Slashdot Mirror
CenturyLink To Buy Level 3 For $34 Billion, Create a More Formidable Competitor To AT&T (bloomberg.com)
In what is seen as a move to build a more formidable competitor to AT&T, rival CenturyLink today announced it is buying Level 3 Communications for about $34 billion in cash and stock. From a report on Bloomberg: Both companies have amassed giant networks to haul internet traffic through deals over the years. Level 3 is one of the largest providers used by internet services including Netflix and Google to route traffic across the web, operations that would bolster CenturyLink's core offerings to businesses. Level 3 was the second-biggest U.S. provider of ethernet services -- running high-bandwidth internet connections for companies -- in the first half of this year, trailing only AT&T, according to Vertical Systems Group. CenturyLink was fifth on the list.
So you're saying they are better than At&T?
"That's the way to do it" - Punch
The government needs to block this merger as well as the AT&T/TW merger. We need to have more choices when it comes to internet providers, not fewer. Pretty soon there will be a single source for the internet and they'll give you a 'bend over, take it or leave it' choice and that's all.
More formidable competitor? Those assholes can't even be bothered to repair their broken pedestal covers throughout the city! They're literally leaving their twisted-pair splice points open to the weather!
They've made a choice to not maintain their network infrastructure, both for legacy and for their DSL broadband customer base. Why should we trust them do do any better with anything else, let alone a build-out that's not even really that far along yet?
Do not look into laser with remaining eye.
Honestly, you probably don't have the redundancy you think you do. I see this crap all the time from the other side. Some hospital or bank orders two connections from Telco A and Telco B, and Telco B turns around and buys the local access from Telco A. Even if the local loops are different facilities, it's almost certain the same long haul fiber is being used. You'd get better reliability if you ordered diverse service from Telco A, then they would know and be able to control both paths.
09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0
AT&T has never sent me someone else's Social Security number. Can't say the same for Century Link.
Sorry, teleporters just kill you and then make a copy. A perfect, soul-less copy.
It's interesting to see the larger (old) AT&T successors competing back again...
New AT&T is, of course, former Baby Bell "Southwestern Bell", which bought the old AT&T, including the AT&T Long Lines department (and, of course, AT&T Long Distance, which is still a functioning corporate unit and is still "the old AT&T"). With CenturyLink, we'll now have two Baby Bells with significant fiber footprints. (As others have pointed out, AT&T / TW doesn't involve TWC/Comcast though.)
It's arguable whether the reconstitution of mega backbones was inevitable. Although divestiture helped competition (MCI, of course) and helped explosively develop the technical capacity needed for internet growth, economies of scale do come back into play. Especially when massive capital outlays come into play.
Hire a Linux system administrator, systems engineer,
Before they go buying up a titan like Level3, they need to be spending at least 1/4 of that cash in client support and relations. I've heard (and experienced) nothing but bad things about their client support. They have a serious disconnect between the call center level and field tech, making for awful ticket response and lousy on site times.
First rule of holes; When in one, stop digging.
I'm not sure that's true. Make "the last mile" a standardized utility. This would make it far easier for many competitors to enter the market because they wouldn't have to string potentially redundant wires to jillions of homes: they'd only have to hook up to routing nodes, set roughly a mile apart from each other.
You can then change ISP and content providers without anyone having to visit your house: it's all done at the routing nodes. (That part could perhaps even be made remote-controlled so that a truck doesn't even have to visit the nodes.)
The last-mile problem is the current bottleneck to competition. Remove that barrier by shifting it to a utility, and we then get real competition instead of the 2 co-shitty ISP's a typical city has to choose between.
Some may argue that a utility would be slow to add speed improvements, which would typically increase over time based on past patterns. But I'd sacrifice growing speed for reliability and choice. Reliability and choice are something the current oligopolies consistently suck at: they crawl on weekends and force you buy crap you don't want to get what you do want (bundling). Let alone crappy customer service.
Table-ized A.I.
Why on earth would you want two or more ISPs to compete over you over a common backbone? That would require competition, something the ISPs obviously don't want. It would also require someone to foot the bill for that last mile infrastructure. ISPs have fought for years to not have to share their lines, nor have a city be able to put in their own lines. It's much better the way it is where one ISP control the area and kick and scream and pout whenever there's an effort to try to modernize the service.
I've had them terminate in different exchanges, only to find out that the second exchange was daisy chained off the first. The idiot who bough the circuit even specified (And paid for) diverse penetration for a line that laid in the same trench. And the companies saw that CIOs were idiots, and banked on it.
Learn to love Alaska
As someone who has priced and engineered these services from the carrier side, what you have to do is request maps. If it's a layer 3 service you also need to make sure the two terminating routers are fed diversely. It's pretty common to be able to get a Google Earth KMZ file of the physical path, but expect to pay much more because last mile diversity usually requires construction.
09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0
If you don't request a map, the specific diversity details might not actually get into the hands of the people engineering the circuit/service. It's very common for the order to be entered into some system and the actual service order document to only be looked at by people who don't even really understand the language around diversity requirements.
However, if you request a map and (if needed) raise a stink, someone on the sales/order management side has to get in touch with someone in Engineering. When I was in network engineering I would always go to great lengths to wring actual signed service orders out of order management. Usually I'd be told that I didn't need it a few times before I got it. Pretty often I'd find something that wasn't put into the order database, like a diversity requirement or some technical requirement like supporting 9000+ byte frames or transparent to double-tagged traffic. Sometimes I'd find out that the OC192 was actually 10GigE or vice versa.
Ask for maps, trigger that Engineering involvement.
09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0