, or ripping the socket from the wall, to be able to connect your inverter.
Slashdot Mirror
Using four unsigned longs is horrible please read out this address: 536937584.520688358.33825791.4270907832
The hex quads can at least be read out without error.
Quite frankly a 128 bit address is too large for almost anyone to remember you have to use names. You'll probably get a /64 that means that the first half of the address isn't controlled by you eg: 2001:470:1f09:12e6::/64. you can then assign any you like for the other half eg: 2001:470:1f09:12e6::7 even making everything as nice as I can your version would be 536937584.520688358.0.7, horrific.
The double colon is a neat feature eg: 2a00:1450:8006::68 but multiple sets are ambiguous, you'd have to start counting colons, and that's a shitty job.
Anyway, where have you been the last 15 years?
Except it's not a jungle, it's more like the space between galaxies.
You throw a packet into the IPv4 address space and you'll probably hit a machine but it's a different scale in IPv6 space.
I have a little solar system, a few tiny points of light in a single /64. There's a (very) simple firewall across the whole /64 but I've yet to see a single packet that's even a slightly possible candidate for an unsolicited connection of any sort. The few packets the firewall has blocked are traceable to leftovers of old connections.
I'm not using Teredo, I've signed up of a 'Hurricane Electric' 6to4 tunnel (no charge) so I've got static IPv6 addresses but even just the Teredo section of the address space is vast and scanning it is basically impossible.
That's the ideological argument.
The pragmatic argument is a combination
But the same mentality means that if a problem in the BLOB doesn't make Windows crash too much then it's just ignored and will never be fixed. This is ALWAYS seen as a buggy Linux driver.
Hee hee
Sorry, misapplication.
If we were talking satellite data you'd be right, a few gigabits per second shared between the 20 million people in the footprint of one satellite is never going to be enough, and going much higher would be changing the laws of physics. (Shannon's law IIRC)
But enough 8kbit/s voice channels could fit.
The problem is money.
That's the point; by a 'wavebox' I mean a machine that runs the wave locally so 90% (asstat) of the traffic doesn't even touch Google's servers. Google would be used mainly as the trusted third party to startup a wave.
That way when you lose access to the Google account it actually can be a few days before it gets to be a problem.
Though, that was probably the bit I wasn't thinking through. Expecting Google set up a communication link that they can't monitor ... what was I thinking, I'm not usually that dumb!
Sure the internet is reliable in your basement and there's loads of open wifi's in the student block next door.
My DSL is pretty reliable too; no basements or students round here though so Wifi is in short supply. You could notice it too if you went further than the local Starbucks; not everywhere has internet connection in fact some of the best places to visit you don't get any internet connection for days.
Personally I thought the first GoogleOS netbook was going to be a Wavebox for just this reason.
But joined up thinking was never a corporate strong point.
Linux and BSD the OS's under Android and iPhone both have solid security tools. Linux's version was written by the NSA FFS. But once the machine leaves the hands of G&A the 'integrators' have full control over what goes in and what stays out. For Linux the major security enhancements can be turned off with a single switch & kernel recompile.
The only way either company could force the issue is to use legal means and renegotiate their agreements with these 'integrators' AKA 'Phone companies'.
Good luck with that!
We need everyone to verify their age via CC or SS to browse the web
The ISP ALREADY DOES THIS!
They even bill you every month so you won't forget you took responsibility for the connection.
What can they do if you abuse this trust and let someone else browse the web?
As for trusting some random website owner can read you mind and work out what you would be offended by and give a shit once they've done this ... are you a complete idiot, a moron or just a dolt!?
Your message better be a sarcasm overload, because if you can't learn you have a severe retardation problem.
I've done a quick scan and the autogroups thing creates a new scheduling group whenever a user process sets or clears the current CTTY value for the process. This should give a process grouping almost identical to the session id (setsid(3)). You can see your current session ids with ps(1) in the SID column.
I can't tell for sure without trying it but I think this will give a new process group for each application started from many GUI menu programs too because they actively detach you from the CTTY. Even if it doesn't there are many user programs that change the CTTY, the simplest is probably script(1).
But, this is definitely one of those changes that demands the question of why it wasn't though of before; why didn't I think of it, I know what session IDs are and I know about process scheduling groups. It's obvious, in hindsight, that they are a pretty good match for each other.
Gosu is an unrefined mix of cobalt oxide, sodium and other minerals mined in China.
Yup, so very right.
I dunno who you mean by "The GPL", might be the FSF but they're a charity and a stand in copyright holder.
There are quite a few more sites than the one you mention that release hardware designs (GPL and other free licenses) but they run into the problem that there are significant costs to duplication and distribution of hardware. So the closest you're going to get to your silly example is initially a single sponsor leading to a large group of people and companies. Say perhaps Google, they could sponsor an open phone design, they'd need a good name like "the robot phone" or perhaps something a little more human. They would make a lot of cold hard cash out of free software.
The universities were a good source, but now the better sources of paid for free software development are companies, like the afore mentioned Google, or IBM or other similar (mostly) service oriented companies. Companies that know their assets are their people and the skills they have. Companies that think of their software as an asset would never be a good source, even when the source becomes available (eg the copyright expires, or it's just released) it's likely to be severely over valued and mostly out of date.
As for the MIT license; it's a completely different fish from the GPL. The MIT license is moderately close in form to the BSD license but even with that there are significant differences. But you're right about one thing; the GPL is very well written. So well written in fact that most respectable lawyers will tell you to start running now. Begging for forgiveness all the way.
Um, Copyright law in general does apply without distribution, if you (eg Apple) break the GPL license it's automatically revoked in toto so copyright law then comes into play. Also the GPL's clauses that come into play only for distribution are rather unusual, most copyright licenses (including the Apple ones) come into play straight away and so if they attempt to be a restraint on distribution they immediately revoke the GPL license; at least that how it's intended.
BTW: The GPL does have clauses that come into play immediately, eg: everything to do with granting you the right to use the software. But as they're grants or very standard clauses they're rarely contested. It's just the right to distribution that get contested.
I'm pretty sure that taken alone Apple are in breach of the GPL (The DRM and related license restraints) but there's lots of wriggle room if they're legally acting as an Agent.
I suppose Apple have tried to set themselves up as an agent so that they aren't legally doing the distribution, they say they are just providing a distribution service in the same sort of way that DHL or UPS or a shop chain does.
I suspect that would fly very well with a judge; especially if they pull things from the store promptly.
The only issue I'd see from Apple's point of view is that they could still be held partially responsible if it could be shown that they should know that the license that the software is being distributed under (GPLv2) is in direct conflict with their agreement with the legal distributor or the customers. That could possibly be seen as a conspiracy to commit copyright infringement.
With "GNU Go" I understand there was such a 'direct conflict' because their agreement tried to add terms to the GPL. But they've changed things since then.
NO.
All the provisions of the GPL apply to distributing exact and modified copies of the software.
If you make changes to GPL software but do not distribute it there is nothing that says you must provide your changes to anyone. ONLY if you distribute the software must you make provision for distributing the source of the version you distribute.
I'm not sure where you've got the idea of a 'ridiculous amount of proxying' from. The best solution to having a host connecting to both IPv4 and IPv6 hosts is to use dual stacks; that way you don't get any fragile hacks with routers inspecting and modifying the contents of the connection that your idea (which after all is a form of NAT) would involve. BUT it's quite possible to use NAT to connect between domains that have to be either pure IPv4 or IPv6. It's even got a special name: IPv6 NAT-PT. It lets a pure IPv6 host connect to any IPv4 host through a 'NAT' gateway and allows a pure IPv4 host to connect to selected IPv6 hosts through a 'reverse' NAT.
It works as well as NAT ever does.
Well, if it sucks, maybe you should implement a vacuum cleaner with it.
Tried that, it chokes.
Because an "interactive protocol" would not work. The core idea of IP is to allow global communications using only local decisions, for any machine connected to the internet you can route any packet with just a small list of "routes" which tell you which port to send the packet and the values in the packet header itself.
Adding some sort of negotiation phase would mean that this information would have to be saved, you have to record the fact that a successful "connection" had been made and what sort of connection it was and probably broadcast the fact that you've made a successful connection to anyone who's interested. But IP doesn't have connections ... so it wouldn't be IP any more. You'd lose the massive advantages of IP, the local decisions and the tremendous speed that this allows.
This has been tried before, there are lots of connection oriented protocols around, they are very good for working out how much to charge people and once a connection is established you can easily do things like guaranteed bandwidth allocations. But it takes a long time to setup a connection, every node along the route has to agree to and record the connection (and possibly even talk to an authentication database to decide if the connection is allowed) which takes time, memory and other resources.
In fact it gets so complicated and expensive that most systems end up dumbing down the nodes in the middle so they're fast and light, so they just send packets without worrying where they come from, just pointing them to the next hop on their journey ... IP gets reinvented and eventually the inner nodes actually run IP because the hardware for a given guaranteed bandwidth is cheaper; even if there's a need for "over provision".
Quite simply, many people skip all that hassle nowadays and just use IP from the start.
Such things will mean that no, it's not a safe nor a stable orbit, but which direction the planet goes will be entirely up to chance (or rather chaos).
18,446,744,073,709,551,616 seems like an apparently inexhaustible number of addresses, but who knows what else we'll find to add connectivity to?
That's what the IPv6 people thought, so IPv6 has 340,282,366,920,938,463,463,374,607,431,768,211,456 addresses, to get to that limit before IP timeouts you need faster than light communications.
NAT in IPv6 is only of any use when connecting from an IPv6 only host to an IPv4 only host and in that case it's supposed to belong to the IPv4 network.
IPv4 addresses would have run out years ago if it hadn't been for more aggressive allocation policies ( you snooze you lose) and LOTS of NAT devices on the current internet.
There are currently well over 5 billion devices connected to the internet and it's still growing exponentially with all the smartphones being the most recent jump.
Bin there, done that, or haven't you noticed that nearly all ISPs offer /29 assignments (normally 5 available addresses) for small business customers.
Ask politely and there's a very good chance that they'll give you larger assignments too; for a fee.
It's quite possible to run multiple pppoe connections across one DSL line each pppoe connection could then be assigned a single IP address. It's a bit of a pain to setup on windows though. Better for Windows (or DSL using pppoa) would be to make your default address a NAT address and give you two VPN/tunnel connections to a server on their network. Of course, this is still much more of a pain than just routing IP addresses as they do now. AND they would have to provide a reliable VPN endpoint server.
But, as a matter of fact, you've been assigned NINE IPs, not just five, which can all be used.
First there's the single that's normally assigned to the public IP of your router. If terminate the pppoe link on your grown up router your can use that as the public IP for your NAT machines. Then you just use tunnels, proxy ARP or pptp links to assign all the other eight; including the two that are officially broadcast and the one normally assigned to the inside port of your router.
Again Windows can get a little tetchy with some of the IPs (with some tunnel types) but Unix machines usually have no problems as they don't put pointopoint links into a different class.
But that doesn't mean that some packets aren't more important than others.
My LAN network is designed in the way that LAN networks should be; it's fast enough that it's never the bottleneck, the gigabit switch is very fast and rather expensive.
But the WAN links cannot be provisioned to that level, maybe someday for wired networks, but NEVER for wireless because it's a broadcast medium. So every packet heading for my ADSL line has a TOS/DiffServ value branded on it. Each known TOS value gets put into one of several priority queues (the rest are treated as bulk) each queue has a peak allowed bandwidth. ie. If something tries to flood the 'max-override' queue they get 'max override' packets but only a few percent of the available bandwidth. The result is that even when the link is being flooded with email connections (or other PtP style connections) the "remote desktop" connections still work perfectly.
The important thing here is control, the 'user' has the choice of which queue any packet goes into, the network delivers the packet depending on that choice, not port numbers nor IP addresses nor any other sort of 'inspection'. If the user sends everything out as 'max override' that's what they get, 15kbit/s of top priority data. OTOH the worst case of 'bulk' data might be just 10kbit/s but normally it'll be several Mb/s just like now. (Of course it should be round-robin'd between "users" not "connections" like it is now.)
The 3G networks are designed with just this sort of traffic sorting, the designers knew what they were doing ... some of the operators, not so much. (at least that's what they would have us believe!)
I would make the same assumption about two year old pictures as the AC.
This despicable law is just a crutch for lazy investigators.
Quite simply the guy was not at his computer 24 hours a day and it would have been easy to install a key logger that records the first 200 keys after boot. Such a 'wiretapping' process is currently well covered by existing laws.
Point of order; you can't touch any part of a UK plug that becomes live until it's disconnected from the mains. Pulling it half way out does NOT expose anything you can feed power into or get a shock from.
You'd have to do something much more dangerous, like cutting open the cable, or ripping the socket from the wall, to be able to connect your inverter.