DISQUS

DISQUS Hello!  The comments on this profile are unclaimed and thus are unverified.

Do they belong to you? Claim these comments.

joey5's picture

Unregistered

Feeds

aliases

  • joey5

joey5

3 years ago

in The Broadband Buffet on The Technology Liberation Front
Burying terms in the fine print of service contracts that are contradictory to product labeling and advertising is a deceptive business practice. As long as the ISPs persist in this practice, I'll persist in picking the terms out of the contradiction that I find most favorable to my own situation.

I imagine that this will all resolve itself eventually, with metered access or tiered service agreements. This is how telephones work, right? And telephone networks are probably our closest analogy.

There's also a technical aspect here: if I hide all my data by tunneling it through https, there is no way that an ISP will be able to determine if the requests/responses originated with me or with my neighbor. The data is indistinguishable.

Why is that important? If the ISP is to take action against bandwidth "thieves" (as you'd call them), they must necessarily punish the honest as well. My penchant for, say, downloading lots of songs from iTunes will be indistinguishable from me sharing my connection with a few dozen neighbors who just do average surfing. What can the ISP do in this case? Punish anyone who uses "excessive bandwidth."

And there is where the whole argument falls apart: my bandwidth is already capped by the modem I lease from the ISP and the routers that it connects to. I can't get more than X kbits. And I *pay for* X kbits. So the ISP can only punish me for using what I paid for.

Again, truth in advertising would fix this -- ISPs could advertise that you are really only getting the equivalent of a 56K dialup modem (on average) if thats what they expect their customers to use, instead of touting their service as 6Mbit, etc.

3 years ago

in The Broadband Buffet on The Technology Liberation Front
All analogies have problems. But the all-you-can-eat-buffet has a big one: the buffet doesn't meter/restrict your throat/stomach bandwidth.

When I pay for a broadband internet connection, I'm paying for X kbits of bandwidth. Nowhere in my contract does the ISP explain that this bandwidth is only to be available in bursts only, and that my average bandwidth should really be Y kbits/s (where Y < X). If they did, okay, fine; they are really selling me Y kbits of bandwidth, and some truth-in-advertising laws should kick in.

As long as the ISP is selling me a connection with a specified transfer rate, they shouldn't care what bits I fill that transfer rate with. Whether they be bits that I initiated by reading techliberation, bits that my wife initiated by watching videos at youtube, or bits that my neighbor is using to download a security patch, it shouldn't matter. As long as I'm only consuming my alloted bandwidth, the ISP shouldn't know or care what the bits mean, where they came from, or why they are there.

The buffet analogy would work perfectly, if, instead of paying $7 for the buffet, you paid a monthly fee of $X for Z ounces of food per minute. Under such a contract, it wouldn't matter if I ate the food myself, regurgitated it in the bathroom, took it home to the dog, or carried it out and resold it on the sidewalk.

Perhaps the real problem is that the ISPs advertise their services as a data transfer rate, instead of as what they really intend (which you seem to imply as something far less than the advertised data transfer rate). That dishonesty is not my problem. It's the ISPs.

3 years ago

in News Flash: Larry Ellison is a Megalomaniac on The Technology Liberation Front
Oracle didn't even do any core filesystem work -- they just decided to port OCFS (Oracle Cluster Filesystem) to Linux and try to get the kernel maintainers to accept it.

That's not to say that OCFS is no good, or that Oracle didn't make a valid contribution -- they did. But it isn't exactly earth-shattering. RedHat's GFS does the same thing. And Oracle just joins a long line of old-school-unix shops trying to get their hacks into Linux (IBM contributed JFS, SGI contributed XFS, etc) -- none of which form any type of key component for most Linux users, who almost universally just use EXT3 or ReiserFS (the two default filesystems).

Answer me this: is DeLong really siding with Ellison over the Open Source Community? I would have thought he'd at least choose his companions a little more judiciously. Oh well, anything that explains away open source must be good and true, so he gets on the boat. Good job DeLong! You're convincing us all now!
Returning? Login