Australian here. Many of us are currently connected to the internet using pieces of string which stop working when it rains and they get wet. That's why we're less than delighted that the new NBN plan calls for running fiber to the end of the street and using the existing pieces of string from there.

They did, sort of.

That plan had a cheaper monthly charge than the wholesale price, but had a 100 megabyte download quota and huge excess usage fees. Telstra's intent was to abuse clueless users.

Failing if you cut it with a knife isn't unique to fiber optic cables.

Copper corrodes. That means it fails if it's left lying in the ground completely undisturbed. That sort of unreliability is hard to beat.

The Coalition plan assumes that the copper is in good condition and won't require significant repairs, upgrades or ongoing maintenance. Now that the Coalition has to actually implement their plan rather than just talk about it, they have to find out whether or not those things are true. There's a lot of anecdotal evidence from people who work on the copper network to suggest that they aren't.

Also, the $29 billion number doesn't include the cost of going back and replacing the nodes with FTTP in 10 to 15 years time.

Why not use the copper you've got? The short answer is, a lot of it is shit.

That copper has been lying in the ground slowly corroding for some time. Telstra doesn't really bother with maintenance unless customers complain. Customers just get changed over to a spare pair of wires in the conduit when the pair they were on stops working properly. But conduits are running out of spare pairs.

That's assuming you're on copper in the first place. There's aluminium and lead in the ground in some places. They were cheaper and worked fine for telephone service.

We're now in a situation where, in order to do FTTN, we're going to have to dig up and re-run a lot of copper conduits. If you're going to all that expense, there's little reason to put new copper in the ground instead of doing FTTH.

This. A friend was telling me about how superior his early, hard disk based MP3 player was compared to using modern Android or iOS devices. He could plug surround speakers into it, it had an equaliser, etc, etc, etc. So I pulled out my phone and asked him to show his MP3 player to me. He didn't have it with him. Convenience is a completely legitimate advantage. Yeah, a phone camera will always suck compared to a camera that's physically larger, but a camera that you have with you will always take better pictures than a camera that's in a drawer at home.

I'm not so sure, actually. I think a big part of today's relatively low prices is Intel competing with itself.

We're well and truly into the age of 'good enough' computing. You don't need a new computer to run the latest Windows or Office version or even most games. Unless you're transcoding hours of video or playing today's games on high detail mode that four year old Core 2 Duo is fine, and will be for a while yet.

If Intel raise their prices, the risk isn't that their customers will flock to AMD. The risk is that their customers will say, "Screw it, my old computer still works fine." Intel have to put out faster chips each year at around the same prices to convince people to actually buy and not just sit on what they already have.

Nothing like opening some shitty little flash game and hearing your i7's fan spin up to full speed.

Yes, actually! A month or two ago, AMD released VDPAU support for their open source driver.

Bizarrely, the closed source driver is still XvBA only.

I did actually get VA-API / XvBA working on my AMD system, but it could only do h264 and MPEG2. You could forget xvid, forget advanced GPU deinterlacing, etc. Since it was a weak E-350 box, that left it able to play the highest bitrate bluray rips, but not broadcast TV (MPEG2, 1080i). Replaced it with an Intel Atom / Nvidia ION2 box.

This isn't actually anything to do with the devices in question, this is shitty US mobile networks squeezing you for money because they can. Can't you keep paying for a voice-only plan, turn off data on the smartphone and swap the SIM card over? Or are you talking about signing up for a 24 month plan and getting a 'free' smartphone?

Settings --> Wireless & Networks --> VPN. It's right there on my Galaxy Nexus running vanilla Android.

The vast majority of software that exists won't max out an AMD E-350 netbook chip. Put things in perspective here: we're talking about the minority of software that will actually tax a system.

A lot of programs are single threaded or do almost all of their work in a single thread, and don't really benefit from more cores. Other programs scale almost linearly with number of cores. I was only making the point that software that takes advantage of many cores isn't as rare as the great grandparent seems to think. AMD's multi-threading advantage with its 8 core chips isn't just something that AMD fanboys babble about, there's real benefits in real software that people actually use.

AMD don't have anything that can compete with Intel's top end on single thread performance, but for the mid-range and lower end parts of the desktop market their products are quite competitive. Building a basic computer for a relative, or an office PC that's never going to do anything more intensive than run Word and play Youtube videos? AMD's APUs offer quite a lot of power for not a lot of money. Not everyone has a use for an i7-3970X.

You've never run a video transcode or compiled anything, have you?

I transcode Fraps recordings and upload them to Youtube, transcode bluray video for my Nexus 7, my MythTV backend often has transcode and commflag jobs queued that could run in parallel with no performance loss if it had more cores. 7-Zip will happily multithread compression tasks across dozens of cores. None of that is particularly exotic.

When you say "real world shit" you're talking about games, right? Be aware that there are things other than World of Warcraft that will tax a CPU, and they aren't imaginary or hypothetical.

I think you meant Nub or TrackPoint-style Pointer.