Low Cost VOIP Providers

[miso]

If people knew the hidden taxes they paid, they would freak. On gas in
California, there are several taxes, and you end up being taxed on the
tax. I haven't seen this documented (the press doesn't have the time to
do anything other than rewrite press releases). There is a refinery tax.
I guess you would call that a wholesale tax. Since a lot of California
gas leaves the state, you can't blame Ca. for taxing the refiners.
However, they don't credit this tax back to those in California. Then
you pay a gas tax at the pump, plus a sales tax that includes tax on the
federal and state gas tax. The same gas gets piped to Nevada and sells
for much less than in California. The press says there are no refineries
in Nevada, but I found out that is not true. But there are no
significant refineries like you have in LA and the bay area.

Funny thing about Nevada refineries is you find them in the areas where
Nevada pumps oil. The Great Basin for example. So you have gas that
doesn't have a picoliter of OPEC oil in it, yet sells for the same price
as the rest of the state. I've done the Reno to Las Vegas run a few
times, and once in a while the podunk towns that get Nevada refined gas
don't get around to jacking up the price to match the rest of the state.

 

[Don Y]

On 8/17/2013 4:25 PM, miso wrote:

With FOSS, well linux distros, you can pick one that doesn't upgrade
much. FreeBSD for instance. On a web server, less is more. Because a
server spends all it's time on the public internet, every program is an
attack vector.

Yup. But you have to balance "infrequent updates" against (potentially)
"long periods of bugs-in-place".

I ran NetBSD from 0.8. Then, moved to FreeBSD at ~1.0 -- cuz the
NBSD camp was just too slow getting the features I wanted in place.
Then, *left* FBSD when they started pushing "an update every few
months" and returned to the longer update cycle for NetBSD.
Learning to live with workarounds for those bugs that weren't
yet fixed. (e.g., I could care less about KDE, OpenOffice, etc.)

Because updates come out of *my* pocket (time budget), I'm not keen
on wasting time chasing the latest and greatest set of half-baked
"improvements". Fix the *previous* improvements before releasing
a new set of buggy features on your "test subjects"... er, "customers"

Regarding routers, buy one that runs DDWRT or the other similar open
source firmware. That way, your old router has a path to upgrade.
Otherwise the manufacturer won't bother to upgrade your firmware because
they already made their money.

Power supplies do not have to be crap. They are crap because they
OEM/ODM farms them out to fly by night firms that in turn have them OEMd
by the worst of the Chinese assembly houses. This is true of your basic
wall wart. Not so much with the power supply for your notebook.

I've replaced my share of laptop "bricks". Most folks don't see
this because they replace their *laptop* often! :>

The same sort of attitude *should* be equally acceptable with
software, eh? "We'll have a new version out soon enough so why
bother testing/FIXING the old one?" (the laptop will be replaced
in a couple of years so why bother designing a power supply that
will last any longer than that?)

When I met someone who did QA at a software company, my reply was "Oh, I
didn't know software has QA. I figured they let the customers test it."
Amongst all the peddlers, Oracle has to be the worst at this.

Actually, I find much of the FOSS stuff to be really bad in this
regard. Which always amuses me!

The same folks who will COMPLAIN that the reason their 9-to-5
job generates such poor quality software is because "the boss
won't give us *time* to test things, thoroughly"!

Yet, when they put themselves in a situation where *they* can
set the release schedule (i.e., to *include* plenty of testing)
we get the exact same results! ("Hmmm... 'the boss' is no longer
in the equation yet the results are the same. What's the common
factor, here??" :> )

Granted, most folks find testing to be boring. And, many just
don't have the mindset to be able to challenge *their* own
code! I.e., they weren't able to think of these "unexpected
circumstances" when they were developing the algorithms so
they won't be able to think of them when *challenging* the
algorithms!

Forcing your users to test your wares is the basis of the
"release early, release often" snake oil. :<

 

[miso]

The only notebook power supply brick I ever destroyed was when I put it
on a modified sine wave inverter. Those things are just shit cubed.

I keep my notebooks way longer than most people because I dual boot and
don't want to go through the pain of setting things up. I kept the last
one for 8 years.

 

[Jeff Liebermann]

On Sat, 17 Aug 2013 19:53:24 -0700, miso <miso@sushi.com> wrote:

>Satellite VOIP will be terrible. Consider the latency.

Welcome to 2013.
<http://www.exede.com/voip>
<http://www.exede.com/blog/an-inside-look-at-exede-voice>
Wild Blue, Exede, ViaSat (same company) sell VoIP via Ka band
satellite internet.

If you watch the video, notice the 1 second pause between

I just happened to just finish running a test with a friend via Exede.
He was getting 9 Mbits/sec download speed via Speedtest.net. Not so
great on the upload at 1.5 Mbits/sec. Looks like I'll have to do a
dish alignment (again).

He just installed the VoIP service, so I did some testing. I had him
call me on the Exede VoIP phone and with Skype at the same time. I
wanted to see if the VoIP traffic had priority. Yes, it did by quite
a margin. My guess(tm) was that Skype was about 0.5 to 1 second
slower which suggests that Exede is giving their VoIP packets
priority. Both had about a noticeable delay with a little garble from
lost packets and compression. Quality was quite good and far from
your "terrible". It also seemed to be switching between full and half
duplex audio when there was no audio in one direction, which was
distracting, but tolerable. Skype had more dropouts, but that was to
be expected. I forgot to have him run jitter and video tests. Maybe
tomorrow.

The ATA is an InnoMobile MTA 6328-2Re.
<http://www.innomedia.com/products_sohores_mta6328re_features.shtml>
<http://www.innomedia.com/images/datasheets/MTA%206328-2Re%20DS.pdf>
Looks like it has native support for some low bitrate CODECs, which is
my guess(tm) what Exede is using. It also supports both SIP and MGCP.
No clue yet which one Exede is using.

The bad news... $30/month.


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

 

[Don Y]

On 8/17/2013 7:48 PM, miso wrote:

The only notebook power supply brick I ever destroyed was when I put it
on a modified sine wave inverter. Those things are just shit cubed.

I keep my notebooks way longer than most people because I dual boot and
don't want to go through the pain of setting things up. I kept the last
one for 8 years.

Two of my volunteer gigs involve fixing/refurbishing donated
kit of this sort. *Often*, laptops get scrapped solely because
the "brick" that came with it was toast (there is too much
variation among models and manufacturers to be able to piece
together brick A with laptop B). Note these aren't always
*old* laptops, either!

 

[miso]

I don't have the option of using Comcast as they don't service this
area. Cell data connections don't work reliably here either. In fact,
my options for digital connectivity were limited to satellite and
dial-up with a friend on satellite explaining that there isn't a lot of
difference sometimes. Until... someone who provides Internet off of
towers, not cell towers, but dedicated links. He uses a device in the
900 MHz band according to the info I found on it. Initially it didn't
do so well, but after a couple of weeks I have no complaints.

This sounds like WISP, though I find the links in the 2.4G range. I have
some 900MHz spread spectrum backhaul gear, but damn slow.

Satellite VOIP will be terrible. Consider the latency.

 

[Spehro Pefhany]

On Sat, 17 Aug 2013 22:24:43 -0700, the renowned Don Y
<this@isnotme.com> wrote:

On 8/17/2013 7:48 PM, miso wrote:
The only notebook power supply brick I ever destroyed was when I put it
on a modified sine wave inverter. Those things are just shit cubed.

I keep my notebooks way longer than most people because I dual boot and
don't want to go through the pain of setting things up. I kept the last
one for 8 years.

Two of my volunteer gigs involve fixing/refurbishing donated
kit of this sort. *Often*, laptops get scrapped solely because
the "brick" that came with it was toast (there is too much
variation among models and manufacturers to be able to piece
together brick A with laptop B). Note these aren't always
*old* laptops, either!

I've been able to get spare knock-off power supplies for all my
laptops readily, and cheaply, and all have had nice molded cables and
worked properly (so far) (there are some 3rd party ones "universal"
ones with adjustable voltage and a whack of connector adapters which
do NOT work properly with some laptops- Dells I think mostly). I like
to have adapters in multiple locations so forgetting them isn't an
issue. Mind you, I don't wait eight years before trying to get one,
but several years seems to be okay. Typically something like $15 each.

Of course if you're not permitted to spend any money at all in your
activities, then this doesn't work.


Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com

 

[rickman]

On 8/17/2013 7:40 PM, miso wrote:

Funny thing about Nevada refineries is you find them in the areas where
Nevada pumps oil. The Great Basin for example. So you have gas that
doesn't have a picoliter of OPEC oil in it, yet sells for the same price
as the rest of the state. I've done the Reno to Las Vegas run a few
times, and once in a while the podunk towns that get Nevada refined gas
don't get around to jacking up the price to match the rest of the state.

People often think the price of something is defined by what it cost to
make. The price of a commodity has only to do with the market, supply
and demand. Your places in Nevada that charge a little less have
nothing to do with their proximity to refineries or oil wells.

There is a section of central Virginia where the price of gas is some 15
or 20 cents lower than most places in the state. No special reason that
I can see. They are further from refineries and any natural resource
for oil than other areas. I expect it is the economics of the area.
They are a little less affluent than much of the state and when the
price of gas goes up a little, they cut back on usage more than other
areas I would expect.

Supply and demand... that's all it is. Ain't nobody in Nevada gas
stations doing you any favors.

--

Rick

 

[Don Y]

Hi Spehro!

On 8/18/2013 5:26 AM, Spehro Pefhany wrote:

On Sat, 17 Aug 2013 22:24:43 -0700, the renowned Don Y
this@isnotme.com> wrote:

Two of my volunteer gigs involve fixing/refurbishing donated
kit of this sort. *Often*, laptops get scrapped solely because
the "brick" that came with it was toast (there is too much
variation among models and manufacturers to be able to piece
together brick A with laptop B). Note these aren't always
*old* laptops, either!

I've been able to get spare knock-off power supplies for all my
laptops readily, and cheaply, and all have had nice molded cables and
worked properly (so far) (there are some 3rd party ones "universal"
ones with adjustable voltage and a whack of connector adapters which
do NOT work properly with some laptops- Dells I think mostly). I like
to have adapters in multiple locations so forgetting them isn't an
issue. Mind you, I don't wait eight years before trying to get one,
but several years seems to be okay. Typically something like $15 each.

Of course if you're not permitted to spend any money at all in your
activities, then this doesn't work.

That;s the problem with non-profits.

At one of them (with which I am no longer affiliated), we'd
receive between 10 and 100 donated laptops *every* week.
Just don't have the manpower ("volunteer-power"?) to test,
refurbish and distribute that many in that period of time.

So, it becomes a simple matter: anything that has "exotic" missing
components (CD/DVD drives, plastic case parts, etc.) goes in the
"toss" pile -- unless it looks to be really *cherry*! Anything
with broken or missing components of which we don't likely have a
*spare* (stolen from some other piece of kit at some earlier point
in time) also goes to the tip.

You're not going to spend a few hours testing, refurbishing, cleaning,
installing software so you can give away the final "product" *AND*
have to shell out $$$ for a new power supply, battery, etc.

[Folks would always grumble: "The battery doesn't work! What
good is a laptop if the battery doesn't hold a charge?" "Well,
you can buy a new battery!" "But they cost $80!" "So, what
you're saying is you want *us* to spend that $80 for the
laptop that we've just GIVEN to you??" (sigh)]

I've rebuilt power supplies, battery packs, etc. -- but not for
items that will be "coming from" a non-profit. E.g., if I am
given a device with a bad/missing power supply, I'll first test
it with a lab supply (I've got a few old HP digitally programmable
lab supplies that come in handy here). Once I know what state of
(dis)repair it is in, I may choose to scrap it *or* rebuild it.

For a power supply, often just cannabilizing some *other* power
supply (same or greater capabilities) and just marrying the right
cable/connector to it. (I'm not fond of rebuilding encapsulated
supplies because they are tricky to get apart and back together
while still *looking* "good" :> )

C's laptop came to us in a similar way: a neighbor gave me one
of their kids' laptops: "it doesn't work" (bad power supply).
I tested it and shoved it in a desk drawer. Almost a year later,
another friend dropped off the exact same model -- *with* a
power supply (but with some cosmetic damage to the case). It
doesn't take a rocket scientist to piece together one from two! :>

One of *my* laptops came from a friend at a three-letter company
in feenigs. Another "bad power supply" story. I just found a
similar supply and doctored the plug (barrel connector). While
it's not a convenient laptop to lug around (heavy as all sin!),
it's reasonably responsive and has a good set of I/Os for the
types of devices with which I typically have to interface
(e.g., a *real* serial port)

<shrug> You'd be heartbroken to see how much stuff gets
"discarded" -- often because people just want new or don't
want to bother figuring out what's wrong with what they have!
While much of it can be diverted from landfills, it's still
an incredibly inefficient use of resources (e.g., I think
a desktop PC has a "recycle" value of < $10)

 

[rickman]

On 8/17/2013 6:44 PM, Jeff Liebermann wrote:

On Sat, 17 Aug 2013 15:00:16 -0400, rickman<gnuarm@gmail.com> wrote:

On 8/17/2013 1:54 PM, Jeff Liebermann wrote:
On Sat, 17 Aug 2013 03:46:12 -0400, rickman<gnuarm@gmail.com> wrote:

Measure the jitter, especially when there's lots of traffic. Setup
QoS. Then play.

Yeah well, I don't know how to do that.

http://myspeed.visualware.com/indexvoip.php

That link was posted before. It uses Java and I don't have Java on my
browser.

I couldn't find anything that does not use Java. You can measure
jitter using Wireshark to sniff VoIP traffic. I've never tried it:
http://wiki.wireshark.org/RTP_statistics
http://nms.lcs.mit.edu/~hari/papers/CS294/paper/node5.html
http://toncar.cz/Tutorials/VoIP/VoIP_Basics_Jitter.html

Another method is to us JPerf (xjPerf). Unfortunately, the preferred
graphical user interface in Jperf requires Java. However, iPerf is
command line and should work. There are some notes and sample outputs
for jitter at:
http://openmaniak.com/iperf.php> (search for jitter).
http://blog.nexcess.net/2011/11/04/testing-network-performance-and-throughput-with-iperf/
See Scenario #3.

Note that you do NOT need to sniff a VoIP UDP packet stream, as in the
Wireshark test. There are Mac, Windoze, and Linux versions of Iperf
available.

What you should do is setup a typical computah in Iperf server mode
somewhere on the internet and run the client at your end. You won't
have the benefits of QoS unless you use the standard SIP port numbers.
Don't do that on a corporate LAN or where VoIP sessions are active or
you will make some new enemies. If you can't get anyone to cooperate,
bug me and I'll throw together a test server in the office (time
permitting).

Ok, I'm not sure how I would set up such a server. I may try it
sometime if I can get a friend to cooperate. I do have a spare computer
or two. But even if I get the server set up, how would you know if
performance problems were at my end vs. the server end? We *are*
talking about someone else's house for the server and why would their
service be any better than mine?

Maybe I should ask you to set me up with a provider and an adapter?

I'm on vacation and expect to be buried in paying work for at least 2
weeks. I probably won't have time.

Do you understand what I want?

About 80% yes. The big mysteries are the characteristics of the
900MHz link, potential congestion due to limited bandwidth, and
coexistence with other applications. The last one is critical as I'm
finding that bandwidth hogs, such Bitorrent and DropBox are real
problems. We would need to talk on the phone first.

What would you charge me to make it work?

Dunno. My shop rate is $75/hr for bench work. This would probably
turn into building and testing something at my office, and ship
everything to you via UPS. I don't like doing it like that because
there's always an oversight. Please switch to email if you want to
continue this.

Plan B might be easier. We talk. I recommend a service provider that
also delivers hardware. You sign up for a 30 day trial and just see
if it works for you. There will be install and network complications,
but I can hand them using remote control software (Teamviewer).

Right now I am leaning to trying the NetTalk Duo setup. I was all set
to get their $65 wifi unit. Their deal is you buy the unit with a year
of service. Then extending for additional years is $30... plus those
taxes and "fees". But then I found Walmart (who seems to have a deal
with NetTalk) doesn't carry the wifi unit anymore. Still, the price is
right and money back guarantee. What do I have to lose?

The reviews of Ooma are not good and Vonage is overpriced for my needs
at $30 a month plus I think taxes.

Careful with VoIP reviews. Users have the expectation that VoIP will
be "just like POTS". You can come close, but not in every area. For
example, I once read a scathing denunciation of some VoIP vendor by a
user complaining that support was not open after midnight and that
nobody offered to drive several hundred miles to her house to fix her
phones. Also, one line reviews can usually be discarded.

If Ooma and Vonage are inadequate, there are smaller providers that
will also sell you hardware. I'm not thrilled with those that supply
VoIP routers. That solves the QoS setup and port forwarding problems,
but also creates coexistence problem with other common devices. The
problem is that I don't know whom else to recommend. I don't do many
appliance type VoIP installation.

I wouldn't say "inadequate". I would say, "too expensive". $30 a month
is twice what I pay for a phone line. I've heard a lot about how
inexpensive Internet phones should be including the links you have
pointed me to. So I prefer to pay less rather than more.

Why is it no one can tell you in
advance what the taxes are?

Because the courts decided that taxes should be calculated by the
customers address, and not the vendors. So if you deal with a
national vendor, they have to calculate the taxes to include all the
state and local taxes required at the buyers location. They also have
to pay all those taxes to the various taxing agencies, but you don't
see any of that.

That doesn't mean they don't know what the taxes are! They have to
charge them, so they obviously know. However they calculate them when
they bill you, that can be done during sales as well. They just don't
*want* to. The "fees" on the other hand, are not taxes on the consumer
at all. They are taxes that get passed on to the consumer. Totally
different. Like asking a renter to pay the property taxes on the house
they are renting.

I don't actually have a "decor" so I don't need to match any colors.
Does that make you feel better?

No. You probably have an "ambience", "theme", or "style" to replace
your missing "decor". Either way it has to match. It's not really a
trivial problem.

LOL! You greatly overestimate my decorating capabilities. Right now I
have one corner of the living room filled with a computer, an
oscilloscope, other equipment and a cheap bookcase with the Internet
modem cable tied to a lamp with no shade and the wifi router nearby.
Actually in an attempt to get the best signal the lamp is sitting on a
board sticking out from the shelf with a brick to adjust the height.
Yes, we definitely need to match that and it may be difficult. If the
LEDs on the VOIP box aren't the right shade of red and green it just
*won't* fit in...

Trying to find a place to put the black boxes and
wires is a major part of an installation. Vendors don't make it any
easier by putting smooth rounded top on their boxes, so nothing can
placed on top of their box. At least the hi-fi and home theater
designers have recognized the benefits of stacking boxes, but that
hasn't come to the attention of network appliance designers quite yet.

Incidentally, I usually carry at least 2 new routers in my vehicle.
One is black (Linksys) while the other is white (Netgear). I've had
to amuse myself for many minutes while the homeowner considers the
relative merits of each color.

Yeah, you are clearly overthinking my situation.

Yeah, but Future Nine requires you to be knowledgeable in the operation
of the system and I'm not. That one in particular I can't even figure
out what they charge and what I have to buy. I bet you can't either
unless you read it off your statement. Their web site is horrible.

Correct on all counts. Being the lowest priced vendor has its costs.

I don't think "Just some buzzwords and understanding of how things work"
is sufficient for picking a vendor.

True. But I think you got my point. You need to at least know
something about the technology to ask the right questions.

My problem is the vendors don't
explain their service well enough for *anyone* to know what they offer
and what they charge.

True. I must admit that even I don't understand some rate schedules.

For those that roll their own, no explanation is necessary.
For those that don't roll their own, no explanation is possible.

For example, I still don't know if I have to pay
for E911 service with Callcentric after two emails from them.

http://www.callcentric.com/faq/23
http://www.callcentric.com/911/
Kinda looks like E911 service is mandatory, bundled, and billed as
part of the monthly charges.
"You must register with Callcentric the physical location where
you will utilize your service for each Callcentric account.
(...)
You will register your initial location of use when you purchase
a rate plan or a phone number."

Was that a yes or a no?

Yeah, not very clear at all. It kinda reads like it is assumed that
everyone will get E911 service.

Even if they don't have outbound service? How do you use it?

They qualify this with "and have at least one
service under your account"... what does that mean?

An account is a login and a monthly invoice.
A service is a single phone number. You can have many services or
phone numbers under a single account. Not the best choice words, but
good enough methinks.

Outbound only service has no phone number, right? Inbound only service
can't use 911, right?

Why would I have an
account without a service?

The usual way such things are done is to first create an account (so
they can send you a bill) and then add services to the account. It's
much the same with some cellular MVNO's such as PagePlus.

So if I get incoming calls only I still have
to pay for E911 even though I can't call 911???

If you only subscribe to call-in service, it is my understanding that
certain outgoing numbers still function, such as 611 and 911. However,
I have to check on that first as it varies with vendor. The
assumption is that 911 *ALWAYS* works, even on POTS lines used
exclusively for DSL, unactivated cell phones, and some pay phones.

I was aware that cell phones could reach 911, I didn't know that about
POTS. So even if you are turned off for non-payment you can dial 911?
Interesting. That would solve my 911 problem. I had a POTS here many
years ago. I'm pretty sure I don't get dialtone though.

There is no need to consider your system since I won't be using anything
remotely like your system.

I thought you wanted a shopping list of features that you might want
to consider adding or using?

But none of this stuff seem important.

How about considering *my* system? One
adapter and a POTS phone, maybe two since that seems to be a fairly
standard service provided at no extra cost.

Boring... Ok, get a Linksys PAP2-NA. It's cheap, easy to configure,
plenty of versatility, lots of online help, small, portable, and
supported by literally all the VoIP vendors. About $35.
http://www.callcentric.com/support/device/linksys/pap2
Make sure it's a real PAP2-NA as you don't want one that's vendor
locked to a service provider. Configure it and sign up for an account
with any vendor that doesn't have annual contracts. Try it for a
month and see what happens.

Isn't that an out of date unit? One of the links I was given for
equipment indicated a replacement unit, SPA2102 or something similar.

See anything useful?

Mostly no except for the three locations. I might find a use for
softphone when I am not at any of my locations but am using the laptop
on the internet.

You might look into Callcentric "extensions" which are much the same
as simultaneous ring:
http://www.callcentric.com/faq/35

Thanks.

I need a phone replacement, that's all.

Resistance is futile. You will be educated (in VoIP).

Not this way. I expect it will be even more painful.

No pain, no gain. The only difference between drinking water and
drowning is the rate at which the water is consumed. Try going a bit
slower on the VoIP learning exercise and it may be somewhat less
painful.

Ok

The cell service here is poor
and I am spending more on minutes that I would like.

I've installed a few of these for such situations. If the service is
poor, they will help. If the service is non-existent, it won't help:
http://www.zboost.com
http://www.zboost.com/solutions-products/home-office/yx545.html
Yagi or panel antenna on the roof. RG-6a/u coax to the box. About
50ft indoor range. Photo of the insides:
http://802.11junk.com/jeffl/zBoost/

Wow! $400 times two. My roommate has ATT and I'm on Verizon, so we
would need nearly $1000 worth of equipment to improve the signal.

Nope. Only one device is needed. The one I mentioned does 800/1900
Mhz on all the service providers except Nextel. However, it will not
do 4G data on 700/1700/2100/etc. You can find them on eBay for about
$240.

Still, $400 is a bit steep. Once I get VOIP I'll be happy. Everything
else he does with his iPhone can be done through wifi.

I
think VOIP is a better solution. Cell phone suck even when they get a
good signal in my opinion. On the other hand, my roommate lives on his
iPhone.

Show the zBoost page to your roommate and maybe he'll buy one and let
you use it.

I bet not...

--

Rick

 

[Don Y]

Hi Rick,

On 8/18/2013 11:09 AM, rickman wrote:

On 8/17/2013 7:40 PM, miso wrote:

Funny thing about Nevada refineries is you find them in the areas where
Nevada pumps oil. The Great Basin for example. So you have gas that
doesn't have a picoliter of OPEC oil in it, yet sells for the same price
as the rest of the state. I've done the Reno to Las Vegas run a few
times, and once in a while the podunk towns that get Nevada refined gas
don't get around to jacking up the price to match the rest of the state.

People often think the price of something is defined by what it cost to
make. The price of a commodity has only to do with the market, supply
and demand. Your places in Nevada that charge a little less have
nothing to do with their proximity to refineries or oil wells.

There is a section of central Virginia where the price of gas is some 15
or 20 cents lower than most places in the state. No special reason that
I can see. They are further from refineries and any natural resource
for oil than other areas. I expect it is the economics of the area.
They are a little less affluent than much of the state and when the
price of gas goes up a little, they cut back on usage more than other
areas I would expect.

Supply and demand... that's all it is. Ain't nobody in Nevada gas
stations doing you any favors.

Actually, not even that. *Perceived* value. What people can
be made to *expect* to pay for something.

Porter's _The Price of Everything_ is an interesting read. Not
just tangible things (why loss-leaders work, etc.) but also
intangible things: e.g., how much does it "cost" to believe a
certain *faith*? Why increasing a "cost" can lead to *fiercer*
"loyalty" to that product? etc.

Amusing to see how much MORE folks peddling "stuff" to us know
about why we buy and what we will spend then *we* appear to!
(and how that can be used to manipulate your "purchases")

 

[Spehro Pefhany]

On Sun, 18 Aug 2013 15:15:45 -0700, the renowned Don Y
<this@isnotme.com> wrote:

Hi Rick,

On 8/18/2013 11:09 AM, rickman wrote:
On 8/17/2013 7:40 PM, miso wrote:

Funny thing about Nevada refineries is you find them in the areas where
Nevada pumps oil. The Great Basin for example. So you have gas that
doesn't have a picoliter of OPEC oil in it, yet sells for the same price
as the rest of the state. I've done the Reno to Las Vegas run a few
times, and once in a while the podunk towns that get Nevada refined gas
don't get around to jacking up the price to match the rest of the state.

People often think the price of something is defined by what it cost to
make. The price of a commodity has only to do with the market, supply
and demand. Your places in Nevada that charge a little less have
nothing to do with their proximity to refineries or oil wells.

There is a section of central Virginia where the price of gas is some 15
or 20 cents lower than most places in the state. No special reason that
I can see. They are further from refineries and any natural resource
for oil than other areas. I expect it is the economics of the area.
They are a little less affluent than much of the state and when the
price of gas goes up a little, they cut back on usage more than other
areas I would expect.

Supply and demand... that's all it is. Ain't nobody in Nevada gas
stations doing you any favors.

Actually, not even that. *Perceived* value. What people can
be made to *expect* to pay for something.

Porter's _The Price of Everything_ is an interesting read. Not
just tangible things (why loss-leaders work, etc.) but also
intangible things: e.g., how much does it "cost" to believe a
certain *faith*? Why increasing a "cost" can lead to *fiercer*
"loyalty" to that product? etc.

Amusing to see how much MORE folks peddling "stuff" to us know
about why we buy and what we will spend then *we* appear to!
(and how that can be used to manipulate your "purchases")

CBC radio has a good series available as podcasts called "Age of
Persuasion".. now "Under the Influence". Might have been on NPR.

With the clever use of Big Data (especially search terms, web surfing
and other viewing habits), a company could become aware of a potential
need you have LONG before you are (probably longer in the case of
non-computer literate people). A government could also track what you
are going to be thinking before you know what you are going to think.

Just finished "Shadow Factory" by James Bamford .. interesting stuff..
and most of the claims have proven to be right on the money.


Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com

 

[Jeff Liebermann]

On Sun, 18 Aug 2013 17:02:59 -0400, rickman <gnuarm@gmail.com> wrote:

http://openmaniak.com/iperf.php> (search for jitter).
http://blog.nexcess.net/2011/11/04/testing-network-performance-and-throughput-with-iperf/
See Scenario #3.

Ok, I'm not sure how I would set up such a server. I may try it
sometime if I can get a friend to cooperate. I do have a spare computer
or two.

Before you try it over the internet, do it locally. Plug a good fast
computer into your ethernet switch as a server. Go to the
dos|shell|cmd prompt and run:
iperf -s
For now, we'll use the defaults and TCP instead of UDP. iPerf will
display the port number. You'll need to run ifconfig or ipconfig to
get the IP address of the server.

Now, plug another machine into the ethernet switch as a client. Go to
the dos|shell|cmd prompt and run:
iperf -c IP_address_of_server

Get some performance numbers and see if they make sense.

Once that's working, try it via wireless, via the internet, or any
other tests that might be of interest. To get jitter,.on the server
side:
iperf -s -u -i 1
and the client side:
iperf -c Server_IP_address -u -b 10m
The -b option allows you to allocate the maximum bandwidth, which in
this case in 10 Mbits/sec.

But even if I get the server set up, how would you know if
performance problems were at my end vs. the server end?

Well, I do that by looking at graphs produced by the Java version
which you seem reluctant to run. You can probably determine which end
is a problem by introducing a 3rd internet connection to the test. Try
the bidirectional tests between each of the three locations. If two
paths are disgusting, while the 3rd is just fine, the endpoint that
has the two disgusting paths is at fault.

If you find all this to be too much effort, I suggest you fire up a
SIP softphone client on one of your computahs, and get someone else to
do the same. Then make a direct connection between the two computahs.
(This is a bit tricky so feel free to bug me for details). If that
works, you have a chance with VoIP. If that fails because your 900
MHz WISP connection is infested with jitter, then you'll be the first
to know before you spend any cash.

We *are*
talking about someone else's house for the server and why would their
service be any better than mine?

We're not trying to troubleshoot anything at this point. We're trying
to determine if your 900 MHz WISP provider is capable of providing
useable VoIP service. The tests simulate a VoIP connection and
measure jitter, lost packets, and otto order packets. If any of these
are not sufficient for VoIP service, you will be the first to know. I
think it's a fair assumption that almost any wired broadband internet
connection will be better than a 900 MHz wireless link.

Right now I am leaning to trying the NetTalk Duo setup. I was all set
to get their $65 wifi unit. Their deal is you buy the unit with a year
of service. Then extending for additional years is $30... plus those
taxes and "fees". But then I found Walmart (who seems to have a deal
with NetTalk) doesn't carry the wifi unit anymore. Still, the price is
right and money back guarantee. What do I have to lose?

You can only lose your time and money. I was trying figure out a way
to test your WISP connection to make sure it will work for VoIP. Then,
I was trying to figure out how to get any flavor of service without
spending much time and money. That's the $35 PAP2 ATA. If you want
to commit to some service for a year just to find that your WISP
service is inadequate, feel free to burn the time and money.

I wouldn't say "inadequate". I would say, "too expensive". $30 a month
is twice what I pay for a phone line. I've heard a lot about how
inexpensive Internet phones should be including the links you have
pointed me to. So I prefer to pay less rather than more.

I pay about $26/month for my AT&T POTS line, with measured rate and no
long distance service. At $15/month, you're probably on Lifeline
service, which is effectively subsidized home phone service. Please
find a better comparison.

Yeah, not very clear at all. It kinda reads like it is assumed that
everyone will get E911 service.

Even if they don't have outbound service? How do you use it?

Dunno. Since Callcentric sells inbound and outbound service
separately, it's conceivable that they only include E911 service if
someone purchases outbound service.

>Outbound only service has no phone number, right?

Outbound only service has a phone number. For CallCentric is
1777-Your-CCID. CCID is your CallCentric user ID. You can't call
that number from the PSTN (public switched telephone network) but you
can call it from another SIP phone:
<http://www.callcentric.com/faq/4/130>

Inbound only service
can't use 911, right?

<http://www.callcentric.com/faq/23#159>
The $1.50 fee is NOT included in the following services:
Pay Per Call outgoing service
Any incoming service
Under these products you WILL be billed for 911 separately,
at $1.50/month plus a one time $1.50 setup fee, if you state
that your are in the US or Canada.
The way I read this is that you can make 911 calls with incoming only
service.

I was aware that cell phones could reach 911, I didn't know that about
POTS. So even if you are turned off for non-payment you can dial 911?

Dunno. I've never had the experience. However, it's my understanding
that if you have dialtone, you can dial 911.

Interesting. That would solve my 911 problem. I had a POTS here many
years ago. I'm pretty sure I don't get dialtone though.

Without dialtone, it won't work. This might help:
<http://www.consumerreports.org/cro/news/2009/05/update-about-911-and-disconnected-landlines/index.htm>

Boring... Ok, get a Linksys PAP2-NA. It's cheap, easy to configure,
plenty of versatility, lots of online help, small, portable, and
supported by literally all the VoIP vendors. About $35.
http://www.callcentric.com/support/device/linksys/pap2

Isn't that an out of date unit?

Yep. There are plenty of better and later devices. However, the
PAP2-NA is probably the most widely used and widely supported device.
You should not have any problems using it on any VoIP system. The
only gotcha is that eBay is full of counterfeit PAP2 boxes.
<http://www.voipfan.net/other/fakepap2.php>

One of the links I was given for
equipment indicated a replacement unit, SPA2102 or something similar.

That's a later unit. There's also an SPA3102 which supports fail over
to POTS. Lots of options. Incidentally, I just ordered another 4
line SPA941 desk phone because I got tried of only having one line on
my current SPA921. More than one line is really handy.

You might look into Callcentric "extensions" which are much the same
as simultaneous ring:
http://www.callcentric.com/faq/35

Thanks.

Incidentally, CallCentric recently added some new features, some of
which involve extensions.
<http://www.dslreports.com/forum/r28535967-CallCentric-CallCentric-announces-more-new-features>

Still, $400 is a bit steep. Once I get VOIP I'll be happy. Everything
else he does with his iPhone can be done through wifi.

So, put a SIP softphone client in his iPhone and be done with it. One
device does it all.
<https://www.counterpath.com/softphone-products.html>
Try a test call and see if the 900 MHz WISP link can handle it.

Show the zBoost page to your roommate and maybe he'll buy one and let
you use it.

I bet not...

Assuming AT&T 4G data and an iPhone 4S, he's probably paying about
$80/month for the service (and subsidized phone). That's almost
$1,000/year. Somehow, $400 doesn't seem to much any more.

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

 

[Don Y]

Hi Spehro,

On 8/18/2013 4:33 PM, Spehro Pefhany wrote:

On Sun, 18 Aug 2013 15:15:45 -0700, the renowned Don Y
this@isnotme.com> wrote:

Actually, not even that. *Perceived* value. What people can
be made to *expect* to pay for something.

Porter's _The Price of Everything_ is an interesting read. Not
just tangible things (why loss-leaders work, etc.) but also
intangible things: e.g., how much does it "cost" to believe a
certain *faith*? Why increasing a "cost" can lead to *fiercer*
"loyalty" to that product? etc.

Amusing to see how much MORE folks peddling "stuff" to us know
about why we buy and what we will spend then *we* appear to!
(and how that can be used to manipulate your "purchases")

CBC radio has a good series available as podcasts called "Age of
Persuasion".. now "Under the Influence". Might have been on NPR.

I'll have a look.

With the clever use of Big Data (especially search terms, web surfing
and other viewing habits), a company could become aware of a potential
need you have LONG before you are (probably longer in the case of
non-computer literate people).

<http://www.forbes.com/sites/kashmirhill/2012/02/16/how-target-figured-out-a-teen-girl-was-pregnant-before-her-father-did/>

A government could also track what you
are going to be thinking before you know what you are going to think.

Yup. Can't "be sure" of a particular outcome. But, with big
enough datasets, they can be "reasonably sure" -- for large values
of "reasonably"!

E.g., solely looking at "metadata", you could notice outliers in
usage: "Gee, how come this <unknown> user has such *low* usage"
(e.g., a prepaid cell phone being used once every few months)

Or, in my case, why *don't* we see lots of data tied to his "ID"?
(social media, credit card purchases, etc.)

Looking at who and how many "friends" you claim (claim you)
can be a predictor of your behaviors, risks, etc. (e.g.,
facebook should be in the RISK ASSESSMENT business, not
peddling advertisements!)

Here, I have serious concerns over how to prevent my automation
system's observations of your "private" behaviors from being
exploited. "Hmmm... he/she eats a lot of sweets, appears to have
a voracious thirst -- perhaps diabetic??" "Lots of trips to
the bathroom during the night -- prostate problems?" etc.

[Note: I'm just playing with stereotypes, here. With access to
*lots* of data, I'm sure someone could notice: "people who eat
beans on tuesdays and watch lots of westerns are more likely
to own a firearm", etc.]

Just finished "Shadow Factory" by James Bamford .. interesting stuff..
and most of the claims have proven to be right on the money.

 

[rickman]

On 8/18/2013 8:47 PM, Jeff Liebermann wrote:

On Sun, 18 Aug 2013 17:02:59 -0400, rickman<gnuarm@gmail.com> wrote:

http://openmaniak.com/iperf.php> (search for jitter).
http://blog.nexcess.net/2011/11/04/testing-network-performance-and-throughput-with-iperf/
See Scenario #3.

Ok, I'm not sure how I would set up such a server. I may try it
sometime if I can get a friend to cooperate. I do have a spare computer
or two.

Before you try it over the internet, do it locally. Plug a good fast
computer into your ethernet switch as a server. Go to the
dos|shell|cmd prompt and run:
iperf -s
For now, we'll use the defaults and TCP instead of UDP. iPerf will
display the port number. You'll need to run ifconfig or ipconfig to
get the IP address of the server.

Now, plug another machine into the ethernet switch as a client. Go to
the dos|shell|cmd prompt and run:
iperf -c IP_address_of_server

Get some performance numbers and see if they make sense.

Once that's working, try it via wireless, via the internet, or any
other tests that might be of interest. To get jitter,.on the server
side:
iperf -s -u -i 1
and the client side:
iperf -c Server_IP_address -u -b 10m
The -b option allows you to allocate the maximum bandwidth, which in
this case in 10 Mbits/sec.

But even if I get the server set up, how would you know if
performance problems were at my end vs. the server end?

Well, I do that by looking at graphs produced by the Java version
which you seem reluctant to run. You can probably determine which end
is a problem by introducing a 3rd internet connection to the test. Try
the bidirectional tests between each of the three locations. If two
paths are disgusting, while the 3rd is just fine, the endpoint that
has the two disgusting paths is at fault.

If you find all this to be too much effort, I suggest you fire up a
SIP softphone client on one of your computahs, and get someone else to
do the same. Then make a direct connection between the two computahs.
(This is a bit tricky so feel free to bug me for details). If that
works, you have a chance with VoIP. If that fails because your 900
MHz WISP connection is infested with jitter, then you'll be the first
to know before you spend any cash.

Uh, I think about 10 posts ago I said a friend had already tried her
VOIP phone on my Internet connection and it seemed to work fine. So all
of the above jitter tests, etc are overkill if the connection passes the
acid test, no?

We *are*
talking about someone else's house for the server and why would their
service be any better than mine?

We're not trying to troubleshoot anything at this point. We're trying
to determine if your 900 MHz WISP provider is capable of providing
useable VoIP service. The tests simulate a VoIP connection and
measure jitter, lost packets, and otto order packets. If any of these
are not sufficient for VoIP service, you will be the first to know. I
think it's a fair assumption that almost any wired broadband internet
connection will be better than a 900 MHz wireless link.

You used the WISP term which is the name the provider uses for this and
in fact the ISP company name is iWisp. I take it this is a generic
term? I thought it was a brand name. Is there just one major brand of
products doing this function?

Funny note... when the guy was here hooking up a test he asked me for an
email address. Having my own URL I make up email addresses that
correspond to the vendor I am giving it out to, this allows me to track
the source of spam. I gave him iwisp.net@ my URL and he gave me a
really funny look like I was trying to pull something. I found that
very funny... as in ha, ha funny.

Right now I am leaning to trying the NetTalk Duo setup. I was all set
to get their $65 wifi unit. Their deal is you buy the unit with a year
of service. Then extending for additional years is $30... plus those
taxes and "fees". But then I found Walmart (who seems to have a deal
with NetTalk) doesn't carry the wifi unit anymore. Still, the price is
right and money back guarantee. What do I have to lose?

You can only lose your time and money. I was trying figure out a way
to test your WISP connection to make sure it will work for VoIP. Then,
I was trying to figure out how to get any flavor of service without
spending much time and money. That's the $35 PAP2 ATA. If you want
to commit to some service for a year just to find that your WISP
service is inadequate, feel free to burn the time and money.

The money issue is moot, they have a money back guarantee. I think the
tests will be more time and effort than just trying a system to see how
well it works. I had tried to find Internet reviews, but I think you
think these are not of much value and I am not going to disagree with
you. Every one I researched had a number of happy users (otherwise they
would be out of business quick, no?) and a number of unhappy users, most
of who seemed to have valid complaints. So I have to assume that the
products work for most, but not all and more important is how the
company deals with it. No one complained that NetTalk didn't have
reasonable support.

I wouldn't say "inadequate". I would say, "too expensive". $30 a month
is twice what I pay for a phone line. I've heard a lot about how
inexpensive Internet phones should be including the links you have
pointed me to. So I prefer to pay less rather than more.

I pay about $26/month for my AT&T POTS line, with measured rate and no
long distance service. At $15/month, you're probably on Lifeline
service, which is effectively subsidized home phone service. Please
find a better comparison.

I don't know what "lifeline" is, but I assure you my phone service is
*not* subsidized. This is what a phone line costs from Verizon. I pay
$0.10 per call (which are typically few) and more than half the $15 is
taxes and fees. The actual phone connection is just $6-7 a month. I
think that is true about anywhere. My two places are 100 miles apart
and in two different states and the bills were about the same with the
same options for service. When I recently inquired (as part of my
search for Internet, the WISP stuff is very new) Verizon has a very
strange way of presenting the available services on their web site. I
know you can get a simple, minimum cost plan from them because it is
regulated by a public service commission. The web site seems to hide
all this. I think every plan they showed had a bundle of services even
if only long distance, caller ID, etc.

Yeah, not very clear at all. It kinda reads like it is assumed that
everyone will get E911 service.

Even if they don't have outbound service? How do you use it?

Dunno. Since Callcentric sells inbound and outbound service
separately, it's conceivable that they only include E911 service if
someone purchases outbound service.

But that is not what their email says. That's my concern, they are
better than that something-Nine outfit, but they still don't make the
costs clear.

Outbound only service has no phone number, right?

Outbound only service has a phone number. For CallCentric is
1777-Your-CCID. CCID is your CallCentric user ID. You can't call
that number from the PSTN (public switched telephone network) but you
can call it from another SIP phone:
http://www.callcentric.com/faq/4/130

Then that is not a "phone" number.

Inbound only service
can't use 911, right?

http://www.callcentric.com/faq/23#159
The $1.50 fee is NOT included in the following services:
Pay Per Call outgoing service
Any incoming service
Under these products you WILL be billed for 911 separately,
at $1.50/month plus a one time $1.50 setup fee, if you state
that your are in the US or Canada.
The way I read this is that you can make 911 calls with incoming only
service.

I don't get that. They say they will bill you for 911, but they don't
say you can make calls.

I was aware that cell phones could reach 911, I didn't know that about
POTS. So even if you are turned off for non-payment you can dial 911?

Dunno. I've never had the experience. However, it's my understanding
that if you have dialtone, you can dial 911.

That's different, when you are cut off for non-payment you have no
dialtone. But then I don't know for sure, I haven't had that problem in
many, many years.

Interesting. That would solve my 911 problem. I had a POTS here many
years ago. I'm pretty sure I don't get dialtone though.

Without dialtone, it won't work. This might help:
http://www.consumerreports.org/cro/news/2009/05/update-about-911-and-disconnected-landlines/index.htm

You are an amazing wealth of information I have to say.

Boring... Ok, get a Linksys PAP2-NA. It's cheap, easy to configure,
plenty of versatility, lots of online help, small, portable, and
supported by literally all the VoIP vendors. About $35.
http://www.callcentric.com/support/device/linksys/pap2

Isn't that an out of date unit?

Yep. There are plenty of better and later devices. However, the
PAP2-NA is probably the most widely used and widely supported device.
You should not have any problems using it on any VoIP system. The
only gotcha is that eBay is full of counterfeit PAP2 boxes.
http://www.voipfan.net/other/fakepap2.php

One of the links I was given for
equipment indicated a replacement unit, SPA2102 or something similar.

That's a later unit. There's also an SPA3102 which supports fail over
to POTS. Lots of options. Incidentally, I just ordered another 4
line SPA941 desk phone because I got tried of only having one line on
my current SPA921. More than one line is really handy.

One thing I'm not totally clear about is the second FXS jack on many
adapters. The accounts seem to say they support a second phone, but in
what way? Does that second phone have a separate number? Does it use
the same number but you can have two separate calls on the same number?

You might look into Callcentric "extensions" which are much the same
as simultaneous ring:
http://www.callcentric.com/faq/35

Thanks.

Incidentally, CallCentric recently added some new features, some of
which involve extensions.
http://www.dslreports.com/forum/r28535967-CallCentric-CallCentric-announces-more-new-features

Still, $400 is a bit steep. Once I get VOIP I'll be happy. Everything
else he does with his iPhone can be done through wifi.

So, put a SIP softphone client in his iPhone and be done with it. One
device does it all.
https://www.counterpath.com/softphone-products.html
Try a test call and see if the 900 MHz WISP link can handle it.

We've already done that with a friend's phone adapter on more than one
occasion.

Show the zBoost page to your roommate and maybe he'll buy one and let
you use it.

I bet not...

Assuming AT&T 4G data and an iPhone 4S, he's probably paying about
$80/month for the service (and subsidized phone). That's almost
$1,000/year. Somehow, $400 doesn't seem to much any more.

Yes, he pays $100 a month for an unlimited plan (grandfathered in). But
disposable income is hard to come by for a lot of people regardless of
what they consider "essentials".

The NetTalk is the path of least resistance, but I do like the idea of
getting service with a provider and having my own "universal" adapter.
We'll see. NetTalk is so cheap it will be hard to beat on price unless
I lived in NY state (NetTalk actually has totally free accounts in NY
for local service). I'm going into town tomorrow (it's a 25 mile drive
to anyplace that would sell this sort of electronics) and I'll see if I
can pick one up. I think the Walmart web site says they don't carry
them in the stores, but you have to order it online and they ship it to
the store which still takes three or four days. So I might still not
have it tomorrow.

--

Rick

 

[rickman]

On 8/18/2013 3:53 PM, Don Y wrote:

For a power supply, often just cannabilizing some *other* power
supply (same or greater capabilities) and just marrying the right
cable/connector to it. (I'm not fond of rebuilding encapsulated
supplies because they are tricky to get apart and back together
while still *looking* "good" :> )

That is one repair I have done lately. All of my PS adapters wear out
the cable where it goes into the "strain relief". For some reason the
wire in the cable gets covered with something gooey. But I cut the
cord, crack open the case and reattach it. The only hard part is trying
to get the wires clean enough to take solder.

I'd love to get my hands on a couple of old, recycled laptops. I've
gotten a couple of working desktops on Freecycle, but laptops are harder
to come by. I just can't see spending bucks on a new machine when
something five years old would still be perfect for many apps.

C's laptop came to us in a similar way: a neighbor gave me one
of their kids' laptops: "it doesn't work" (bad power supply).
I tested it and shoved it in a desk drawer. Almost a year later,
another friend dropped off the exact same model -- *with* a
power supply (but with some cosmetic damage to the case). It
doesn't take a rocket scientist to piece together one from two! :

One of *my* laptops came from a friend at a three-letter company
in feenigs. Another "bad power supply" story. I just found a
similar supply and doctored the plug (barrel connector). While
it's not a convenient laptop to lug around (heavy as all sin!),
it's reasonably responsive and has a good set of I/Os for the
types of devices with which I typically have to interface
(e.g., a *real* serial port)

Many of the things I do with laptops are as desktop replacements.
Desktops are actually very inconvenient because of the size.

shrug> You'd be heartbroken to see how much stuff gets
"discarded" -- often because people just want new or don't
want to bother figuring out what's wrong with what they have!
While much of it can be diverted from landfills, it's still
an incredibly inefficient use of resources (e.g., I think
a desktop PC has a "recycle" value of < $10)

I know and I *am*... but I understand *why*. I used to think people
were talking through their hats when they said it was all about planned
obsolescence. But computers really aren't any faster than they were
five years ago and they don't do anything they didn't do before or even
better than before, so why do we have to keep upgrading? Because that
is the only way the software and hardware companies can stay in business
cranking out new stuff.

I recently heard that sales of new PCs was *way* down. This may be the
beginning of the end for the traditional PC. It may be all tablets and
smart phones in another five years.

So why can't VOIP advance as fast and become a staple of communications?

--

Rick

 

[rickman]

On 8/17/2013 10:53 PM, miso wrote:

I don't have the option of using Comcast as they don't service this
area. Cell data connections don't work reliably here either. In fact,
my options for digital connectivity were limited to satellite and
dial-up with a friend on satellite explaining that there isn't a lot of
difference sometimes. Until... someone who provides Internet off of
towers, not cell towers, but dedicated links. He uses a device in the
900 MHz band according to the info I found on it. Initially it didn't
do so well, but after a couple of weeks I have no complaints.


This sounds like WISP, though I find the links in the 2.4G range. I have
some 900MHz spread spectrum backhaul gear, but damn slow.

Satellite VOIP will be terrible. Consider the latency.

When I looked up the manual for the equipment I was given they sell
equivalent units to work at a variety of frequencies.

--

Rick

 

[Jeff Liebermann]

On Mon, 19 Aug 2013 02:29:35 -0400, rickman <gnuarm@gmail.com> wrote:

Uh, I think about 10 posts ago I said a friend had already tried her
VOIP phone on my Internet connection and it seemed to work fine. So all
of the above jitter tests, etc are overkill if the connection passes the
acid test, no?

No. One phone call does not constitute a test. You don't have any
measurements, just pass/fail. You should see what happens under worst
case situations, such as you downloading something on another
computer, or running Bitorrent. Otherwise, you run the risk of having
it only work when nobody else is using the internet.

You used the WISP term which is the name the provider uses for this and
in fact the ISP company name is iWisp. I take it this is a generic
term? I thought it was a brand name. Is there just one major brand of
products doing this function?

Sorry. When talking telecom, everything is in acronyms. I tried to
avoid doing that but missed the WISP acronym. It's short for Wireless
Internet Service Provider. I used to work with one in Boulder Creek
about 15 years ago.

So I have to assume that the
products work for most, but not all and more important is how the
company deals with it.

That's typical with VoIP. The quality of the service is largely
dependent on the quality of the internet connection. If you have lots
of bandwidth, little jitter, and a proper router, you'll have no
problem. If any of these are lacking, you will have problem. I kinda
like the vendors where the review sites are full of billing
complaints. That means that the service works just fine and that all
that's left to complain about is the billing.

No one complained that NetTalk didn't have
reasonable support.

I worry a little when that happens. Quality products don't need much
support.

I pay about $26/month for my AT&T POTS line, with measured rate and no
long distance service. At $15/month, you're probably on Lifeline
service, which is effectively subsidized home phone service. Please
find a better comparison.

I don't know what "lifeline" is, but I assure you my phone service is
*not* subsidized. This is what a phone line costs from Verizon. I pay
$0.10 per call (which are typically few) and more than half the $15 is
taxes and fees. The actual phone connection is just $6-7 a month. I
think that is true about anywhere (...).

Sorry. I'm in AT&T territory and know little of Verizon POTS pricing.
I didn't know they could do it that cheap. I would probably use the
same plan if I could get it here.

Since you have two location, you can get two phones setup as
"extensions" on one phone number. That cuts your phone bill in half.

Outbound only service has no phone number, right?

Outbound only service has a phone number. For CallCentric is
1777-Your-CCID. CCID is your CallCentric user ID. You can't call
that number from the PSTN (public switched telephone network) but you
can call it from another SIP phone:
http://www.callcentric.com/faq/4/130

Then that is not a "phone" number.

Correct. It's really a CallCentric ID number.

Inbound only service
can't use 911, right?

http://www.callcentric.com/faq/23#159
The $1.50 fee is NOT included in the following services:
Pay Per Call outgoing service
Any incoming service
Under these products you WILL be billed for 911 separately,
at $1.50/month plus a one time $1.50 setup fee, if you state
that your are in the US or Canada.
The way I read this is that you can make 911 calls with incoming only
service.

I don't get that. They say they will bill you for 911, but they don't
say you can make calls.

The only question is whether the ingoing only service provides
dialtone. If it does, you can make 911 calls. If not, you can't make
911 calls. When I was using CallCentric, I had both incoming and
outgoing, so it wasn't an issue. I don't see it as much of an issue
as you're more likely to be buying both in and out service.

That's different, when you are cut off for non-payment you have no
dialtone. But then I don't know for sure, I haven't had that problem in
many, many years.

I vaguely recall some kind of regrettable incident, where AT&T
suddenly cut off service for non-payment, where the resident needed to
make an emergency call but couldn't. After that public relations
disaster, AT&T leaves dialtone and 911 service enabled for
non-payment. I'm not sure how long they leave it on, but when I took
possession of a rental house that had been empty for about 90 days, I
still had dialtone on the previous owner. Verizon may be different.

>You are an amazing wealth of information I have to say.

I prefer the other type of wealth, where I'm somewhat lacking.

One thing I'm not totally clear about is the second FXS jack on many
adapters. The accounts seem to say they support a second phone, but in
what way? Does that second phone have a separate number? Does it use
the same number but you can have two separate calls on the same number?

It's usually a separate line with a separate phone number or a
separate phone extension on the same phone number as the main line. It
totally depends on how it's provisioned with plenty of creative
options. If you buy your ATA from a VoIP provider, it's most likely
going to be setup as an extension, where you can have two
conversations at the same time over the same internet connection (but
not via two different providers as the VoIP provider isn't terribly
thrilled with you using his equipment on the competitions system).

On a more complexicated ATA, like the SPA3102, the 2nd line can be:
1. A 2nd service provider with a separate phone number.
2. An extension on the same phone number as line 1.
3. A POTS fallback or pass through.
The last is a bit tricky. It's commonly used direct local calls to a
POTS line, where the CLEC provides a large number of bundled "free"
minutes, and switch automatically to the VoIP line, for toll and long
distance calls, which would be charged by the minute.

The NetTalk is the path of least resistance, but I do like the idea of
getting service with a provider and having my own "universal" adapter.
We'll see. NetTalk is so cheap it will be hard to beat on price unless
I lived in NY state (NetTalk actually has totally free accounts in NY
for local service).

Also available from CallCentric:
<http://www.callcentric.com/dids/free_phone_number>
The catch is that you have to be in New York state and you are
required to pay for E911 service. The service originates with
Telengy, which is a NY area CLEC and the parent company of
CallCentric. I don't know how NetTalk got into the puzzle, but my
guess is that they also get their DID (dial in direct) lines from
Telengy and may have a similar billing arrangement.

Good luck.

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

 

[Jeff Liebermann]

On Mon, 19 Aug 2013 02:45:20 -0400, rickman <gnuarm@gmail.com> wrote:

>So why can't VOIP advance as fast and become a staple of communications?

Because most internet services (HTML, SMTP, etc) need only gross
bandwidth and some buffering to work. You can have substantial
delays, such as in satellite communications, and it will still work.

VoIP needs low latency, low jitter, low packet loss, and low out of
order packets. In the rush to broadband, these specifications are
being compromised (or ignored). We are not going to have reliable and
widespread VoIP service until the major providers either improve these
factors, or provide a seperate path (as Comcast had done).

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

 

[miso]

On 8/19/2013 8:48 AM, Jeff Liebermann wrote:

On Mon, 19 Aug 2013 02:45:20 -0400, rickman <gnuarm@gmail.com> wrote:

So why can't VOIP advance as fast and become a staple of communications?

Because most internet services (HTML, SMTP, etc) need only gross
bandwidth and some buffering to work. You can have substantial
delays, such as in satellite communications, and it will still work.

VoIP needs low latency, low jitter, low packet loss, and low out of
order packets. In the rush to broadband, these specifications are
being compromised (or ignored). We are not going to have reliable and
widespread VoIP service until the major providers either improve these
factors, or provide a seperate path (as Comcast had done).

A lot of podcasts and even some low budget news shows like Democracy Now
use VOIP in the form of Skype. It is just awful. Words getting chopped.
Video going haywire.

Some of the network companies came from a telephony background. They
understood the things that make VOIP work. The problem was if you were a
telephony company, you had that big customer base and didn't have to be
so nimble in business. Nortel being a good example of how to have good
technology and totally fail. Lucent would have been tits up except for
the purchase by Alcatel.

So networking isn't about voice, it is about data.