SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : AUTOHOME, Inc -- Ignore unavailable to you. Want to Upgrade?


To: E. Davies who wrote (9332)5/9/1999 6:49:00 PM
From: ahhaha  Read Replies (1) | Respond to of 29970
 
No, I meant ISP's.

No, you have to mean ISPs of the future, cable ones, not ones as they currently exist. We can't validly use the term "ISP" when referring to ATHM. We do it colloquially because they provide an Internet service. The kind of infrastructure of copper ISPs makes the difference here quite a difference. Copper ISPs have the ubiquity of the installed telephony plant and the versatility of access, but the telephony network would allow consumer choice completely if the local market was restructured like the non-local market has. The current arrangement allows you to choose any ISP, but not any local carrier.

I understand Franks comments to mean that each ISP wanting to use the cable network would need to build/lease a data feed to each individual headend in the field. I picture a headend serving several hundred homes so there would be a great number of headends. Right or
Wrong?


Almost both. ISPs mainly lease, they avoid building. Why not have a partner handle all of that? Haven't you heard how fearful AOL is to try to mount a cable system? They need a partner, an MSO.

I understand that "hybrid fiber-coax" means that the network is coax in the local loop from the home to the headend and fiber connecting the headends and the MSO "intranet" together. Right or Wrong?

Wrong is use of terminology. The term "coax" refers to the junk you have hanging out of your tv currently. The local loop which is commonly terminated at the headend has to be HFC, it has to be two way since on the loop everyone wants interactive capabilities. Coax won't support that bidirectional transfer. It will support high speed reception like 100 channels of what you always wanted to see on the tube.

In other words that the MSO runs the local connection and the ISP handles the services like e-mail and content.

Yes. I don't know about content. Maybe cut-rate cable ISPs won't offer any specialized content as part of the monthly rate. Maybe they'll just give you high rate Internet access. What specialized content does ATHM have now?

An ISP has physical equipment does it not?

You are mixing copper ISP and cable service. I think this is causing confusion.

The coax hanging out of your tube goes outside to the local dropdown node and that is where it meets up with HFC which defines the loop's "backbone". Coax can't handle all that chat so you can't have it in the local loop. You can have it handle T-1 between your house and the node, but that's it. It's a very poor arrangement. You should have Pure from your STB out not to a headend, but out to the metro super ring.

The cable headend which terminates the loop contains secondary SGI caching servers which aren't as powerful as the Sun dynamic servers residing further up in the RDCs, but are adequate for their service areas. For purposes of load redundancy support it would be necessary to beef up these servers or co-locate several more of them. It hasn't been determined whether multiple carry of many cable ISPs should proceed by every ISP having allocated caching storage or whether there should be sole server dedication or whether it all should be done virtually only. We don't have an in-place system to go and modify, rather, it is a system in evolution built around the exclusivity assumption of the cable partners. That is, one ISP, ATHM. Why would the cable partners develop such contingency plans since the whole business is so primitive it's a miracle that it works at all.

Does not @home own (or at least lease) the servers that provide local caching for instance?

Yes.

The routers that connect the MSO network to the "Internet" and to @homes "intranet"?

Yes.

The servers that provide the "services" (e-mail, web pages, content etc..) Right or Wrong?

Right.

Your implication here is that the MSO owns and manages the network that connects together all its head ends and directly ties that network to various internet access points. I didnt think they had that level of sophistication. Right or wrong?

Close. MSO purview essentially ends at the headend. From there upwards it is @home's bag. They jointly furnish the headend according to the function of the needed equipment. But the bag doesn't contain all that much significance relative to the topics we talk about or are important to all these "deals". For example, I believe Exodus is handling more and more ATHM traffic. The RDCs are pretty much automated so the NOC which maintains the ATHM system needs to interface with another equivalent network management as the subscribership mounts and the company has to create more NOCs or offload those tasks to someone like Exodus. As with the partitioning of functionality at the headend to accommodate potential co-carry, I'm speculating here, so don't hold me to it. It just seems that it isn't necessary for ATHM to in-house support nation or worldwide traffic management. You concede autonomy for cost efficiency.

In many ways my question boils down to: "What is an ISP anyway?"

I have wandered around a bit above in order to superficially address this question. Fact is. no one knows. It is still in development. Ask the Castro Valley subscribers.

If its just a mail server hung on the internet *anyone* can do that!

AOL is scared to attempt it. Sounds so easy but consider all that is involved: T, the world's greatest cable operators, computer manufacturers, sophisticated equipment manufacturers, telcom carriers, worldwide network, cutting edge technology everywhere, and much more.

Cable modems have features that effect the network that are controlled through the network, otherwise you would have a completely unmanagable situation. Who controls those features?

You are asserting existence of something which doesn't exist and then asking about it. I have no idea to which "features" you are referring.

I understood Franks statement to mean that the 750K-1Ghz bandwidth (available on certain quality of coax wires) is also intended for
bidirectional use.It occured to me that that frequency could do a lot to solve the problems of the overloaded upstream if it was available
for internet data.


I am going through all this boring stuff because you need to know some basics as do most on this thread. However, it is all in this thread. We have gone over all this stuff again and again, so I hope you guys who are still hanging around in the future will get it down so that you can explain it to others. Knowing all this stuff isn't actually of much value in assessing the company's prospects. That kind of thing is in the subscribers added per unit time. Otherwise I'm giving an intro into networks that one presumably gets at the university. That's boring.