That used to be the case, but it also depends *where* in Asia.

From HK or Tokyo to Western Europe, sure, because of lit capacity and paths of least-resistance, but since about 2010 or 2011, there has been a lot more capacity installed and lit up between Mumbai and Europe (England/France/Italy via the Middle East, depending on the cable system) than there is between Chennai and the US via Singapore/HK/Tokyo.

Russia has fairly extensive fiber, but as I understand it, not much diversity. I would hypothesize that it's used mostly by domestic ISPs and AFAIK quite expensive to use because *I believe* they charge in bit-kilometres in a similar way to most sub-sea cables** AND because of the risks (perceived or real) of routing through Russia**. It's also possible that their domestic network might not be suitable for carrying traffic of International operators because the figures I've as far as operational speeds go seem to be a bit slow.**

**These are all theories on my part, don't take them as accurate unless, by chance, they are.

It would, unfortunately, need to be a significant number of people. Airtel has something like 15 crore (150mm) subscribers, so probably 1 crore (10mm) need to port. Being as big as they are, they wouldn't even notice a change of 1 lakh (100k) subscribers - to them, that's just about standard quarterly churn.

Back in 2008-2009 they were the darling of tech-oriented people with the best plans, prices and quality of service (especially for wired options). Then they started pissing off that crowd and some of them have found sanctuaries on providers you really wouldn't expect (based on the people I know personally, Reliance and BSNL seem to benefited most overall, but Idea and Vodafone seem to gain a quite few too).

India is even more complicated than the US.

There are "monopolies" by local cable providers, but these are miniscule, hyper-local operators with 1,000-10,000 customers operating in very small areas rather than behemoths like Comcast. Some of them aren't even legal entities but will still throw their weight around if given half a chance.

Most of the time, you do have a choice of which ISP you'll use via your LCO's infrastructure (strictly speaking, they are open networks, but they won't always admit it because they might be trying to push one particular ISP that week). Some LCOs have been known to illegally switch subscribers from their ISP of choice to a competitor (illegal not just because it would equate to a bait & switch, but illegal for the ISP because the ISP is required to have documentation on every single subscriber, and the DoT has direct access to networks so in theory, they can check) - not that anybody really seems to give many shits about contract law in India (or, as it turns out, in the US, but I digress). Anyway, they seem to do it because each ISP has different revenue percentages (kickbacks) to the LCO (I've seen rates in the vicinity of 80%) and if they can get away with pocketing an extra 10 rupees per subscriber and the subscriber doesn't notice, they will.

Of course, not every LCO offers every ISP. Sometimes because the ISP already has connectivity with a competing LCO and they don't like each other, or because the LCO doesn't like someone who works at the ISP, or they just never tried because there are hundreds of LCOs around so it's fairly difficult for any ISP to partner with all of them.

In other areas where LCOs don't exist (particularly in certain building projects), ISP choices are artificially limited because the building company doesn't build an open-access network which any ISP can use and/or they are loosely affiliated with an ISP because up the chain they're part of the same conglomerate so only that ISP is allowed in, and not even BSNL/MTNL can build (despite the laws requiring the public companies pretty much have to build and have nearly automatic right of way).

In still other areas, wireless is the only option because nobody (not even the PUCs) even bothers to run cables because they don't want to (like the US in that respect), or equally likely, they can't run cables due to lack of permission and stonewalling (by officials or VIPs) whenever permission is sought (whether the proposed cables are underground or overhead) - or because, weirdly, ROW charges are often a lot higher for telecoms than for power companies.

If the NOFN gets off the ground and is built properly, it'll be a big help in solving some of these problems.

Ah, so it seems we are, in fact, in partial agreement. Unfortunately the Cable-cos choose not to ignore 4G/Wireless when it comes to choices.

That seems to be a rarity, but that does explain why some of my wholesale pricing comes through as BH/TWC, and proves what I've said in various diatribes elsewhere on the web that even on DOCSIS, infrastructure sharing is possible, despite the cable cos (probably including Time Warner but I haven't checked) saying that sharing is not.

Verizon's FIOS growth has, as far as I understand, stalled. Some of it has even been sold off to (bleargh) Frontier.

Google is stalled in some areas, and despite it now being "available" in subsets of 3 cities and all the expansion plans, it's all very nice and well for the populus that lives within reach, but that doesn't include about 99% of the country. I admire what they're doing very much and want them to expand fast, but they'll need a lot of help. I personally would like to see them enter as a retailer on some of the already-existing municipal fiber that's already up and running all over the country. They don't *have* to build or buy all the infrastructure, do they? Just let the municipalities sell them L2 or L3 access and be done. THAT will scare the pants off the incumbents.

AT&T seems to be doing a lot of FTT-PR (and then back-and-forthing with "we will", "we won't", "actually we will").

The latter 2 do at least seem to be growing (excellent), however, while Google's growth rates might look impressive (let's say for the sake of argument 50% year on year or whatever they might be) compared to AT&T's (say 2%), AT&T's *ACTUAL* growth rates would be far more impressive: as we know, 50% of 10k is far less than 2% of 10mm.

Absolutely not, but I do agree with the statement for the most part: choice between DSL or Cable is not really "choice", it is simply the lesser of 2 evils - and it certainly does not resemble anything that could be called "a free market".

HOWEVER, if it were such that you had 2 wires in to your house (one DSL and one Cable, or even 3 wires if you were lucky enough to have fiber as well) and 10 different retailers supplying services on each set of infrastructure, THAT, is what I would call genuine choice, and this is what America needs IMHO.

Isn't the penalty higher when the infringement is commercial?

"Some OEM" is a funny way to put it when you're talking about Foxconn :)

Based on your username, I'm going to make an educated guess that you're Finnish and resent the sale of what was once Finland's flagship company to Microsoft, which is fair enough. But in some ways though, maybe Nokia's influence on the Finnish economy was (until Elop the destroyer came along) a bit *too* large - I lived in Finland for a while and I'm fairly proud of the fact that I was the only non-Finn I knew of that never held a job that involved working (directly or indirectly) for Nokia.

That's not how "shared infrastructure" (in terms of sharing the same infrastructure with multiple ISPs) works though.

In many countries, most or all ISPs offer their services through the same set of cables as their competitors. The ISP merely has some equipment at a central location (street cabinet, exchange or CO) which sends your data off to it's own network for distribution out to the rest of the Internet.

You're right as far as the choice between DSL/Cable goes, even then only sort-of. The rest is nonsense.

4G=Expensive and restrictive (data caps etc). Not practical for every day use unless you literally are only checking emails and reading the paper online (with adblock, of course, because those ads can be heavy).

Satellite=Expensive and horrible latency. Plus not viable for many because HOAs and apartment complexes won't let you put up a dish. Realistically, Satellite is a last resort I'd only suggest in a really rural area with no other choice.

Wireless (Wimax/LR WiFi etc)=Similar deal to satellite and not something I'd recommend unless there was really no other choice.

Fiber=Fiber running past or near your house does not necessarily mean available to you for use. Last stats I saw were not a very high percentage of homes had fiber available and in some cases, that was in place of DSL (so still only leaving a maximum of 2 viable ISP choices), so while I'd love to see more proliferation, that's not currently the case.

So, repeat after me: 2 ISPs to choose from (incumbent DSL or fiber and incumbent cable) is not competition.... 10 or 20 or 50 ISPs utilizing any one of those technologies (as it happens in some countries) is competition.

You'll find that smaller ISPs (especially municipal and/or little fiber ISPs like this one) tend not to have data caps, and usually have big honking backhaul connections and peering just as the incumbants do - sometimes with as much bandwidth as the incumbants do (or at least MUCH more per subscriber).

Well yeah, in a case like that, I kind of would expect better service depending on the circumstances.

The same doesn't got for every product/service, of course, but if the product/service I'm buying specifically offers me a higher level of service than the average joe, I should damn well get it.

For example, I would be a bit miffed if an AAA classic member got priority (or better benefits) than me as an AAA premier tier - after all, I paid for premier for a reason, damnit. Similarly, if I talk business class or dedicated Internet over residential, that probably comes with some form of elevated level of customer service, so why shouldn't I get priority if that's what I'm paying for?

As a Kiwi myself, I feel pretty good about our telecommunications infrastructure when I compare it to the US, and I've been trying to explain the concepts of what's been going on in NZ to Americans since I got here. It's rare to find people here who understand that it is in fact a good thing (AT&T, Verizon, Comcast etc would actually have to compete for your business!!) and the counter-arguments are always about profits (HOW WOULD $COMPANY MAKE ANY MONEY!?) or "incentive to upgrade" -- whatever that's supposed to mean. Then I get pointed at things like Akamai's quarterly SOTI report or rankings on Ookla with silly comparisons like "at least we're better than Mongolia" or something.

They just don't get that maintaining/upgrading the infrastructure is the sole job of the infrastructure company, and that it would be prudent for that infrastructure company to keep up with the technology. They just don't get that since NZers now have 30 or 40 retail ISPs to choose from, irrespective of their address, it's difficult for ISPs to abuse their customers the same way customers get abused in the US, and we can get prices similar to or sometimes cheaper than services in the US and far far superior quality of service without all the gimmicks and special offers of only $20 per month (which then jumps to like $80 per month after some period of time) and not to mention the difficulty of actually finding the damn price on the ISP's website in the first place (which will be advertised exclusive of taxes and local fees, of course).

That being said, there is a VISP market in the US, but it's a bitch: the difference in price between wholesale and retail is often miniscule or even negative (eg wholesale account costs VISP $50/month while the customer can get it direct for $45 or something stupid like that), so I'm not completely certain how well it works.

You're not talking about something like WPA2-Enterprise, are you? I seem to recall that requiring some certificates last time I set one of those up, and that worked OK - while they don't HAVE to be signed by a big CA (you can use a self-signed cert), it works better when they are otherwise you can get authentication problems and such.

Having done this sort of thing for hotels and apartment complexes (and having once briefly worked on the front-lines), the problem with something like WPA2-Enterprise boils down to the people on the front-lines: generating the usernames and passwords on the RADIUS server, which, while not *technically* demanding, might not be something you want to add to the front-desk staff's to-do list -- unless APIs are written for their existing software and it's generated on-the-fly when the guest checks in, of course. The obvious benefit (other than security) being that their username & password could be made to be valid for the duration of their stay.

On the other hand, solutions like what is available at Changi airport (and probably lots of other places, I would imagine, it just sprang to mind first) whereby you just walk up to the desk and receive a one-time code to enter in to the portal, that's a 5 minute job if you include cutting the individual codes from the sheet you'd print them on. Sadly, little security on this part, but if I were paranoid about what I was doing on such a hotspot, that's where a VPN would come in.

Sadly, the line between ease of use by the staff versus ease of use by the customer (and the security implications of each solution) becomes a bit blurry and there isn't much middle ground.

They already exist. 3.5GHz and 900MHz IIRC. And 26GHz or so as well. Many manufacturers have gear available, but if you're looking for off-the-shelf/easy to get ahold of, one place to start would be Ubiquiti.

Problem is, the consumer devices that can't have a USB WNIC plugged in that we all carry don't tend to have radios for those frequencies (having dual-band wifi in most devices is a big enough ask, it would seem), so that might be a bit of a dead-end.

Even *IF* the FCC says "go ahead, jam that wifi", is there any reason we couldn't start moving to 5.8GHz hardware so as to avoid the nonsense they want to pull on the 2.4GHz channels?

There are several options, especially now that 802.11ac has started taking hold:
*Use a USB 3G/4G stick in a small router that supports WAN over USB and has 5.8GHz WiFi (probably the most clunky yet permanent solution because it would have Ethernet ports as well)
*Use a 3G/4G hotspot that offers 5.8GHz wifi.
*Go old-school and connect your phone to your laptop with a good-old USB cable?
*Skip the tethering altogether and use a USB 3G/4G stick plugged directly in to your laptop.

This might not be an option for some clients (older phones/ipads/many cheaper tablets) but they've been putting 5.8GHz radios in the last couple of generations of iphone/ipad/google nexus devices and some laptops, so it's not out of the question.

And while changing to another frequency range might only be a stop-gap solution, but how long until these hotel chains upgrade their hardware? Probably quite a while... and by then, we'll just have to figure out another way of circumventing their BS.

I should clarify that I mean this sort of thing goes on in all sorts of industries when it involves large government projects.

In my view, the question that needs to be asked is less about what is spent, but what is gained or returned. If the rate of return is not good enough, then yes, it's a bad inefficient project and if it's not particularly important, should probably have less allocated to it or be scrapped.