Death of Decent Australian Broadband

iamplasma writes: "As reported by several Australian newspapers, Optus cable internet services will be switching their standard plan to a 3gb "soft-limit" broadband service (once the limit is passed, the service slows to 28kbps). This is effectively the end of decent broadband in Australia, with Optus being the only major provider to offer a service without a highly restrictive usage cap. This is also the ISP who proudly promoted themselves over their main competitor specifically over the issue of the competitor's 3gb limits."

Comments

Live With it or Build Your Own Network

[Taliesan999]

Having started and run an ISP in the mid nineties, there is a simple equation. In Australia, bandwith costs money and is considerably more expensive than in the US (because of the poor deal Australia gets on bandwith to/from the US).

Asking Australian service providers, no matter how large to foot the bill for file sharing networks, movie downloads etc. is a non starter as an idea. Would you like Optus to go the way of One.Tel? No bandwith is considerably worse than limited bandwith.

As far as file sharing goes, why not start building networks using wireless links etc. in urban areas (I realise this is a non starter in rural areas), or perhaps start an ISP who's emphasis is on file sharing (connections provided via wireless or ADSL (I have a feeling such an ISP would quickly attract the attention of the Australian equivalent of the RIAA)).

As far as distributing Linux ISOs via CD/DVD is a far more efficient method while bandwith is still limited. Perhaps talk to Optus about putting certain large files like this online for download at reduced bandwith cost (i.e. the bandwith used is say 10% when downloaded from their "mirror"). This could be a different way for Optus to distinguish their service from Telstra's (perhaps some sort of voting system could be implemented to request files)

Re:I'm going to disagree with some of what you sai

[Combuchan]

I've heard I don't know how many times that "xxxx is enough for the average user" in computer-speak, and every time it's short-sighted.

It is short-sighted, but complain when the Next Big Thing actually arrives. We have 2.4 GHz Pentium IV's, and I'm 'chugging' along on my 500 MHz Celeron, and the most processor intensive operation I do (aside from dnetc, but that doesn't count) is run mplayer, which eats 25% CPU. This is why the tech industry's in a slump, we don't have a Big Thing and haven't had a latest Big Thing for a while now.

Look at the latest greatest game, look at how beautiful the 3D is, and then look out your window and realize how truly shitty it looks and you get the idea. We will need more computing horsepower for graphics and AI and everything else, and we will definitely need more bandwidth than 3 lousy GB per month.

And in twenty years when we're at that point, we'll have a far better Internet infrastructure than what we have now. You presume that the 3GB a month limit will still be the same in the year 2022, and it won't. Sorry.

The problem is, the Internet is as much an entertainment medium as anything else, and it's competing with technologies that are not pay-as-you-go, like television and so forth.

While the Internet is an entertainment medium, it differs from traditional devices by the method of transport: Waves vs. bytes. You can put a 100,000 watt FM tower and cover millions of people with your radio station. With radio and television, you don't pay for each user like you do with the internet distribution.

Calculate the bandwith costs to cover four million people listening to 128 kbps Internet radio instead. To serve this, you'd have to be thinking 4,000,000 * 16 * 1024 bytes per second. Each OC unit (Optical Carrier, as in OC-3) transmits data at 51.84 Mbps, or 6,794,792 bytes per second. Divide out and you're going to need an OC-9645.

Even if this were a regular day, ie, not four million people listening, and you had a fraction of the total listeners, you'd still have to serve massive bandwith out; the costs of which would be far more than any large-market FM radio station could cover.

I'd rather put up my 100,000 watt FM antenna.

I hate to rain on your parade, but the Internet is not the best method of distribution for, uh, packaged entertainment, like pay per view and radio and television shows. Maybe in 2022 when we have your true-to-life 3D, things will change, but it is unfortunate that in 2002 we have advanced so far but still have a long way ahead of us.

If you disagree, reply.

Re:Perhaps broadband should charge 'per megabyte'?

[newt]

The thing that always chokes me up about these sorts of conversations is that the people who have strong ideas about how much it "should" cost don't actually know what they're talking about.

Just a hint: Anyone who is even remotely interested in thinking about how much bandwidth in Australia "should" cost really needs to understand that there is no such thing as a T-1 in Australia. Trying to make judgements without that kind of basic knowledge is a bit like making judgements about how much cars "should" cost without knowing that they're made out of steel.

If the deficiencies in your knowledge are really that basic, you just aren't qualified to comment about how much it "should" cost. A fundamental understanding of the market conditions in Australia is required before you can put yourself into the position of making authorative statements about costs.

What you are really doing is taking a US-centric view of the Internet, and applying it to other parts of the world -- And anyone who lives outside US territory will be able to tell you that that's just nonsense.

- mark
Network Engineer, Internode