bdb

I agree that Comcast is in fact trying to classify this service as a Title VI service. The implementation of this specific service seems to have far more in common with Netflix than standard cable service (even including VoD streams delivered to your STB.)

Whether the FCC considers this as a Title VI service or not is yet to be seen.

hey Zac. I'm the author of the post.

That, or something like it, is Comcast's argument. Which makes no sense. This is a video streaming service, it's prioritized, and it's delivered just like other video streaming services are. Hopefully the data reflects this.

Hi. I'm the author of the original post; thanks for your comment.

To be clear, VOD streams to your STB are delivered over IP, just not through your cable modem, router and home network. You're right, of course, that Comcast claims these streams are no different than that VOD service.

The point where I think Comcast's argument falls apart is that streams to their Xbox app are delivered through the exact same network elements both inside and outside your home, as e.g., Netflix streams.

Doubtful. I think both are too expensive, but both data and SMS plans are only able to be priced where they are because of breakage[1].

[1] http://en.wikipedia.org/wiki/Breakage

I'd be willing to bet they subsidize their data pricing with that text messaging revenue. How many of AT&T's iPhone activations come with a fat text-messaging plan? Data plans are low-margin; text messaging is high-margin.

AT&T, along with most other carriers, made a tremendous amount of capital investment in the last few years to improve the service they provide to smartphone owners. iMessage is a threat to their margin-stealing model.

If they made text-messaging free—or just counted its data use against your data plan—but raised the price of your data plan by $15/mo, how would you feel?

Glad I'm not the only one who manages to order just one slightly incorrect part for every hardware project I work on.

In certain cases, e.g., the cables AT&T laid between Hawaii and the mainland US, they were decommissioned from active service and then used to power a network of undersea sensors used for research by the University of Hawaii:

http://aco-ssds.soest.hawaii.edu/

Actually, there are much better photos in this PDF: http://www.nec.com/en/global/prod/nw/submarine/pdf/submarine...

Look at the last page for a photo of one of the amps with people in the frame for scale.

Submarine cable systems often employ repeaters (hardened EDFAs). I remember one vendor bragging that not one of their units had EVER failed in service.

If you want to see what they look like: http://www.nec.com/en/global/prod/nw/submarine/product/ns-se...

Ooh, neat. What kind of hardware are you using? (And are you using gnuradio? Does the gr-smartnet module on CGRAN actually work?)

Using the MAC address as a unique identifier alone seems like a terrible idea(that is, without other secret, constant data included in the hash, similar to how UDIDs are generated.)

Your device is literally broadcasting your unique identifier to everyone in a several hundred foot radius.

Pure speculation on my part, but given the limited suitable equipment space available in each BART station, it's likely that BART has deployed a neutral host (i.e., BART-owned, not carrier-owned) distributed antenna system[1] throughout their stations. In that case, it's easy for them to cut power to the actual RF gear without carrier intervention (and without having to worry about those pesky strings of 48V batteries keeping things running.)

The bigger issue to me is that is almost certainly illegal to willfully disrupt 911 service, and it's unlikely that whoever flipped whatever switch, be it disrupting commercial power to carrier-owned base stations or a BART-owned DAS, was able to disrupt texts/voice calls to other parties selectively without disrupting 911 service.

[1] http://www.infinigy.com/PDF/InfinigyNeutralHostWhitepaper.pd...

No source yet, but there's a demonstration VM.

Unfortunately, it looks like they decided to choose to license the code under the AGPL[1], so it will probably be of little use to most of us.

[1] http://hpccsystems.com/print/518

Awesome.

Here's a link to the actual project: http://hpccsystems.com/

There have been two mentioned on HN recently: zStack and StackMob (the latter is run by a friend and former coworker of mine.)

Ironically, zStack has the OP listed as their only reference customer. Er, guess that's not working out so well?

Here's a link to the relevant part of the Chromium source tree: http://src.chromium.org/viewvc/chrome/trunk/src/chrome/brows...

Where is Amazon's?

A few years ago, I signed up for a Borders rewards card. Gave them an email address of borders@mydomain.com. Weeks later, began getting tons of spam to that address.

So uh, I'm guessing that this Borders employee isn't the only one who knows about this.

Funny, I had that exact thought.

This is a pretty clear example of someone not even having a basic grasp of how a network functions, yet making a big inflammatory website about it anyway.

I predict this app will be renamed within the next six months.

Really interesting that your submission has to be in the form of something that can run on Native Client. I wonder if they're using this to stress-test implementations of NC.

Which falls over when db01 ain't a database machine anymore.

When I ran a large network, machines didn't physically move very much (at least, far less often than their logical functions changed); we had a naming scheme based on datacenter code, rack ID and rack position:

sjc02-X0501-19

would be the machine in sjc02, rack X05.01, unit 19.

Far more important than the actual scheme you use is consistency. Pick one scheme and stick with it. Easier when you pick the right scheme the first time. :)

> In short, any IPv6 transition plan involves everyone having an IPv4 address, right up until everyone has an IPv6 address, at which point we can start dropping IPv4, which means IPv6 will start being useful.

Sigh, no. Not correct -- in the last 8 or 9 years, engineers have been cooking up various ways to bridge the gap. The transition landscape looks a lot different today than it does in 2003. Comcast (the largest MSO in the US) has detailed their transition plans here (tl;dr, it's v6-only customer addressing, with v4 reachability via DS-lite and CGN): http://www.comcast6.net/

Summary here: http://en.wikipedia.org/wiki/IPv6_transition_mechanisms

This plus the 4G Nexus S is going to make it really hard to hang onto my old T-Mobile/Nexus One combo. No equivalent of my T-Mobile no-contract post-paid service, though.

Hercules is cool -- I last goofed around with it almost 10 years ago, but it was complete enough then to run VM/370 and MVS/370. If you're interested, start here: http://www.cbttape.org/vm6.htm and http://www.bsp-gmbh.com/turnkey/

If you had a truck and a bunch of free time, say, $499?

http://cgi.ebay.com/IBM-S-390-G5-Parallel-Enterprise-Server-...

The interesting thing about mainframe hardware is that it's designed to last forever, so the "old" stuff (which really means "brand new" stuff in mainframe terms) is still interesting.

That's fine. Apple "sells" HD movie rentals (~3 GB) which are only useful for 24 hours.

Gerrit feels so... heavy. Are any of you using it at a small-ish company? Does it get in the way of getting work done?

We take for granted today that there are several fast, stable DVCS generally understood by an average engineer (in other words, the bar for using a DVCS has been lowered in the last few years and it's not a purely academic endeavor.) When Google was ramping up, Perforce was dramatically faster and more reliable than the almost all of its competitors. It was one of the most common source control systems to find at Valley companies.

Why are they still using it? Inertia, I assume.