how-to block ads
|
funchords
Hello
Premium,MVM
join:2001-03-11
Washington, DC
 ·Verizon Online DSL
 ·Skype
| reply to funchords
Re: Complain about Comcast=Have your Reputation Scrutinized
My reply to the attack -- | |
Nerdtalker
Working Hard, Or Hardly Working?
Premium,MVM
join:2003-02-18
Tucson, AZ
clubs:
| said by funchords  :My reply to the attack -- I think your reply was extremely professional, to the point, and exacting.
The part at the end about the blatant mudslinging going on from Comcast's side is very well done.
I find it ironic that they see fit to make claims that you don't understand the network, when it is the blatant level of secrecy they approach these kinds of supposed "end-user-transparent" network changes which prevents you from knowing about it in the first place.
Honestly, all this stuff is making me wonder whether ISPs shouldn't be required by the FCC to deploy, maintain, and update some basic standardized network topology map for their customers. This way, there's no guessing about if/where/to-what the NebuAD boxes are connected to, and where Sandvine is running from. Some sort of transparency may well benefit the consumer, as well as the FCC in dealing with these kinds of issues.
The problem with the whole thing is that, for some reason, the "last mile" has decided it can arbitrarily violate the most sacrosanct of TCP/IP routing rules; only devices at the "edge" of the network actually change or modify data. The stipulation is that everything in-between doesn't do SPI, doesn't deliver user-tailored ADs, doesn't track browsing metrics. All that this news does is tell me that the edge to edge stipulation of network routing is basically dead. If ISPs can do it, what's stopping Level3? What's stopping all the hops between you and me from inserting their two cents, sending reset packets, or throttling?
--
"Some people never see the light till it shines thru bullet holes." -Bruce Cockburn
I'm testing Gmail's spam filters: Broadbandreports1@gmail.com
Spam: 12900+ messages currently using 406 MB. | |
netcool
@comcast.net
| reply to funchords
In reference to the bit about network upgrades.
Doesn't changing upstream modulation profiles count as an upgrade? Going from 16QAM to 64QAM for example gives you a nice bit of extra upstream bandwidth. Or even going from a 3.2mhz 16QAM carrier to a 6.4mhz 16QAM carrier for that matter. Comcast may not have implemented DOCSIS 2.0 everywhere but it seems to be implementing it on a as-needed basis. Which is probably what prompted them to give everyone a boost in upload speed in the first place.
And to your point about the placement of sandvine servers, I think you should have visited Comcast so you could re-evaluate that point.
It's not surprising that this whole debate has generated into an attack on your character sadly. It is a legal move to discredit one of Freepress' star witnesses so to speak. | |
funchords
Hello
Premium,MVM
join:2001-03-11
Washington, DC
·Verizon Online DSL
·Skype
| said by netcool :
Doesn't changing upstream modulation profiles count as an upgrade? Going from 16QAM to 64QAM for example gives you a nice bit of extra upstream bandwidth. Or even going from a 3.2mhz 16QAM carrier to a 6.4mhz 16QAM carrier for that matter. Yes, if and where that happened. But Comcast has been rather vocal about where it does network improvements and it hasn't been saying much about doing DOCSIS 2 upgrades. I think it's been holding out for DOCSIS 3. (Wouldn't you if you were them?) It has DOCSIS 2 in the former Adelphia areas and anywhere else it may have inherited the equipment from a network purchase.
It's possible that they've been quietly upgrading the DOCSIS 1.1 network to DOCSIS 2.0 in order to change the upstream modulation. But their silence on the matter, pressure by investors to keep capital costs down, and the advent of DOCSIS 3 indicate that, by and large, the DOCSIS 1.1 areas will remain DOCSIS 1.1 until DOCSIS 3 rolls around. said by netcool :
Comcast may not have implemented DOCSIS 2.0 everywhere but it seems to be implementing it on a as-needed basis. Which is probably what prompted them to give everyone a boost in upload speed in the first place. Can you give some examples?
said by netcool :
And to your point about the placement of sandvine servers, I think you should have visited Comcast so you could re-evaluate that point. Even if they told me point blank (which they have, in filings), I'd still have to point back at the data which indicated that it was at the Access Router. The test result is the result. Why would the TTLs be exactly right for the Access Router and be all wrong for the CMTS? We are talking about TTLs on forged packets, though -- but that they would choose exactly the wrong TTL is somewhat of an unlikely coincidence.
I also wonder if things changed. I first detected this in March of 2007 and my tests that showed TTL were in August '07 (I think). Could Comcast have bought more Sandvine and/or reconfigured between then and now?
But as I said in the filing, it's ultimately a moot point since Sandvine doesn't look at congestion so where to place the box in the network has no effective difference (and it is more efficient to put it at the Access Router since Congestion is not even looked at before blocking uploads).
said by netcool :
It's not surprising that this whole debate has generated into an attack on your character sadly. It is a legal move to discredit one of Freepress' star witnesses so to speak. Although I see where you're coming from, that's ultimately no excuse and no way to treat a very-long-term customer. We should be debating with data being shared openly, not personal attacks to divert attention from the facts.
--
Robb Topolski -= funchords.com =- Hillsboro, Oregon
Comcast: We never did anything wrong, and we'll never do it again...
| |
netcool
@comcast.net
| Yes, if and where that happened. But Comcast has been rather vocal about where it does network improvements and it hasn't been saying much about doing DOCSIS 2 upgrades. I think it's been holding out for DOCSIS 3. (Wouldn't you if you were them?) It has DOCSIS 2 in the former Adelphia areas and anywhere else it may have inherited the equipment from a network purchase. DOCSIS 3.0 does not fix the problem of using old modulation schemes on the upstream. It introduces channel bonding on both the up and downstream but what would you rather channel bond, QPSK channels or 64QAM channels? Getting the plant to DOCSIS 2.0 is an important step in the grand scheme of things. To get to a 50/10 speedtier you basically need to have DOCSIS 2.0 upstream modulations deployed. Upstream channel bonding will not be available for awhile yet.
As to your question about which areas are seeing DOCSIS 2.0, well we all know Minnesota has DOCSIS 2.0 deployed. I believe quite a few markets have deployed 6.4mhz channels but that is hard to glean from one's cable modem.
As you mentioned in your rebuttal letter virtual node splits can be implemented with relative ease and with little impact to the end subscriber. With the way most CMTS line cards are designed there usually ends up being quite a few extra Upstream ports available for use. Using multiple returns to a node is another way to effectively "uprade" capacity. As is changing modulation schemes. I think Freepress' generalization of the upgrades as window dressing is based merely on opinion and not fact. We know the options discussed above are available to Comcast and would cost them relatively nothing to implement. My point is that it is fine to remain skeptical with Comcast but to note that upgrades are possible without costly upgrades to "the core network." | |
Sturm
@alconlabs.com
thumbs down from:
TKJunkMail
| reply to funchords
It is clear at this point that the only way for Comcast to deliver unhindered service that they oversold to the customers is to have the company massively fined and split apart into smaller entities that would be more customer service oriented.
Hopefully the federal agencies will see this massive problem Comcast has become and will act accordingly to resolve it. Comcast has chosen to buy many other smaller cable providers and the experience has been negative for the many of the acquired customers (InsightBB, Time Warner being some). Clearly, those customers were much better off before both technologically and price wise.
Clearly the company does not intend to deliver service to meet the rising demand of the market, nor does it even want to introduce metered access, because even if the customer paid more they would still not be willing to provide the service they already sold, therefore they choose to delve into legalese explanations why really black is white.
Maybe they simply can't. It is difficult to know since they are so opaque in their policies and confused in daily communications. It is too bad that rather than work with the many knowlegeable customers toward a mutually rewarding business relationship the company chooses the heavy handed approach typical of a monopoly, but such is the way sometimes things go.
When they accuse knowlegeable people as Mr. Topolski of not knowing their network, they are only themselves to blame. We, as customers, have as sources of information only confused CSR's or threatening "abuse" department reps that would not even disclose usage caps that supposedly do not even exist. We can draw conclusions only empirically and reactively regarding network architecture and topology vs. preferred cherry picked companies that Comcast chooses to disclose minimal network information to.
We have to deal with the stress of trying to figure out what affects our traffic while we have purchased "unfettered Internet access".
It is important for our country to be a leader in Internet access and development of new technologies as both a promoter of technological growth and as a stream of revenue. The greed of a huge cable video provider should not be the determining factor why this should not happen. | |
-
|
