In many places such passing is perfectly legal for cyclists.
I live in the UK. AFAIK it is illegal for *any* vehicle to pass on the left (including cycles). Not that anyone pays attention to this.
And yes, its illegal to cycle at night without lights, but I frequently see people doing it (usually dressed in dark clothes too). They are visible (although not noticable) when they are on their own in a street-lit area. They are invisible when in the glare of oncoming headlights.
Cyclists are not "targets" or "obstacles". They're road users that should have the same rights as bikers, car drivers, etc. Apart from children, they're not really small either. I'd say that if a driver has trouble seeing cyclists, then it's time to take their licence away because they're clearly unfit to drive.
I dunno, the number of cyclists I see cycling down the gap between stationary cars and the footway, at high speed with no lights at night... Hint: if I'm pulling out of a junction and having to look down a line of traffic with their lights all pointing at me, its not going to be entirely easy to see a cyclist illegally undertaking the traffic if they have lights on their bike. If they don't have lights they have no chance of being seen.
Originally, you had "Apple Corps Ltd", a British company. And "Apple Computer, Inc", a US company.
My comment was more specifically related to the GP's proposed.tm.uk domain for trademarks, whereby Apple Records and Apple Computers could both legitimately claim apple.tm.uk since they both hold that trademark.
That's exactly the kind of injury that cycle helmets aren't much use at preventing -- the speeds are too high.
Also, some research showed that drivers overtook helmeted cyclists with less room compared to unhelmeted cyclists, i.e. the drivers take a higher risk because they assume the helmet is protecting the cyclist.
I know a few people who have done an over-the-handlebars manouver at high speed (one because a car pulled out of a junction in front of her and she hit it and went flying over the bonnet). In these cases, the helmets usually end up in pieces rather than the rider's skull.
Personally I've been knocked off by cars clipping my handlebars while passing too close. The result is my bike making a swift turn and throwing me into the centre of the road. Again - the helmet works well here since it prevents your head hitting the road surface.
Sure, if a car hits you head-on at 50mph the helmet's not going to do a lot to help, but there are so many collisions that result in an impact with the road surface that a helmet does help with, I think you'd be nuts not to wear one.
That's the kind of injury the helmet might help with, and people cycling for sport should probably wear helmets. (Just like people driving for sport wear helmets.)
I'd be willing to bet that a commuting cyclist who goes flying over the front of a car that pulled out infront of them and lands head-first on the road surface is going to be in much better shape than a mountain biker who gets thrown head-first into a tree, so I don't see why you think that commuting cyclist collisions are too extreme for helmets to be useful whilst sport accidents aren't.
Having used a Royal Mail PO box in the past, I can say that its not that easy to set up such a thing unless you tell lies. You do need to already have a valid postal address in the post town that the PO box will be
Doesn't the RM offer regionless PO Boxes? I'm pretty sure I've seen addresses published which are just "Company Name, PO BOX 123" without a town.
Also, you can have the street addressed mail automatically diverted to your PO box, so that it is not obviously a PO box, which is also useful since banks at least won't allow your registered address to be a PO box.
Well, in the context of this discussion, if you had a real street address in the UK then you wouldn't need to worry about a PO Box anyway.
Well, but the shops trading as "Village Grocer" aren't limited companies, right?
I would expect them to be. Frankly, anyone running a vaguely sizable business as a sole trader is an idiot - as soon as you start employing people you'd be crazy not to register as a limited company.
Also, as for trade names, they would not fall under the ltd namespace because they are not corporation names.
That was my point, because it makes the.ltd.uk namespace far less useful by virtue of the fact that your domain name is not necessarilly obvious to your customers. Your customers know you by your trading name, not your limited company name - if they want to find you on the web, which do you think they will look for?
For tradenames, there could be a separate tm namespaced ("trademark").
Again, tradenames are only protected from collisions within their specific sector. For example, the tradename "Apple" is owned by both Apple Records and Apple Computers, and this is ok because Apple Records deals with music and not computers, whiles Apple Computers deals with computers and not music... (right? right? ahhem:)
What I'd be interested on knowing: How fast would we have to spin something to approximate 1G, and how big would it have to be? (Several times the height of a human is my guess, in order to prevent having stratified gravity.)
Is 1G even optimum or necessary to retain bone mass and a strong heart?
Smaller diameters of "space station" require a higher angular velocity, but in principle there is no specific size restriction if you simply want to achieve a 1g accelleration at floor level. However, using a small diameter has a couple of problems: 1. The accelleration gradient is more extreme (equivalent to gravitational tidal forces). e.g. for a capsule twice the height of a human, your feet would be at 1g, but your head (being in the centre) would be in 0g. Use a bigger diameter and a slower angular velocity and you will reduce the gradient. 2. The coriolis effects associated with high angular velocities make it extremely unpleasant to move around in a fast spinning (hence small diameter) capsule. According to Wikipedia you need to spin at under 7 RPM, preferably around 2 RPM, to make this manageable. At 2 RPM you need a diameter of about half a kilometer to achieve 1g. 7 RPM is a bit more managable, requiring a diameter of 40 metres. Rather than building a cylindrical capsule, a better option might be to have a pair of capsules tethered together with a 500 metre tether.
If my company is in the EU, but not the UK, I can't get a ".uk" domain name? Doesn't that violate EU rules?
Can you get a UK postal address? If not, doesn't that violate EU rules? (I don't think there is any EU rule that says a French business (for example) has a right to try to convince their British customers that they are based in the UK even when they aren't).
Sure, you can pay a UK business to use their postal address and forward mail to you, and similarly you could pay a UK business to let you use their domain name. You can probably pay the Royal Mail to hold a PO Box for you, but a PO Box address is extremely obviously a PO Box rather than a real address.
How about an Ltd secondary level domain? It would cover limited companies (corporations). Since this namespace is already controlled (you can't have the same name as another corp, AFAIK), you would automatically be allocated "your" domain name. That, or it would be reserved for your purchase.
So, you'd have britishgas.ltd.uk
.ltd.uk already exists, but I've never seen anyone actually use it.
However, you're wrong on this preventing namespace collisions - companies are allowed to have the same name so long as they are in completely different lines of business (so there is no confusion).
Also, the trading names of limited companies are often not the same as the limited company name itself, so this probably doesn't help too much. e.g. there are probably quite a few shops that trade as "Village Grocers" or similar, but they can't all have that as their limited company name. Similarly, a single limited company may own several distinct business units trading under different names, which may either be an intentional attempt to segregate the business in the eyes of the customer (this is often a good thing if those shops specialise in different things - the customer knows which shop to go to for the thing they want without needing to care whether they are run by the same company or not), or may be through aquisition (its common for merged businesses to continue trading under separate names to avoid customer confusion, even though they have merged to become a single limited company).
This would work similarly to how the.com.au works in Australia. I know there's numerous work-arounds, but for the most part you need an ABN (Australian Business Number) registered for the domain name you're after.
I'm not a fan, but it has reduced much of the cyber squatting and other issues (sorry, can't site sources).
This is what the.ltd.uk TLD is supposed to do, and AFAIK you have to prove you're a registered limited company. However, I don't think I've ever seen anyone use this TLD.
Rather than inventing TLDs at random, what is needed is some kind of joined up thinking on how best to structure domain names so that the TLD actually reflects something useful and doesn't just result in each company having to register their domain name under every TLD in existence.
For example, categorising businesses based on TLD would be vaguely sensible - e.g. fred.plumber.glasgow.uk for your friendly local plumber in glasgow... (yes, I know, its not short and catchy, but in a world where there are probably lots of businesses called "Fred's Plumbing", "Fred's Butcher", etc they can't all have fred.com).
When I had AT&T, my friends with free texting pre-paid plans would text me all the time and AT&T would charge 25 cents for each and every one sent or received.
So your choice seems obvious. Either: 1. Change to an MNO with a less fucked up pricing model (WTF? Charging for received SMS messages? I don't even get that when I'm roaming) or 2. Tell AT&T to block all SMS messages and use a different service (e.g. XMPP)
If you want something even more cross-platform, Google Talk, which uses XMPP, also works great and is a full-fledged IM service.
I tend to use my own XMPP server, but don't usually leave the XMPP client running the whole time because it sucks battery keeping the XMPP session open (if anyone finds a good Android XMPP client that doesn't suck battery, I'm all ears). Its a shame it doesn't integrate with the normal messaging app though.
Least they are not going down without a fight after the GPS industry screwed them over. They PAID spectrum to start a business on but interference with GPS devices WHICH clearly is the fault of companies that made the GPS devices screwed them bad.
They PAID for spectrum which was designated to be used for a different purpose, then tried to convince everyone that there was no problem with changing the purpose of that spectrum, even though it caused interference for existing users of neighbouring spectrum.
Trade secrets lead to a closed, uncooperative system where "the wheel" so to speak is constantly reinvented and the pace of techological innovation is significantly slowed.
When was the last time you looked up a patent rather than reinventing the wheel? Certainly whith software I am reinventing wheels on a daily basis, but it is easier and quicker for me to do this than find an appropriate patent and adapt it to my situation.
Most software patents document the obvious. Those things that weren't obvious when they were filed will be considered obvious by the time they are granted. Modern patents are also so badly obfuscated by the patent writers that they probably can't be used as a basis of implementation anyway.
There are some (non-software) patents that cover large portions of a whole product that I think may be beneficial uses of the patent system, but patents that cover only small components within a device are really not beneficial to society because no one is going to spend the time looking for a patent that covers what they want to do, and those who infringe almost always do so by independently inventing something without realising it was already patented.
I support the idea of having to pay an inventor in situations where their invention has saved you from the R&D expense of developing it yourself, but I don't support the notion that you should have to pay them just because you inadvertently invented the same thing as them (and haence already had the R&D expense yourself.)
Android simply isnt designed for hardware keyboard use the way Blackberries were.
It isn't? My old HTC Dream had excellent hard keyboard support, as does my current Samsung Captivate Glide. Now, it is quite hard to find a decent hard keyboarded Android phone (I had to import mine from the US), but the software itself supports it fine.
I once had the misfortune to work for a self-taught coder who had no formal qualifications (don't get me wrong, I'm self-taught to, but I did bother to take the time to do some formal learning too). Whenever I told him how long something would take, he took the attitude that that was too long because he could've coded it quicker himself. The problem was he _could_ have coded it quicker himself, but the resulting heap of spaghetti code would've been buggy as hell and utterly unmaintainable. Trying to justify to hime spending a longer time on development so you don't have to rewrite the whole thing from scratch the next time you need to do any kind of maintenance was hard... especially if I wanted to avoid insulting his coding abilities.
Marketing gets a bad rep, but if you were a tech company would you want your coders or engineers selling your products to the general population?
No, but similarly I'm also concerned about marketers selling my products since my company hinges on customers trusting us, and the trust is likely to go out of the window as soon as your average salesman starts making up any old bullshit in order to sell the product.
Solution? Have these conversations behind closed doors, well away from the street.
My thoughts exactly. Non-engineers really shouldn't be directly involved with purchasing engineering solutions.
The evolving paradigm in computing is that there is no such thing as 'don't have a reliable internet connection'.
Yes, and I'm all for having ubiquitious and cheap internet connectivity wherever you go, but "evolving" != "evolved" - we're nowhere near there yet. Some situations I am frequently in:
- On a train. Yes, I can get a flakey 3G signal, it comes and goes. It is about good enough for occaisonally looking stuff up on the web so long as you're happy to twiddle your thumbs for 10 minutes every so often while the 3G signal completely vanishes. I certainly wouldn't want to be relying on this for doing actual work.
- In the middle of nowhere. As a mountaineer, I often go to pretty much the middle of nowhere for days on end. Whilst I don't usually take work with me on such trips, I do often take a laptop and it is handy to be able to do random stuff on it. For example - if I spend a week on the Isle of Skye, that is a week without even a GSM signal, let alone something I can access the Internet over. In this situation if all my applications were cloud app, my laptop would be as useful as a brick.
- For whatever reason, most of my customers seem to be in locations where 3G signals are quite weak. Go a little way into the building and the signal disappears completely. Having to ask them to set up wifi access for me on their network just so I could use a cloud app (which could easilly have had a just-as-functional non-cloud version) would be a lot of faff. (Yes, I accept that sometimes I need network access from my laptop anyway so I already have to do this, but there are a lot of times where I don't and I'm just using applications and data stored on the laptop itself).
In a decade, maybe we will be closer to "internet everywhere", but we're certainly not there yet. However, one thing I've noticed is that the trend from the wireless providers is largely not in that direction: for example the new EE LTE network is going to be primarilly rolled out only in high population locations (i.e. places where you could *already* see a bunch of wifi hotspots practically everywhere you go anyway), whilst the places that have no 3G or even GSM coverage are still going to be left uncovered. In a decade or two I really wouldn't be surprised to see the current dead spots still being dead spots whilst the currently well covered areas will just end up with more redundent networks offering identical coverage.
I welcome cloud stuff where it makes sense, but I just don't see stuff like doing your word processing on the web as making any sense at all...
If they can allocate consistently in a 51/8 network, they can allocate consistently in a 10/8 network.
Ok, you've got two separate organisations (lets call them A and B). They both use the same RFC1918 network (lets say, 10/8). Now, as part of a collaborative project, organisation A needs to access some resources on organisation B's network, so they set up a VPN. Except they are both sharing addresses, so they can't just route traffic between the existing networks.
The solution here is to allocate non-conflicting addresses to all the resources that need to be shared between sites - So we allocate 192.168.0/24 to A and 192.168.1/24 to B. The important point here is you're not renumbering entire networks, you're just adding an extra address to everything that needs to be shared - renumbering the entire network is a *lot* of work, adding an extra address to a few machines, and approrpiate routing on the routers is easy.
Now, organisation A needs to collaberate with a third organisation, C. But oh shit, organisation C uses 192.168/16 internally. We now end up in a complete mess that can really only be solved by doing a load of renumbering.
The more sensible alternative i that organisation A has their own globally unique IP address space. Lets call it 51/8. So A now allocates 51.0.0/24 to its shared resources, B allocates 51.0.1/24 to its shared resources, and when C comes along they get 51.0.2/24. No one needs to do any renumbering, and you can continue to add partner organisations without needing to worry about coordinating address allocations for their internal networks.
This is also why IPv6 unique local addresses were depricated many years ago, because it was widely recognised as better for each organisation to have globally unique addresses, even if they aren't using them to talk to the internet, because it guards against unforseen future address conflicts caused by mergers, collaberative projects, etc.
Tproxy. We use it all the time on our customers' networks.
Transparent proxies require that the router redirects the relevant packets to them instead of the real server
Correct.
(DNAT target on iptables) and would not work without NAT, unless the DNS requests also got intercepted
Incorrect. No NAT required. The TPROXY netfilter target will intercept matched traffic and send it to a local process instead - no NAT happens here, the packet just gets _routed_ to a local process with the original IP and TCP headers unmodified.
If your proxy is not on the gateway router machine itself, then you have 2 options: 1. Add routing rules to the border router to route the appropriate traffic to the proxy (where it will be intercepted by the TPROXY target). If there are other routers between the proxy and the border router then these need the same rules, which can be a little messy so we only usually use this option if the border router and proxy server are on the same subnet. 2. Do (1), but over a GRE tunnel between the border router and the proxy so that any routers between the two don't need to care about special routing rules.
On the whole, because no connection tracking is involved, you don't need to care about avoiding asymmetric routing so this works out easier to manage than using DNAT to redirect the traffic (for this reason I'd advise against using DNAT to implement transparent proxying even in IPv4-only situations).
The only thing more evil than forced upgrades in the world of OS developers would be subscription-based OS.
No one is *forcing* you to upgrade. But supporting multiple OSes costs money so everyone will drop support for old versions eventually because continuing to support it isn't profitable any more. This applies to *all* OSes, not just Windows.
I don't require MS to support my XP install. I use XP because I like it, and I don't WANT to switch to 7 (don't need any of the features, and it has a much higher overhead on my older hardware).
So don't switch. You're free to continue using Windows XP (under the terms of the licence you originally _agreed to_). But don't expect third parties like Google to bother supporting it for you either.
I am finding less and less shiny new Win7 laptops with serial ports
And that's why god invented FTDI USB devices...
As far as my milage goes: Not had to use IE in over 10 years for anything other than occasional testing. All the hardware we've got is happy with firefox, banking and government web apps are all fine too.
Oh dear god, there are people who still sell this obsolete piece of shit (complete with obsolete H323 phones... seriously, why is anyone using anything other than SIP these days?)... some of my customers recently had them installed. Worst thing is the installers, who claim to be "VoIP experts" pretty much univerally don't understand the first thing about "IP" (they look at you completely stumped when they discover they are working with a routed IP network instead of a site-wide broadcast network).
Slashdot Mirror
In many places such passing is perfectly legal for cyclists.
I live in the UK. AFAIK it is illegal for *any* vehicle to pass on the left (including cycles). Not that anyone pays attention to this.
And yes, its illegal to cycle at night without lights, but I frequently see people doing it (usually dressed in dark clothes too). They are visible (although not noticable) when they are on their own in a street-lit area. They are invisible when in the glare of oncoming headlights.
Cyclists are not "targets" or "obstacles". They're road users that should have the same rights as bikers, car drivers, etc. Apart from children, they're not really small either. I'd say that if a driver has trouble seeing cyclists, then it's time to take their licence away because they're clearly unfit to drive.
I dunno, the number of cyclists I see cycling down the gap between stationary cars and the footway, at high speed with no lights at night... Hint: if I'm pulling out of a junction and having to look down a line of traffic with their lights all pointing at me, its not going to be entirely easy to see a cyclist illegally undertaking the traffic if they have lights on their bike. If they don't have lights they have no chance of being seen.
Right, I was going to bring up Apple.
Originally, you had "Apple Corps Ltd", a British company. And "Apple Computer, Inc", a US company.
My comment was more specifically related to the GP's proposed .tm.uk domain for trademarks, whereby Apple Records and Apple Computers could both legitimately claim apple.tm.uk since they both hold that trademark.
That's exactly the kind of injury that cycle helmets aren't much use at preventing -- the speeds are too high.
Also, some research showed that drivers overtook helmeted cyclists with less room compared to unhelmeted cyclists, i.e. the drivers take a higher risk because they assume the helmet is protecting the cyclist.
I know a few people who have done an over-the-handlebars manouver at high speed (one because a car pulled out of a junction in front of her and she hit it and went flying over the bonnet). In these cases, the helmets usually end up in pieces rather than the rider's skull.
Personally I've been knocked off by cars clipping my handlebars while passing too close. The result is my bike making a swift turn and throwing me into the centre of the road. Again - the helmet works well here since it prevents your head hitting the road surface.
Sure, if a car hits you head-on at 50mph the helmet's not going to do a lot to help, but there are so many collisions that result in an impact with the road surface that a helmet does help with, I think you'd be nuts not to wear one.
That's the kind of injury the helmet might help with, and people cycling for sport should probably wear helmets. (Just like people driving for sport wear helmets.)
I'd be willing to bet that a commuting cyclist who goes flying over the front of a car that pulled out infront of them and lands head-first on the road surface is going to be in much better shape than a mountain biker who gets thrown head-first into a tree, so I don't see why you think that commuting cyclist collisions are too extreme for helmets to be useful whilst sport accidents aren't.
Having used a Royal Mail PO box in the past, I can say that its not that easy to set up such a thing unless you tell lies. You do need to already have a valid postal address in the post town that the PO box will be
Doesn't the RM offer regionless PO Boxes? I'm pretty sure I've seen addresses published which are just "Company Name, PO BOX 123" without a town.
Also, you can have the street addressed mail automatically diverted to your PO box, so that it is not obviously a PO box, which is also useful since banks at least won't allow your registered address to be a PO box.
Well, in the context of this discussion, if you had a real street address in the UK then you wouldn't need to worry about a PO Box anyway.
Well, but the shops trading as "Village Grocer" aren't limited companies, right?
I would expect them to be. Frankly, anyone running a vaguely sizable business as a sole trader is an idiot - as soon as you start employing people you'd be crazy not to register as a limited company.
Also, as for trade names, they would not fall under the ltd namespace because they are not corporation names.
That was my point, because it makes the .ltd.uk namespace far less useful by virtue of the fact that your domain name is not necessarilly obvious to your customers. Your customers know you by your trading name, not your limited company name - if they want to find you on the web, which do you think they will look for?
For tradenames, there could be a separate tm namespaced ("trademark").
Again, tradenames are only protected from collisions within their specific sector. For example, the tradename "Apple" is owned by both Apple Records and Apple Computers, and this is ok because Apple Records deals with music and not computers, whiles Apple Computers deals with computers and not music... (right? right? ahhem :)
What I'd be interested on knowing:
How fast would we have to spin something to approximate 1G, and how big would it have to be? (Several times the height of a human is my guess, in order to prevent having stratified gravity.)
Is 1G even optimum or necessary to retain bone mass and a strong heart?
Smaller diameters of "space station" require a higher angular velocity, but in principle there is no specific size restriction if you simply want to achieve a 1g accelleration at floor level. However, using a small diameter has a couple of problems:
1. The accelleration gradient is more extreme (equivalent to gravitational tidal forces). e.g. for a capsule twice the height of a human, your feet would be at 1g, but your head (being in the centre) would be in 0g. Use a bigger diameter and a slower angular velocity and you will reduce the gradient.
2. The coriolis effects associated with high angular velocities make it extremely unpleasant to move around in a fast spinning (hence small diameter) capsule. According to Wikipedia you need to spin at under 7 RPM, preferably around 2 RPM, to make this manageable. At 2 RPM you need a diameter of about half a kilometer to achieve 1g. 7 RPM is a bit more managable, requiring a diameter of 40 metres. Rather than building a cylindrical capsule, a better option might be to have a pair of capsules tethered together with a 500 metre tether.
If my company is in the EU, but not the UK, I can't get a ".uk" domain name? Doesn't that violate EU rules?
Can you get a UK postal address? If not, doesn't that violate EU rules?
(I don't think there is any EU rule that says a French business (for example) has a right to try to convince their British customers that they are based in the UK even when they aren't).
Sure, you can pay a UK business to use their postal address and forward mail to you, and similarly you could pay a UK business to let you use their domain name.
You can probably pay the Royal Mail to hold a PO Box for you, but a PO Box address is extremely obviously a PO Box rather than a real address.
secondary level domains:
How about an Ltd secondary level domain? It would cover limited companies (corporations). Since this namespace is already controlled (you can't have the same name as another corp, AFAIK), you would automatically be allocated "your" domain name. That, or it would be reserved for your purchase.
So, you'd have britishgas.ltd.uk
.ltd.uk already exists, but I've never seen anyone actually use it.
However, you're wrong on this preventing namespace collisions - companies are allowed to have the same name so long as they are in completely different lines of business (so there is no confusion).
Also, the trading names of limited companies are often not the same as the limited company name itself, so this probably doesn't help too much. e.g. there are probably quite a few shops that trade as "Village Grocers" or similar, but they can't all have that as their limited company name. Similarly, a single limited company may own several distinct business units trading under different names, which may either be an intentional attempt to segregate the business in the eyes of the customer (this is often a good thing if those shops specialise in different things - the customer knows which shop to go to for the thing they want without needing to care whether they are run by the same company or not), or may be through aquisition (its common for merged businesses to continue trading under separate names to avoid customer confusion, even though they have merged to become a single limited company).
This would work similarly to how the .com.au works in Australia. I know there's numerous work-arounds, but for the most part you need an ABN (Australian Business Number) registered for the domain name you're after.
I'm not a fan, but it has reduced much of the cyber squatting and other issues (sorry, can't site sources).
This is what the .ltd.uk TLD is supposed to do, and AFAIK you have to prove you're a registered limited company. However, I don't think I've ever seen anyone use this TLD.
Rather than inventing TLDs at random, what is needed is some kind of joined up thinking on how best to structure domain names so that the TLD actually reflects something useful and doesn't just result in each company having to register their domain name under every TLD in existence.
For example, categorising businesses based on TLD would be vaguely sensible - e.g. fred.plumber.glasgow.uk for your friendly local plumber in glasgow... (yes, I know, its not short and catchy, but in a world where there are probably lots of businesses called "Fred's Plumbing", "Fred's Butcher", etc they can't all have fred.com).
When I had AT&T, my friends with free texting pre-paid plans would text me all the time and AT&T would charge 25 cents for each and every one sent or received.
So your choice seems obvious. Either:
1. Change to an MNO with a less fucked up pricing model (WTF? Charging for received SMS messages? I don't even get that when I'm roaming)
or
2. Tell AT&T to block all SMS messages and use a different service (e.g. XMPP)
If you want something even more cross-platform, Google Talk, which uses XMPP, also works great and is a full-fledged IM service.
I tend to use my own XMPP server, but don't usually leave the XMPP client running the whole time because it sucks battery keeping the XMPP session open (if anyone finds a good Android XMPP client that doesn't suck battery, I'm all ears). Its a shame it doesn't integrate with the normal messaging app though.
I think he's probably saying that voice and text messages are hugely overpriced. I don't know about voice but SMSes certainly are.
Are voice calls and SMS messages *really* overpriced when people continue to use them despite there being cheaper alternatives?
Least they are not going down without a fight after the GPS industry screwed them over. They PAID spectrum to start a business on but interference with GPS devices WHICH clearly is the fault of companies that made the GPS devices screwed them bad.
They PAID for spectrum which was designated to be used for a different purpose, then tried to convince everyone that there was no problem with changing the purpose of that spectrum, even though it caused interference for existing users of neighbouring spectrum.
Trade secrets lead to a closed, uncooperative system where "the wheel" so to speak is constantly reinvented and the pace of techological innovation is significantly slowed.
When was the last time you looked up a patent rather than reinventing the wheel? Certainly whith software I am reinventing wheels on a daily basis, but it is easier and quicker for me to do this than find an appropriate patent and adapt it to my situation.
Most software patents document the obvious. Those things that weren't obvious when they were filed will be considered obvious by the time they are granted. Modern patents are also so badly obfuscated by the patent writers that they probably can't be used as a basis of implementation anyway.
There are some (non-software) patents that cover large portions of a whole product that I think may be beneficial uses of the patent system, but patents that cover only small components within a device are really not beneficial to society because no one is going to spend the time looking for a patent that covers what they want to do, and those who infringe almost always do so by independently inventing something without realising it was already patented.
I support the idea of having to pay an inventor in situations where their invention has saved you from the R&D expense of developing it yourself, but I don't support the notion that you should have to pay them just because you inadvertently invented the same thing as them (and haence already had the R&D expense yourself.)
Android simply isnt designed for hardware keyboard use the way Blackberries were.
It isn't? My old HTC Dream had excellent hard keyboard support, as does my current Samsung Captivate Glide. Now, it is quite hard to find a decent hard keyboarded Android phone (I had to import mine from the US), but the software itself supports it fine.
I once had the misfortune to work for a self-taught coder who had no formal qualifications (don't get me wrong, I'm self-taught to, but I did bother to take the time to do some formal learning too). Whenever I told him how long something would take, he took the attitude that that was too long because he could've coded it quicker himself. The problem was he _could_ have coded it quicker himself, but the resulting heap of spaghetti code would've been buggy as hell and utterly unmaintainable. Trying to justify to hime spending a longer time on development so you don't have to rewrite the whole thing from scratch the next time you need to do any kind of maintenance was hard... especially if I wanted to avoid insulting his coding abilities.
Marketing gets a bad rep, but if you were a tech company would you want your coders or engineers selling your products to the general population?
No, but similarly I'm also concerned about marketers selling my products since my company hinges on customers trusting us, and the trust is likely to go out of the window as soon as your average salesman starts making up any old bullshit in order to sell the product.
Solution? Have these conversations behind closed doors, well away from the street.
My thoughts exactly. Non-engineers really shouldn't be directly involved with purchasing engineering solutions.
The evolving paradigm in computing is that there is no such thing as 'don't have a reliable internet connection'.
Yes, and I'm all for having ubiquitious and cheap internet connectivity wherever you go, but "evolving" != "evolved" - we're nowhere near there yet. Some situations I am frequently in:
- On a train. Yes, I can get a flakey 3G signal, it comes and goes. It is about good enough for occaisonally looking stuff up on the web so long as you're happy to twiddle your thumbs for 10 minutes every so often while the 3G signal completely vanishes. I certainly wouldn't want to be relying on this for doing actual work.
- In the middle of nowhere. As a mountaineer, I often go to pretty much the middle of nowhere for days on end. Whilst I don't usually take work with me on such trips, I do often take a laptop and it is handy to be able to do random stuff on it. For example - if I spend a week on the Isle of Skye, that is a week without even a GSM signal, let alone something I can access the Internet over. In this situation if all my applications were cloud app, my laptop would be as useful as a brick.
- For whatever reason, most of my customers seem to be in locations where 3G signals are quite weak. Go a little way into the building and the signal disappears completely. Having to ask them to set up wifi access for me on their network just so I could use a cloud app (which could easilly have had a just-as-functional non-cloud version) would be a lot of faff. (Yes, I accept that sometimes I need network access from my laptop anyway so I already have to do this, but there are a lot of times where I don't and I'm just using applications and data stored on the laptop itself).
In a decade, maybe we will be closer to "internet everywhere", but we're certainly not there yet. However, one thing I've noticed is that the trend from the wireless providers is largely not in that direction: for example the new EE LTE network is going to be primarilly rolled out only in high population locations (i.e. places where you could *already* see a bunch of wifi hotspots practically everywhere you go anyway), whilst the places that have no 3G or even GSM coverage are still going to be left uncovered. In a decade or two I really wouldn't be surprised to see the current dead spots still being dead spots whilst the currently well covered areas will just end up with more redundent networks offering identical coverage.
I welcome cloud stuff where it makes sense, but I just don't see stuff like doing your word processing on the web as making any sense at all...
That assumes you always upgrade, something many companies do not do every time a new version comes out.
If you're never going to upgrade then this whole conversation is pretty much irrelevant to you anyway.
If they can allocate consistently in a 51/8 network, they can allocate consistently in a 10/8 network.
Ok, you've got two separate organisations (lets call them A and B). They both use the same RFC1918 network (lets say, 10/8). Now, as part of a collaborative project, organisation A needs to access some resources on organisation B's network, so they set up a VPN. Except they are both sharing addresses, so they can't just route traffic between the existing networks.
The solution here is to allocate non-conflicting addresses to all the resources that need to be shared between sites - So we allocate 192.168.0/24 to A and 192.168.1/24 to B. The important point here is you're not renumbering entire networks, you're just adding an extra address to everything that needs to be shared - renumbering the entire network is a *lot* of work, adding an extra address to a few machines, and approrpiate routing on the routers is easy.
Now, organisation A needs to collaberate with a third organisation, C. But oh shit, organisation C uses 192.168/16 internally. We now end up in a complete mess that can really only be solved by doing a load of renumbering.
The more sensible alternative i that organisation A has their own globally unique IP address space. Lets call it 51/8. So A now allocates 51.0.0/24 to its shared resources, B allocates 51.0.1/24 to its shared resources, and when C comes along they get 51.0.2/24. No one needs to do any renumbering, and you can continue to add partner organisations without needing to worry about coordinating address allocations for their internal networks.
This is also why IPv6 unique local addresses were depricated many years ago, because it was widely recognised as better for each organisation to have globally unique addresses, even if they aren't using them to talk to the internet, because it guards against unforseen future address conflicts caused by mergers, collaberative projects, etc.
Transparent proxies work just fine.
How?
Tproxy. We use it all the time on our customers' networks.
Transparent proxies require that the router redirects the relevant packets to them instead of the real server
Correct.
(DNAT target on iptables) and would not work without NAT, unless the DNS requests also got intercepted
Incorrect. No NAT required. The TPROXY netfilter target will intercept matched traffic and send it to a local process instead - no NAT happens here, the packet just gets _routed_ to a local process with the original IP and TCP headers unmodified.
If your proxy is not on the gateway router machine itself, then you have 2 options:
1. Add routing rules to the border router to route the appropriate traffic to the proxy (where it will be intercepted by the TPROXY target). If there are other routers between the proxy and the border router then these need the same rules, which can be a little messy so we only usually use this option if the border router and proxy server are on the same subnet.
2. Do (1), but over a GRE tunnel between the border router and the proxy so that any routers between the two don't need to care about special routing rules.
On the whole, because no connection tracking is involved, you don't need to care about avoiding asymmetric routing so this works out easier to manage than using DNAT to redirect the traffic (for this reason I'd advise against using DNAT to implement transparent proxying even in IPv4-only situations).
The only thing more evil than forced upgrades in the world of OS developers would be subscription-based OS.
No one is *forcing* you to upgrade. But supporting multiple OSes costs money so everyone will drop support for old versions eventually because continuing to support it isn't profitable any more. This applies to *all* OSes, not just Windows.
I don't require MS to support my XP install. I use XP because I like it, and I don't WANT to switch to 7 (don't need any of the features, and it has a much higher overhead on my older hardware).
So don't switch. You're free to continue using Windows XP (under the terms of the licence you originally _agreed to_). But don't expect third parties like Google to bother supporting it for you either.
I am finding less and less shiny new Win7 laptops with serial ports
And that's why god invented FTDI USB devices...
As far as my milage goes: Not had to use IE in over 10 years for anything other than occasional testing. All the hardware we've got is happy with firefox, banking and government web apps are all fine too.
avaya i think it was called
Oh dear god, there are people who still sell this obsolete piece of shit (complete with obsolete H323 phones... seriously, why is anyone using anything other than SIP these days?)... some of my customers recently had them installed. Worst thing is the installers, who claim to be "VoIP experts" pretty much univerally don't understand the first thing about "IP" (they look at you completely stumped when they discover they are working with a routed IP network instead of a site-wide broadcast network).