Thanks for Jeff Liebermann for suggesting the Costco cable modem!

[ceg]

Thank you Jeff Liebermann for being a technical wiz-kid genius!
http://i.imgur.com/obYQsU9.jpg

I just want to shout out to Jeff Liebermann for suggesting the Costco
cable modem (Motorola/Arris SB6141) for about $85 out the door, and again
thanks go to Jeff for suggesting I ask for a discount on the Comcast $50
"installation fee".

I hooked it up with a $1 CATV cable from Goodwill and then called Comcast
who first wanted to charge me $50 for installation, and then $15 when I
argued for free and then the guy used a "promo code" to get me down to $6
for their "installation" fee.

I couldn't get him any lower than $6, so I took that.

In ten minutes it was working fine. They did everything over the wires.

They charge about $45 monthly (I don't know what the taxes will be
though) for 45Mbps down and 5Mbps up.

They try to talk you into 100Mbps down and 10 Mbps up for $50 but we
tested it at 90Mbps out the modem RJ45 to a laptop and about 60Mbps using
the 2.5GHz WiFi out of the router attached to the modem using
speedtest.comcast.net as our test web site (and Firefox 39.0.3).

The only router Costco had was too expensive though (at $130 + about $13
tax for the Netgear Nighthawk AC1750). So, at Frys, I bought a $75 TP-
Link Archer C5 (do you think that was a good tradeoff?) for roughly about
half the price. http://i.imgur.com/obYQsU9.jpg

With Jeff's suggested Arris/Motorola cable modem from Costco, the $85 out
the door will pay for itself in a few months:

1. Saved $45 minus $6 = $39 on installation fees (but I had to supply a
CATV coax cable so let's call that a $30 savings to make the math easy).
2. Saved $10 each month on modem rental fees
3. I get a GREAT modem (fast!).

So, in the first month, I saved about half the modem cost alone, at about
$40; and in the next four or five months, the modem is paid for. Then,
each month, the modem, in effect, saves me another $10 every month.

You can't beat that equation!
Thanks Jeff for being so smart!
Santa Cruz is lucky to have you!

I bought the following, but will return the Costco router I think:
http://i.imgur.com/obYQsU9.jpg

TP-Link Archer C5 router ($75 at Fryes)
Netgear Nighthawk AC1750 router ($130 at Costco)
Motorola/Arris SB6141 cable modem ($75 at Costco)

Only question left is whether the Costco router is worth it for just two
kids sharing an Internet connection away at school.

[Phil Hobbs]

One good option for a router is to get a used Netgear WNDR3700 from eBay
for $20 and install OpenWRT on it. You get a decently secure router
with a lot of stooch for very cheap. I have a couple like that, and I'm
very happy with them.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

[Ed Pawlowski]

On 8/15/2015 7:11 PM, ceg wrote:

>

> Only question left is whether the Costco router is worth it for just two
> kids sharing an Internet connection away at school.
>

Sure, half the dorm will be downloading the latest movies and porn clips.

[Jeff Liebermann]

On Sat, 15 Aug 2015 23:11:31 +0000 (UTC), ceg
<curt.gul...@gmail.com> wrote:

>Thank you Jeff Liebermann for being a technical wiz-kid genius!

Y'er welcome. However, I'm no longer a kid and certainly have never
been a genius. Please keep your superlatives reasonable and knock off
the nymshifting.

>I just want to shout out to Jeff Liebermann for suggesting the Costco
>cable modem (Motorola/Arris SB6141) for about $85 out the door, and again
>thanks go to Jeff for suggesting I ask for a discount on the Comcast $50
>"installation fee".

I certainly suggested buying the SB6141 at Costco, but I don't believe
it was me that suggested negotiating the installation fee. At the
time, Comcast considered the fee as mandatory. However, subsequent
threats of litigation based upon charging for services that were not
rendered has magically made the fee negotiable. The problem was that
anyone that physically picked up their equipment at the local Comcast
store would not be charged an installation fee since it was a "self
install". However, Comcast didn't have a check box for those that
purchased their own equipment, and therefore billed everyone possible
for non-service. I ran into this problem with one new customer, so I
suggested a rather ridiculous alternative. They ordered service along
with the rental of a gateway device, which they picked up at the store
as a self install. They then purchased an SB6141 at Costco and
activated it. Once working, they returned the unopened gateway device
for credit. The customer is rather old, so I walked him through the
entire ridiculous ceremony, which included informing everyone within
range of what was happening and why.

>In ten minutes it was working fine. They did everything over the wires.

Impressive. Much depends on who you get on the phone. I've heard of
hour long ordeals and have experienced 20 minute activation acrimony.

>They charge about $45 monthly (I don't know what the taxes will be
>though) for 45Mbps down and 5Mbps up.

I've always found it odd that Comcast is unable to provide an accurate
estimate. Perhaps they're afraid of sticker shock?

>They try to talk you into 100Mbps down and 10 Mbps up for $50 but we
>tested it at 90Mbps out the modem RJ45 to a laptop and about 60Mbps using
>the 2.5GHz WiFi out of the router attached to the modem using
>speedtest.comcast.net as our test web site (and Firefox 39.0.3).

I've seen 179/12 Mbits/sec in Felton. One customer that had the old
12/1 Mbits/sec service now has more than double at 88/6:
<http://results.speedtest.comcast.net/result/999633544.png>
What happened is that on Tues Aug 11, Comcast magically doubled the
speed of residential internet service.
<http://www.santacruzsentinel.com/business/20150810/comcast-unveils-faster-internet-for-santa-cruz>
I'll spare you my conspiracy theories, but will add an anecdote. The
speed did NOT increase for business class Comcast customers. So, I
walked over to the Comcast store and asked if they could find out the
status and plans, if any. I was blessed with several useless phone
numbers, that connected me to people that didn't know, didn't care,
and didn't bother to obtain the information. Welcome to the legendary
Comcast customer service.

>The only router Costco had was too expensive though (at $130 + about $13
>tax for the Netgear Nighthawk AC1750). So, at Frys, I bought a $75 TP-
>Link Archer C5 (do you think that was a good tradeoff?) for roughly about
>half the price. http://i.imgur.com/obYQsU9.jpg

TP-Link routers are amazingly good products hidden inside amazingly
weird packages. I have a few in service that have had no problems
(after I updated the firmware). With the cable speeds that you're
getting, you should be running 802.11ac something to get something
near full speed. However, the reason the TP-Links is probably cheaper
is that it only does AC1200 (2x2) while the Netgear does AC1750 (3x3).
Whether you actually get the rated speeds depends heavily on your
client radios and RF environment. If it's with a tablet or
smartphone, forget it.
<http://www.smallnetbuilder.com/wireless/wireless-reviews/32651-ac1200-router-roundup-part-1>
Wireless performance is allegedly about 7 to 10 times your Comcast
performance, so you'll probably do ok.
<http://www.smallnetbuilder.com/images/stories/wireless/ac1200_roundup/ac1200_router_roundup_benchmark_summary.jpg>
Make sure you got hardware version v2.0, not v1.20.

>Only question left is whether the Costco router is worth it for just two
>kids sharing an Internet connection away at school.

Probably not. The extra wirless speed is if your kids are copying or
moving massive files via wireless. Hopefully, that's not happening.
Another is reduced latency if they are doing gaming. If you monitor
the wireless mode (with a sniffer), you'll find that in the presence
of even the slightest amount of interference, the router will switch
down to 802.11g speeds or about 25 Mbits/sec throughput maximum. The
higher speeds do have their use (such as using less air time) but
usually require an ideal RF environment in order to be used full time.

It might be fun to lock the wireless mode and speeds to 802.11ac.
Let's see if it can be done:
<http://www.tp-link.com/resources/simulator/C5_v2_simulator/Index.htm>
Yep. On 2.4GHz, you can select 802.11n mode only. On 5GHz, you can
select 802.11ac only. Try AC only, copy something across the network
via wireless, and see what manner of range you can get.

--
Jeff Liebermann je...@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

[ceg]

On Sat, 15 Aug 2015 19:57:19 -0700, Jeff Liebermann wrote:

> Make sure you got hardware version v2.0, not v1.20.

Nymshifting is part and parcel with privacy. Sorry. I hate it just
as much as I hate locking my bicycle up and removing the seat, lights,
and mirrors every time I park it. But, in today's world, we're forced to
do it.

Sorry about that. You're amazingly "open" with who you are, but I can't
be that open. Too much at stake.

You know who I am anyway - since I have emailed you in the past.

Anyway, thanks for the advice on the Archer C5 hardware version.
The router is up at the college but I have the box.

The serial-number sticker on the box says it's version 2.0 so I
got must have gotten lucky on the hardware version at Frys.

[ceg]

On Sat, 15 Aug 2015 19:57:19 -0700, Jeff Liebermann wrote:

> I certainly suggested buying the SB6141 at Costco, but I don't believe
> it was me that suggested negotiating the installation fee. At the time,
> Comcast considered the fee as mandatory.

The SB6141 modem is great, as it's fast, and it will pay for itself in
less than a year, what with Comcast dropping the installation fee to six
bucks and also dropping the ten-dollar monthly rental fee.

However, as I recall, the speeds went from 90Mbps down to about 60Mbps
down simply by switching from a modem-to-laptop wired connection to a
modem-to-router-to-lapop-wireless connection.

That's a huge drop in speed (measured at speedtest.comcast.net).

Is that normal to lose 30 Mbps just by switching from wired directly to
wireless?

NOTE: I just realized while writing this that we never checked the wired
speed directly out of the router, which I'll ask the kids to do for me.

[ceg]

On Sat, 15 Aug 2015 19:57:19 -0700, Jeff Liebermann wrote:

> Probably not. The extra wirless speed is if your kids are copying or
> moving massive files via wireless. Hopefully, that's not happening.

These are good kids.
They're not like I am. :)

The worst they do is *ask* me to get them a movie or two!

(jk)

[Ed Pawlowski]

Sure, your kids are special and won't do anything while away at school.

[Jeff Liebermann]

On Sun, 16 Aug 2015 05:33:58 +0000 (UTC), ceg
<curt.gul...@gmail.com> wrote:

>The SB6141 modem is great, as it's fast, and it will pay for itself in
>less than a year, what with Comcast dropping the installation fee to six
>bucks and also dropping the ten-dollar monthly rental fee.

Yep. The math is simple enough. If you're going to remain a Comcast
customer for more than 9 months, you might as well buy your own modem.

The only gotcha is if you decide to subscribe to the Comcast phone
service. In the past, it was possible to buy an Arris TM722G
telephony modem, which was on the Comcast approved list. No more.
Now, you're stuck with buying a TM862G/CT which is the very device I
was trying to avoid:
<http://mydeviceinfo.comcast.net/device.php?devid=421>
There are a few on eBay, but be careful. Many were previous on rental
and cannot be activated.

>However, as I recall, the speeds went from 90Mbps down to about 60Mbps
>down simply by switching from a modem-to-laptop wired connection to a
>modem-to-router-to-lapop-wireless connection.
>
>That's a huge drop in speed (measured at speedtest.comcast.net).
>
>Is that normal to lose 30 Mbps just by switching from wired directly to
>wireless?

Probably. I say probably because I don't have any idea what the kids
have for equipment other than the router. I also don't know if
they're connecting on 2.4 or 5Ghz.

60 Mbits/sec throughput suggests that they're doing 64-QAM if they're
on a 20 MHz channel.
<https://en.wikipedia.org/wiki/IEEE_802.11ac#Theoretical>
The problem is that the router should be capable of shoveling data at
a much greater rate. For example, the performance graphs show much
higher than 60 Mbits/sec wireless speed on 5Ghz:
<http://www.smallnetbuilder.com/images/stories/wireless/ac1200_roundup/ac1200_router_roundup_5_throughput_vs_attenuation.jpg>
but much slower on 2.4GHz:
<http://www.smallnetbuilder.com/images/stories/wireless/ac1200_roundup/ac1200_router_roundup_24_throughput_vs_attenuation.jpg>
My guess is that 60 Mbits/sec is on the low side of reasonable for
2.4GHz in the presence of interference from other routers in the dorm.

>NOTE: I just realized while writing this that we never checked the wired
>speed directly out of the router, which I'll ask the kids to do for me.

I also had a short list of tests using Iperf and Jperf for them to
try.

Incidentally, my fabulous DSL downloads at 1.2 Mbits/sec and you're
complaining about losing 30 Mbits/sec. I'm jealous and you'll get no
sympathy from me.

[ceg]

On Sat, 15 Aug 2015 23:10:38 -0700, Jeff Liebermann wrote:

> Probably. I say probably because I don't have any idea what the kids
> have for equipment other than the router. I also don't know if
> they're connecting on 2.4 or 5Ghz.

It was a cheap Costco $500 Win8 laptop that apparently only seems to
have 2.4GHz (I didn't see the 5GHz AP show up).

All we did was go to speedtest.comcast.net

[ceg]

On Sat, 15 Aug 2015 23:10:38 -0700, Jeff Liebermann wrote:

> My guess is that 60 Mbits/sec is on the low side of reasonable for
> 2.4GHz in the presence of interference from other routers in the dorm.

I must admit, I had never seen so many access points in one place.
The screen was *filled* with access points!

[ceg]

On Sat, 15 Aug 2015 23:10:38 -0700, Jeff Liebermann wrote:

> Incidentally, my fabulous DSL downloads at 1.2 Mbits/sec and you're
> complaining about losing 30 Mbits/sec. I'm jealous and you'll get no
> sympathy from me.

Likewise, the kids, at school, on cable, are doing much better than I am,
on WISP, at home, pointing at an antenna miles away in the mountains.

[ceg]

On Sun, 16 Aug 2015 01:57:36 -0400, Ed Pawlowski wrote:

> Sure, your kids are special and won't do anything while away at school.

I think they watch "Desperate Housewives" on the net. At least I saw that
as one of the web sites they visited.

[Ed Pawlowski]

One of the nice features that browsers have now is a privacy viewing
setting that does not save any tracks.

But none of that matters since your kids are so well behaved. Not
like the kids and grandkids I have.

[devnull]

On 08/16/2015 06:33 AM, ceg wrote:
> Is that normal to lose 30 Mbps just by switching from wired directly to
> wireless?

In my situation, yes.
And it was easier for me to install some network jacks than to deal with wifi's idiosyncrasies.

[devnull]

You mean "uploading", don't you. (A lot of porn is made in dorm rooms.)
Thank God I didn't have any daughters. ;-)

[sms]

On 8/15/2015 7:57 PM, Jeff Liebermann wrote:
> On Sat, 15 Aug 2015 23:11:31 +0000 (UTC), ceg
> <curt.gul...@gmail.com> wrote:
>
>> Thank you Jeff Liebermann for being a technical wiz-kid genius!
>
> Y'er welcome. However, I'm no longer a kid and certainly have never
> been a genius. Please keep your superlatives reasonable and knock off
> the nymshifting.
>
>> I just want to shout out to Jeff Liebermann for suggesting the Costco
>> cable modem (Motorola/Arris SB6141) for about $85 out the door, and again
>> thanks go to Jeff for suggesting I ask for a discount on the Comcast $50
>> "installation fee".
>
> I certainly suggested buying the SB6141 at Costco, but I don't believe
> it was me that suggested negotiating the installation fee. At the

<snip>

Comcast always seems to charge the $50 then waits for the customer to
call to argue about it. Probably many don't call.

I also got the Motorola modem at Costco and did a self-install and
initially was charged the installation fee.

The problem I had was with getting them to install a drop, with RG6,
from the pole. I told them I needed a drop because the old drop, which I
had never used, was RG59. The first three trucks they sent were
unprepared to do a new drop. Finally I used a tree trimmer to cut down
the RG59 so there was no old drop there to confuse them, and then they
were able to put in a new drop.

[ceg]

On Tue, 18 Aug 2015 01:53:29 -0700, sms wrote:

> Comcast always seems to charge the $50 then waits for the customer to
> call to argue about it. Probably many don't call.
>
> I also got the Motorola modem at Costco and did a self-install and
> initially was charged the installation fee.
>
> The problem I had was with getting them to install a drop, with RG6,
> from the pole. I told them I needed a drop because the old drop, which I
> had never used, was RG59. The first three trucks they sent were
> unprepared to do a new drop. Finally I used a tree trimmer to cut down
> the RG59 so there was no old dr

That brings up a good question which I didn't know the answer to.

Being a cheapskate, I bought my coax cable from the Goodwill for a buck.

I know there are different impedances for coax cable, but I didn't know
what letting to look for on the Goodwill coax cables.

So, the one I bought said "CATV" but I don't know if that's the right
cable.

Is it?

[ceg]

On Tue, 18 Aug 2015 13:28:22 -0600, Tony Hwang wrote:

> To check speed one has to use same server at busiest time of the day
> like 10 in the morning. I use Oookla speedtest.net. I you have a router
> and use QoS, it'll affect speed, bit less than what it should be.

We should probably list the various servers.
I'll make a start, but this is just from memory.

1. speedtest.net
2. speedtest.comcast.net
3. speakeasy.net/speedtest
4. ?

[ceg]

On Tue, 18 Aug 2015 13:28:22 -0600, Tony Hwang wrote:

> To check speed one has to use same server at busiest time of the day
> like 10 in the morning. I use Oookla speedtest.net. I you have a router
> and use QoS, it'll affect speed, bit less than what it should be.

I wonder if Jeff knows whether it's best to use the speed test supplied
by the ISP, since that's most likely to follow the same pipes?

That is, if I'm using AT&T, wouldn't the most accurate speed test be
one that AT&T supplies, which uses AT&T pipes?

Likewise, for comcast?

[Jeff Liebermann]

On Tue, 18 Aug 2015 19:30:42 +0000 (UTC), ceg
<curt.gul...@gmail.com> wrote:

>I know there are different impedances for coax cable, but I didn't know
>what letting to look for on the Goodwill coax cables.

You want something in 75 ohms, not 50 ohms. That means one of the
many mutations of RG-6/u.

>So, the one I bought said "CATV" but I don't know if that's the right
>cable.

CATV means CAble TV. That term hasn't been used in many years, so
it's probably old cable with potentially ancient connectors. It may
also be RG-59/u, which you should not be using (because most of it is
garbage).

Incidentally, a quick test of the connector is to just pull on the
connector. If it seperates from the coax cable, it's a lousy crimp,
bad connector, or both. If you want to make your own cables, buy
compression connectors and the right tools.
<http://www.ebay.com/itm/171227302904>
<http://www.ebay.com/itm/191352097155>
<http://www.ebay.com/sch/i.html?_nkw=coax+cable+stripper>

[Jeff Liebermann]

On Tue, 18 Aug 2015 19:38:23 +0000 (UTC), ceg
<curt.gul...@gmail.com> wrote:

>I wonder if Jeff knows whether it's best to use the speed test supplied
>by the ISP, since that's most likely to follow the same pipes?

That used to be the case, when the number of hops and total latency
limited the maximum speed that it would test. However, with jumbo
packets and large windows, that's no longer the case. The speed test
program will usually select the closest or fastest test server anyway.
I'm on Comcast Biz class service. I sometimes run Comcast speed test
from California to Washington DC. The latency figures are high (88
msec instead of the usual 13 msec) but the speed figures are roughly
the same as local. I do tend to favor using the ISP's test server,
but that's from habit, not necessity.
<http://speedtest.comcast.net>
<http://www.speedtest.net>
<http://speedof.me>
<http://testmy.net>
<http://www.att.com/speedtest/>
<http://www.speakeasy.net/speedtest/>
<http://myspeed.visualware.com>
and bunch of private iperf, iperf3, Jperf test servers.

>That is, if I'm using AT&T, wouldn't the most accurate speed test be
>one that AT&T supplies, which uses AT&T pipes?

Accurate usually means highest speed indication and minimum congestion
effects. At slow connection speeds, almost any test server will work.
When you get up to 100 Mbit/sec speeds as Comcast is doing, you'll
need to check a few test sites. Things get messy if your route is
through a router that blocks, throttles, or delays speed tests.
Probably the highest number is the least congestied, and therefore the
most accurate. Those that limit the number of simultaneous or
consecutive tests are probably the best. Comcast does that as when
they boosted the local speeds, it sometimes took me 10 mins to start a
tests because everyone in the area was running speed tests.

>Likewise, for comcast?

Generally true, but not always. Peering between ISP nodes is commonly
shared. Run a traceroute to a test server and see what's on the
route.

[do...@93.usenet.us.com]

In alt.internet.wireless Jeff Liebermann <je...@cruzio.com> wrote:
> I certainly suggested buying the SB6141 at Costco, but I don't believe

I bought an SB6121, a month or two before it would have made sense to by a
6141, but in any event, the Motorola/Arris anything is better than the
Arris/Arris that Comcast provides.

> suggested a rather ridiculous alternative. They ordered service along
> with the rental of a gateway device, which they picked up at the store
> as a self install. They then purchased an SB6141 at Costco and
> activated it. Once working, they returned the unopened gateway device
> for credit.

I knew I hadn't paid an install fee, and forgot how that happened. I took
the "$8/month" modem/router, because once upon a time, I had so much
trouble with cable that they replaced the modem several times before they
replaced the drop, and I didn't want to have that argument with customer
owned equipment. But, after a month or two of horrible WiFi, I went with
the 6121 and an Asus router.

> Impressive. Much depends on who you get on the phone. I've heard of
> hour long ordeals and have experienced 20 minute activation acrimony.

Mine took a couple of minutes on the phone.

In a reply to another post, speedtest.comcast.net shows IPv4 and
IPv6 numbers, speedtest.net only shows IPv4.

--
Clarence A Dold - Santa Rosa, CA, USA GPS: 38.47,-122.65

[ceg]

On Tue, 18 Aug 2015 16:21:52 -0700, Jeff Liebermann wrote:

> The speed test
> program will usually select the closest or fastest test server anyway.

This is good to know as I knew you'd have experience with the answer.

> <http://speedtest.comcast.net>
> <http://www.speedtest.net>
> <http://speedof.me>
> <http://testmy.net>
> <http://www.att.com/speedtest/>
> <http://www.speakeasy.net/speedtest/>
> <http://myspeed.visualware.com>

These are great test URLs to have handy, and I put them in my
database. Others will find them useful also. Thanks.

> When you get up to 100 Mbit/sec speeds as Comcast is doing, you'll
> need to check a few test sites.

I find, strangely enough, that if I run two or three sequential tests,
that the second and/or third test are far faster than the first. I dunno
why, but, it's almost as if the first test "cleans the pipes" or
something obscure like that.

> Probably the highest number is the least congestied, and therefore the
> most accurate.

This makes sense, although I had never thought about that before.
So, basically, run a few tests, and the highest result is the best.

[ceg]

On Tue, 18 Aug 2015 15:57:30 -0700, Jeff Liebermann wrote:

> You want something in 75 ohms, not 50 ohms. That means one of the
> many mutations of RG-6/u.

I don't remember it saying what the ohms were, but I'll visit this
weekend, and look at the cable. It had nice long copper wires sticking
out, so, I think it was home made, but sturdy nonetheless.

> CATV means CAble TV. That term hasn't been used in many years, so
> it's probably old cable with potentially ancient connectors. It may
> also be RG-59/u, which you should not be using (because most of it is
> garbage).

I'll look, this weekend, for RG figures, but I don't remember seeing
them on the cable when I first bought it. They're getting 90Mbps down
(and 10 Mbps up) out of both the Costco cable modem and out the back
of the TP-Link router Ethernet cable, so, it's not too debilitating
if it's the wrong cable.

The only bad thing is that they're getting only 60Mbps down over
the air, but they're stuck on 2.5GHz in a crowded spectrum, so, I'll
bring an Android phone with InSSIDer installed to check the noise.

I might even bring a spare ubiquiti nanobridge (Jeff knows what they
can do) to check the noise spectrum, and if I do, I'll try to capture
it to show you folks. That won't happen until the weekend though.

[ceg]

On Wed, 19 Aug 2015 03:09:23 +0000, dold wrote:

> I bought an SB6121, a month or two before it would have made sense to by a
> 6141, but in any event, the Motorola/Arris anything is better than the
> Arris/Arris that Comcast provides.

I never heard of "Arris" before. Why do they bother with two names
anyway? Why not just an "arris" or just a "motorola" modem?

> I knew I hadn't paid an install fee, and forgot how that happened. I took
> the "$8/month" modem/router

Comcast told me it's $10 a month for a modem, and $6 for the self install,
so, prices are different here in California.

> Mine took a couple of minutes on the phone.

It was pretty quick for me too. The guy just made all the lights blink a
few times and that was it. About 10 minutes. Maybe 15 but not much more
(I wasn't timing it though - but it was pretty quick).

> In a reply to another post, speedtest.comcast.net shows IPv4 and
> IPv6 numbers, speedtest.net only shows IPv4.

I never understood IPv6, so, I wouldn't know the difference.

[Jeff Liebermann]

On Wed, 19 Aug 2015 03:48:49 +0000 (UTC), ceg
<curt.gul...@gmail.com> wrote:

>I find, strangely enough, that if I run two or three sequential tests,
>that the second and/or third test are far faster than the first. I dunno
>why, but, it's almost as if the first test "cleans the pipes" or
>something obscure like that.

Not strange at all. What you're seeing is a problem with the speed
testing algorithm caused by various devices along the path buffering
or cacheing the data. Instead of getting an end to end test without
any intermediate buffering, you're measuring the speed from some
intermediate devices cache. This happens if the speed test uses
identical data files for each test, instead of randomizing the data
for each test. I have not checked if this is the case for any of the
previously listed speed tests. The usual fix of downloading a large
amount of data to flush the cache before performing the actual speed
test doesn't really work because it slows everyone else down and ISP's
hate that. This is another reason why you want to use the closest
server, with fewer buffers and caches in the path. Of course, the ISP
has a vested interest in producing the highest numbers and could
easily "optimize" their system to produce amazing results that can't
be duplicated by real applications.

>> Probably the highest number is the least congestied, and therefore the
>> most accurate.

>This makes sense, although I had never thought about that before.
>So, basically, run a few tests, and the highest result is the best.

Or, the highest number is what your ISP is throttling your
performance. In my office, the best I can do is bursts of 25
Mbits/sec and sustained traffic of about 12 Mbits/sec. However, when
it was first installed, I was getting 160 Mbits/sec because on a new
building installation, the cable installers wanted to know if the
system could handle the traffic from three business class customers.
So they temporarily turned off the throttling in order to make the
test and forgot about it. 160 Mbits/sec is about the maximum that the
Netgear(?) router could do with DOCSIS 3.0.

Another performance problem is your local wireless speed. If your
laptop tests faster with an ethernet connection, than with a wireless
connection, obviously the wireless is what's slowing you down. For
example, if you're using 802.11g only, your maximum download speed via
wireless will be about 25 Mbit/sec. If your cable modem can do 60
Mbits/sec, the "problem" is in the wireless link.

I've previously ranted on how to setup a local iperf3 server for
testing local network speeds, without the need for an online test
server or even an internet connection. It's interesting to see how
bad some network devices are when the speed is NOT limited by the
internet connection. Find your wireless router on this list for a
clue.
Wired download:
<http://www.smallnetbuilder.com/tools/charts/router/view>
2.4 GHz wireless download:
<http://www.smallnetbuilder.com/tools/charts/router/bar/111-2_4-ghz-dn-c>
Note that only a few of top wireless routers can do over 100 Mbits/sec
on 2.4GHz.

[Jeff Liebermann]

On Wed, 19 Aug 2015 03:09:23 +0000 (UTC), do...@93.usenet.us.com wrote:

>In alt.internet.wireless Jeff Liebermann <je...@cruzio.com> wrote:
>> I certainly suggested buying the SB6141 at Costco, but I don't believe
>
>I bought an SB6121, a month or two before it would have made sense to by a
>6141, but in any event, the Motorola/Arris anything is better than the
>Arris/Arris that Comcast provides.

Every time some company produces a decent modem, Motorola buys them.
The good stuff was made by Netopia and Cayman, both now Motorola
companies. However, there was quite a bit of absolute junk being
shipped during the transitions. Now that Motorola own Arris, I expect
more of the same until things settle down. Incidentally, some of the
Comcast "gateways" that I detest are made by Pace, which now owns
2-wire.
<http://customer.xfinity.com/help-and-support/internet/wireless-gateway-compare/>
I think TG and TC are Arris, SMC is SMC, DPC are Pace. The Gateway 3
drives seem decent (i.e. they do dual band). If you simply take
whatever Comcast is leasing, you're likely to get a Gateway 1 or 2
until the supply runs out. If you try to buy your own, the only one
of these on the approved modem list is a Gateway 1 (TG862G).

>I knew I hadn't paid an install fee, and forgot how that happened. I took
>the "$8/month" modem/router, because once upon a time, I had so much
>trouble with cable that they replaced the modem several times before they
>replaced the drop, and I didn't want to have that argument with customer
>owned equipment. But, after a month or two of horrible WiFi, I went with
>the 6121 and an Asus router.

Yep. I'm doing much of the same thing for my customers. The only
problem is if they order phone service from Comcast, I'm stuck with a
very small list of acceptable "telephony gateway" devices. The latest
irritation is trying to get an Arris TM722/TM822 activated. They're
both on the approved modem list:
<http://mydeviceinfo.comcast.net>
but not listed as "retail". Customers are buying perfectly legal and
brand new devices from various vendors, only to find that Comcast
claims that they can't be activated. There's a screwup somewhere. Of
course, the ONLY retail telephony gateway available is the Gateway 1
TG862G, which has the wireless problems you describe. I suppose that
the TM722 and TM822 will magically go back onto the approved retail
devices as soon as the stock of TG862G junk gateways is depleted.

>> Impressive. Much depends on who you get on the phone. I've heard of
>> hour long ordeals and have experienced 20 minute activation acrimony.
>
>Mine took a couple of minutes on the phone.

I did one yesterday that took about 20 minutes online total. However,
this was a new modem transplant for an existing customer. I plugged
in an SB6141, waited about 10 mins for things to settle down. I had a
computah plugged directly into the modem. I went to some random web
site and Comcast redirected me to the activation page. I gave it the
account number and associated phone number. It thrashed around for
about 10 minutes updating the modem firmware, rebooted twice, and
worked as expected. However, the old modem was still working so I
called Comcast support to make sure that the old (customer owned)
modem will disappear from the bill. I was blessed with a very
concientious support person, who answered every question except the
ones that I asked. I guess we'll have to wait for the bill to arrive.

>In a reply to another post, speedtest.comcast.net shows IPv4 and
>IPv6 numbers, speedtest.net only shows IPv4.

True. Since Windoze 10 seems to use IPv6 before trying IPv4, that's
important. I've had a few odd problems which I've avoided by just
turning off IPv6 in the router until things settle down and I have
some time to assign the blame.

[ceg]

On Wed, 19 Aug 2015 05:10:53 -0700, Jeff Liebermann wrote:

> Every time some company produces a decent modem, Motorola buys them.

:)

[Robert Green]

"Jeff Liebermann" <je...@cruzio.com> wrote in message

<stuff snipped>

> CATV means CAble TV.

It started out being Community Antenna (or Access) TV . . .

<<The abbreviation CATV is often used for cable television. It originally
stood for Community Access Television or Community Antenna Television, from
cable television's origins in 1948: in areas where over-the-air reception
was limited by distance from transmitters or mountainous terrain, large
"community antennas" were constructed, and cable was run from them to
individual homes. The origins of cable broadcasting are even older as radio
programming was distributed by cable in some European cities as far back as
1924.>>

https://en.wikipedia.org/wiki/Cable_television

--
Bobby G.

[Jeff Liebermann]

Ok. Thanks for the history. However, I'm correct that "CAble TV" is
the current acronym of choice.

Some anecdotal history. I got my start in cable with STV
(Subscription TV) in Smog Angeles in about 1966(?). Back then, the
acronym had already morphed into "CAble TV". Oddly, the wireless TV
[1] companies were also calling their stuff "Subscription TV". The
cable companies needed something to differentiate themselves from the
wireless companies, so they borrowed the CATV acronym. Notice how
nobody wanted to invent a new acronym as it was easier to just steal
one from the competition.

[1] Don't ask about this:
<http://802.11junk.com/jeffl/antennas/Misc/slides/subscription-TV-antenna.html>

[Robert Green]

"Jeff Liebermann" <je...@cruzio.com> wrote in message

news:fhsctatsthmja7d3b...@4ax.com...

> On Thu, 20 Aug 2015 09:11:25 -0400, "Robert Green"
> <robert_g...@yah00.com> wrote:
>
> >"Jeff Liebermann" <je...@cruzio.com> wrote in message
> >
> ><stuff snipped>
> >
> >> CATV means CAble TV.
> >
> >It started out being Community Antenna (or Access) TV . . .
> >
> ><<The abbreviation CATV is often used for cable television. It originally
> >stood for Community Access Television or Community Antenna Television,
from
> >cable television's origins in 1948: in areas where over-the-air reception
> >was limited by distance from transmitters or mountainous terrain, large
> >"community antennas" were constructed, and cable was run from them to
> >individual homes. The origins of cable broadcasting are even older as
radio
> >programming was distributed by cable in some European cities as far back
as
> >1924.>>
> >
> >https://en.wikipedia.org/wiki/Cable_television
>
> Ok. Thanks for the history. However, I'm correct that "CAble TV" is
> the current acronym of choice.

Indubitably. I just remember recently seeing the two acronyms and their
original meaning and thought I would share. (-"

> Some anecdotal history. I got my start in cable with STV
> (Subscription TV) in Smog Angeles in about 1966(?). Back then, the
> acronym had already morphed into "CAble TV". Oddly, the wireless TV
> [1] companies were also calling their stuff "Subscription TV". The
> cable companies needed something to differentiate themselves from the
> wireless companies, so they borrowed the CATV acronym. Notice how
> nobody wanted to invent a new acronym as it was easier to just steal
> one from the competition.

To that I say "Digital Versatile Disk!"

> [1] Don't ask about this:
>
<http://802.11junk.com/jeffl/antennas/Misc/slides/subscription-TV-antenna.ht
ml>

Ever see an early Saturday Night Live where (I believe) Dan Ackroyd and
Steven Martin are dressed as farmers looking off into the distance and
saying: "What the heck IS that dang thing?"

My question exactly. It looks like a roto-tiller for Lilliputians.

--
Bobby G.

[Jeff Liebermann]

On Fri, 21 Aug 2015 16:57:11 -0400, "Robert Green"

<robert_g...@yah00.com> wrote:

>"Jeff Liebermann" <je...@cruzio.com> wrote in message

>news:fhsctatsthmja7d3b...@4ax.com...

>To that I say "Digital Versatile Disk!"

That's what happens when acronyms are chosen by marketing or
management. The theory is that the acronym has to be clever, while
what it represents can be totally contorted and insane, because
everyone is only going to use the acronym. The company with the most
acronyms (and patents) wins.

>> [1] Don't ask about this:
>>
><http://802.11junk.com/jeffl/antennas/Misc/slides/subscription-TV-antenna.html>
>
>Ever see an early Saturday Night Live where (I believe) Dan Ackroyd and
>Steven Martin are dressed as farmers looking off into the distance and
>saying: "What the heck IS that dang thing?"
>
>My question exactly. It looks like a roto-tiller for Lilliputians.

Sigh. I told you not to ask, which proves that nobody listens to me.
It's an early version of a bootleg wireless subscription TV antenna
and receiver front end. It was favored by the Z-channel[1] wireless
TV pirates of the early 1970's in Smog Angeles because it had more
gain than the official antenna and therefore worked at longer
distances. Officially, it's a "disk yagi" antenna, which is roughly a
yagi TV antenna, using disks instead of rods. It's actually a very
good antenna that quite easy to design and build:
<http://www.idealantenas.com.br/ingles/produto/yagi%20disc19/yagidisc19_ing.htm>
Notice the lack of side and back lobes.

The uglier and stranger looking the antenna, the better it works.

Now, go away please. I just returned from Costco with a new
Chromebook and some computah goodies for me and I want to play.


[1] Extra credit to Jeff Angus for reminding me of the company name.
<http://www.amazon.com/Z-Channel-A-Magnificent-Obsession/dp/B000AQKV0I>

[Robert Green]

"Jeff Liebermann" <je...@cruzio.com> wrote in message

news:28dfta5a34240mbtp...@4ax.com...

> On Fri, 21 Aug 2015 16:57:11 -0400, "Robert Green"
> <robert_g...@yah00.com> wrote:
>
> >"Jeff Liebermann" <je...@cruzio.com> wrote in message
> >news:fhsctatsthmja7d3b...@4ax.com...
>
> >To that I say "Digital Versatile Disk!"
>
> That's what happens when acronyms are chosen by marketing or
> management. The theory is that the acronym has to be clever, while
> what it represents can be totally contorted and insane, because
> everyone is only going to use the acronym. The company with the most
> acronyms (and patents) wins.

Somehow the rule of "expand the acronym" for the first use in an article has
pretty much fallen by the wayside. I come across at least a few every day
that I have to look up because they're not as self-explanatory as the author
may have thought. To be fair, it might have been an editor that elided the
acronym expansion, but as far as I can tell, very few websites, newspapers
and RATV stations use editors anymore. )-:
FWIW, the Global Language Monitor also named the Most Confusing Tech Acronym
of 2012:


The winner is SOA (solutions oriented architecture).

I think it actually means "Shi+ Out of Acronyms."

> >> [1] Don't ask about this:
> >>
>
><http://802.11junk.com/jeffl/antennas/Misc/slides/subscription-TV-antenna.h
tml>
> >
> >Ever see an early Saturday Night Live where (I believe) Dan Ackroyd and
> >Steven Martin are dressed as farmers looking off into the distance and
> >saying: "What the heck IS that dang thing?"
> >
> >My question exactly. It looks like a roto-tiller for Lilliputians.
>
> Sigh. I told you not to ask, which proves that nobody listens to me.

I listened, I just didn't obey. (-: I don't obey anyone. You wouldn't
want to infringe on my personal freedom by expecting me to make an exception
for you, would you?

> It's an early version of a bootleg wireless subscription TV antenna
> and receiver front end. It was favored by the Z-channel[1] wireless
> TV pirates of the early 1970's in Smog Angeles because it had more
> gain than the official antenna and therefore worked at longer
> distances. Officially, it's a "disk yagi" antenna, which is roughly a
> yagi TV antenna, using disks instead of rods.

Ah, yagi, another word origin to look up. I didn't expect there would be
homework.

Alas, it's not Yaw Aligned Geosynchronous Inductor or any such thing:

A highly directional and selective shortwave antenna consisting of a
horizontal conductor of one or two dipoles connected with the receiver or
transmitter and of a set of nearly equal insulated dipoles parallel to and
on a level with the horizontal conductor.

That's what it is alright. Origin. 1940s: named after Hidetsugu Yagi
(1886-1976).

When I bought my first radial arm saw I went to Sears to buy a dado blade
kit and of course, there was an attractive young salesclerk working the
power tools register that day who clearly didn't know what a dado blade was
but probably thought it was something that sounded similar.

http://dictionary.reference.com/browse/dado is full of dadoes.

FTR, its origin is from 1655-65 and might be from Italian: die, cube,
pedestal, or perhaps an Arabic dad game, Now what, you might ask, is an
Arab dad game? Dunno. Google is not being helpful:

Angry Arab dad over card game - YouTube
www.youtube.com/watch?v=vCzmTXQ7UXg

Ah, Google. It's just not very good at such searches, still.

> It's actually a very good antenna that quite easy to design and build:
>
<http://www.idealantenas.com.br/ingles/produto/yagi%20disc19/yagidisc19_ing.
htm>
> Notice the lack of side and back lobes.

Upon first read I thought we were talking about physical lobes on the
antenna and then I realized you're talking about the polar graph of the
radiated power. I poked around Google but couldn't find out why this style
of antenna is side/rear lobe-lacking. I'll keep looking. I assume it's the
size and linear design that does it. Obviously I am not a radio geek but
rather a cross-post asylum seeker from AHR, which seems to have a terminal
nitwit troll infection.

> The uglier and stranger looking the antenna, the better it works.

Apparently. While the specs don't match the parabolic antenna I use for
Wi-Fi, the size of this yagi antenna and its reduced wind load certainly
have advantages.

> Now, go away please. I just returned from Costco with a new
> Chromebook and some computah goodies for me and I want to play.

Don't get me started on Chromebooks, Android, Stagefright and Google. My
favorite experience with the Chromebook was trying to get connectivity help.
No help available unless you're on line, no ability to get on line unless
you can get help with the various settings. A bit of a paradox.

Another fine experience for these aging eyes was to discovering how tiny the
icons are even on a large screen. Still haven't found a good way around
that. Also, no tool-tips but plenty of oddball things happen with
mouseovers sending me to pages I am pretty sure I didn't select.

Still, at $150 it beats the hell out of a lot of other options, has HDMI
output and sort of even works with my old PS-2 keyboard and Intellimouse
trackball using a USB to PS/2 adapter. Unfortunately, the trackball seems
to require three times as many revolutions to travel the same distance as it
does hooked into a PC. Can't find any settings that alter that behavior.
Solve that problem and I will publicly proclaim you as "hero patriae" (for a
day).

> [1] Extra credit to Jeff Angus for reminding me of the company name.
> <http://www.amazon.com/Z-Channel-A-Magnificent-Obsession/dp/B000AQKV0I>

Netflixers out there can rent it (as I just did - sounds good) at:

http://dvd.netflix.com/Movie/Z-Channel-A-Magnificent-Obsession/70022311

Now go play with your Chromebook which I've renamed my Crohnbook because it
gives me such a bellyache. Leaving in a positive note, the Crohnbook does
have a far more sensitive wi-fi card than many of my other wi-fi devices and
works in places the others won't.

I also understand the newest Chromebooks can run lots of older legacy
applications but don't know the details. I guess the industry has finally
realized that end users are getting quite reluctant to abandon something
that works for the next "greatest" thing in computing.

--
Bobby G.

[Ralph Mowery]

"Robert Green" <robert_g...@yah00.com> wrote in message news:mramg4>>

It's actually a very good antenna that quite easy to design and build:
>>
> <http://www.idealantenas.com.br/ingles/produto/yagi%20disc19/yagidisc19_ing.
> htm>
>> Notice the lack of side and back lobes.
>
> Upon first read I thought we were talking about physical lobes on the
> antenna and then I realized you're talking about the polar graph of the
> radiated power. I poked around Google but couldn't find out why this
> style
> of antenna is side/rear lobe-lacking. I'll keep looking. I assume it's
> the
> size and linear design that does it. Obviously I am not a radio geek but
> rather a cross-post asylum seeker from AHR, which seems to have a terminal
> nitwit troll infection.
>

Yes, the lobes are the often unwanted directions and ammount of signal
pickup/radiated of the antenna.

The Yagi type antennas can be designed for maximum gain or minimum side
lobes with decent gain. It all has to do with the spacing of the elements,
the number, and the lengths of them. The physical size (length of the
boom) does not have too much to do with the side lobes, but the longer it
is, the larger and narrower the main beam usually is.

If designed for maximum gain, it will usually have larger and more side
lobes. Sometimes the bandwidth (frequency range) will have a factor in the
lobes. OUtside of the design frequency range the lobes become very
pronounced and the main beam may be distorted and lots smaller.

[Robert Green]

"Ralph Mowery" <rmower...@earthlink.net> wrote in message
news:85GdndZMnsGvfUXI...@earthlink.com...

>
> "Robert Green" <robert_g...@yah00.com> wrote in message news:mramg4>>
> It's actually a very good antenna that quite easy to design and build:
> >>
> >
<http://www.idealantenas.com.br/ingles/produto/yagi%20disc19/yagidisc19_ing.
> > htm>
> >> Notice the lack of side and back lobes.
> >
> > Upon first read I thought we were talking about physical lobes on the
> > antenna and then I realized you're talking about the polar graph of the
> > radiated power. I poked around Google but couldn't find out why this
> > style
> > of antenna is side/rear lobe-lacking. I'll keep looking. I assume it's
> > the
> > size and linear design that does it. Obviously I am not a radio geek
but
> > rather a cross-post asylum seeker from AHR, which seems to have a
terminal
> > nitwit troll infection.
> >
>
> Yes, the lobes are the often unwanted directions and ammount of signal
> pickup/radiated of the antenna.
>
> The Yagi type antennas can be designed for maximum gain or minimum side
> lobes with decent gain. It all has to do with the spacing of the
elements,
> the number, and the lengths of them.

I assume that spacing corresponds in some way to the frequency you wish to
transmit or receive. Conceptually I still don't get why a rod with all
those disks "skewered" like a shish-kabob can focus the radio energy like a
parabolic dish.

> The physical size (length of the
> boom) does not have too much to do with the side lobes, but the longer it
> is, the larger and narrower the main beam usually is.
>
> If designed for maximum gain, it will usually have larger and more side
> lobes. Sometimes the bandwidth (frequency range) will have a factor in
the
> lobes. OUtside of the design frequency range the lobes become very
> pronounced and the main beam may be distorted and lots smaller.

Thanks for the info. I would like to be able to send and receive wifi from
a nearby free hot spot but the signal's not quite strong enough. A yagi
antenna looks like it could do the trick except that my tablet and netbook
don't have external antenna ports for wifi. I asked someone once and they
said an antenna connection would cause serious signal loss at wifi
frequencies but I don't know if that's true. I know that some people have
opened up various wireless devices and added antenna jacks, but I doubt I
will be doing that.

--
Bobby G.

[Ralph Mowery]

"Robert Green" <robert_g...@yah00.com> wrote in message

news:mrbr7a$8pm$2...@speranza.aioe.org...

> I assume that spacing corresponds in some way to the frequency you wish to
> transmit or receive. Conceptually I still don't get why a rod with all
> those disks "skewered" like a shish-kabob can focus the radio energy like
> a
> parabolic dish.
>
>

It would take a good diagram to explaine that. I doubt I can explain it in
words.

Try to picture a radio wave comming toward the antenna like a flat line. If
only a single antenna element the line would mostly pass the antenna and
only a small portion would be received. Now picture the antenna with many
elements and the line comming toward the antenna. As it hits the first
element, it bends like a rubber band, the more elements it passes the more
it bends. By the time it gets to the driven element (the one that actually
picks up the signal) the signal is bent into a long cigar type shape and
more of it will hit the driven element.

The length and the spacing of the elements are related to the frequency you
want to receive or transmitt.

There are several software programs that let you do computer modleing of the
antennas. Here is a link to one that has a free demo program that allows
for simple antenas.
http://www.eznec.com/

Most all of them are based on the same basic software program.

[do...@93.usenet.us.com]

In alt.internet.wireless Jeff Liebermann <je...@cruzio.com> wrote:
> Every time some company produces a decent modem, Motorola buys them.

I think that's far better than Motorola being purchased by Arris.

> made by Pace, which now owns 2-wire.

Shudder... 2-wire was so much fun when AT&T rolled DSL into Lake County.
They brought guys over from Santa Rosa to do some of the installs, and I
think they cleared out the Santa Rosa dumpsters of all the old 2-wire boxes.
(but we've chatted about that already.)

> problem is if they order phone service from Comcast,

I had phone service from Mediacom for a while. Now that portability works
so well, you can change VoIP telco providers as easily as gas stations.
I thought the CableCo voice was a "good thing" because it used a separate
channel, just like the XfinityWiFi. No impact on your modem speed, no
collisions, better voice.

> devices as soon as the stock of TG862G junk gateways is depleted.

If they reuse returned modems, that might take a long time.
Mediacom didn't even rebox the modems. They'd give you one in a plastic
bag.

> True. Since Windoze 10 seems to use IPv6 before trying IPv4, that's
> important. I've had a few odd problems which I've avoided by just
> turning off IPv6 in the router until things settle down and I have
> some time to assign the blame.

IPv6 was a real headache for my son's new XBoxOne. You couldn't disable
IPv6 in the TG862G, but it didn't work well, or XBox didn't, or something.
On Ubuntu, I see that connections to some of the major sites are IPv6.

[Jeff Liebermann]

On Tue, 1 Sep 2015 00:00:38 +0000 (UTC), do...@93.usenet.us.com wrote:

>In alt.internet.wireless Jeff Liebermann <je...@cruzio.com> wrote:
>> Every time some company produces a decent modem, Motorola buys them.
>
>I think that's far better than Motorola being purchased by Arris.

True, but Motorola is far from perfect. Every time they buy a modem
company, the first thing that happens is the pipeline is crammed full
of rejects, returns, refurbished, and junk modems for about 3 months.
I think the company they bought counted their backlog of repair jobs
as shippable inventory when the company was appraised, and Motorola
just shipped everything.

>Shudder... 2-wire was so much fun when AT&T rolled DSL into Lake County.
>They brought guys over from Santa Rosa to do some of the installs, and I
>think they cleared out the Santa Rosa dumpsters of all the old 2-wire boxes.
>(but we've chatted about that already.)

Yep. I saw the dumpster crammed full of failed and rejected 2-wire
power supplies delivered from the local AT&T yard. Why AT&T kept so
many dead power supplies around their yard will remain a mystery.

>> problem is if they order phone service from Comcast,

Well, I lied a little. While the Arris TM722G and TM822G telephony
modems are not listed as "retail" on the Comcast web site, Comcast
will activate them if you scream or complain loudly. They can be had
for $30 to $70 on eBay.

>I had phone service from Mediacom for a while. Now that portability works
>so well, you can change VoIP telco providers as easily as gas stations.
>I thought the CableCo voice was a "good thing" because it used a separate
>channel, just like the XfinityWiFi. No impact on your modem speed, no
>collisions, better voice.

Comcast really wants to be a phone provider (without all the telco
common carrier restrictions) and is doing their best to become the
carrier of choice. Of course, that doesn't include remembering to
supply backup batteries for the modems and gateways, but I'll forgive
them for trying to gouge and extra $35 from those that complain. If
they keep up such practices, they may achieve their goal of emulating
all the bad parts of AT&T.

The cable telephony modem, which uses a separate channel is the better
way to do phone for the reasons you indicate. Another advantage is
that the jitter on the telephone RF channel is quite good, while the
jitter on the Comcast data channel is variable, apparently depending
mostly on channel loading. I'm on Comcast but using Future-Nine for
VoIP on the data channel. There are plenty of dropouts and garble but
at $6.25/month, I won't complain (much).

>> devices as soon as the stock of TG862G junk gateways is depleted.

>If they reuse returned modems, that might take a long time.
>Mediacom didn't even rebox the modems. They'd give you one in a plastic
>bag.

I don't know what Comcast does with their returns.

>> True. Since Windoze 10 seems to use IPv6 before trying IPv4, that's
>> important. I've had a few odd problems which I've avoided by just
>> turning off IPv6 in the router until things settle down and I have
>> some time to assign the blame.
>
>IPv6 was a real headache for my son's new XBoxOne. You couldn't disable
>IPv6 in the TG862G, but it didn't work well, or XBox didn't, or something.
>On Ubuntu, I see that connections to some of the major sites are IPv6.

Chuckle. I should have predicted what is happening. Comcast cranked
up the speed for home users about 2 weeks ago. I'm seeing 90
Mbits/sec down and either 6 or 12 Mbits/sec up for both IPv4 and IPv6.
Both are obviously rate limited. However, that was 2 weeks ago. Now,
I'm seeing the same 90 Mbits/sec for IPv4, while IPv6 has dropped to
about 40 Mbit/sec. What seems to be happening is that the new Windoze
10 machines favor IPv6 if it's available. Comcast seems to have some
kind of IPv6 to IPv4 gateway for those web sites and servers that only
terminate with IPv4. The Windoze 10 machine favors IPv6 for
everything, so the gateway is probably getting swamped by the Windoze
10 traffic. I've again had to turn off IPv6 in some customers routers
to maintain performance as the congestions seems to produce some
dropped packets.

Tomorrow, I do a service call for a dialup customer. I wonder if I
even remember how to setup and use dialup internet?

[do...@93.usenet.us.com]

In alt.internet.wireless Jeff Liebermann <je...@cruzio.com> wrote:
> >In a reply to another post, speedtest.comcast.net shows IPv4 and
> >IPv6 numbers, speedtest.net only shows IPv4.

> turning off IPv6 in the router until things settle down and I have

I usually retest URLs that I post, just to make sure they are still live
(Jeff had a discussion with someone else who needed to take a little time
out on that issue, recently.)

I discovered that my home IPv6 wasn't IPv6-ing.
I used my set of URLs:
http://ipv6-test.com/ (look at the "api" tab)
http://test-ipv6.comcast.net/
http://speedtest.comcast.net/

hmmm...
Oh, right! Someone suggested that IPv6 has no stateful filters, and it
would be impossible to block inbound attacks.
Is that true?
I left "Enable Router Advertisement" turned off, which killed my external
IPv6 while I investigated, and I guess I never finished investigating.