6 May
2014
6 May
'14
9:57 p.m.
My son who is 9 loves vintage computers and telephones like me, So we
figured why not combine our hobbies into one big project.
We want to recreate the fun of the unix dial in days, and dialing in via
PPP to the internet. The goal eventually will be to have 12 lines
serving up Shell and PPP access via VOIP inbound at speeds around 14.4.
So all of you who have old machines that have modems unused, youll be
able to call into one of the inbound numbers and get online via shell
and PPP. Accounts on the system will be free. Depending on what
type of system you use will determine the access number.
We plan to have a Sun Sparcstation 5 running SunOS 4.1.4 for dial in
access, an HP 712 running HP UX 9.0, and a NeXTstation. Adding more
as we find more older unix machines to put on the system as we go.
Our project for the Mason Michigan Phone show was demonstrating analog
modem connections from computer to computer through a PBX. It was a
hit, We had the commodore 64 calling the Mac 512 and then both machines
calling out to DiversiDial Station #34 on C*NET.
We plan to use the following hardware
Asterisk Server to handle call routing and C*NET connections
Sun SparcStation 5 Running SunOS 4.1.4
HP 712 running HP UX 9.0.4
NeXTstation running NeXTstep 3.3
A regular old linux box providing shell and PPP access.
It gives you a good reason to dust off those old computers, fire up
your favorite terminal program and off you go :)
6 May
6 May
11:55 p.m.
New subject: Son and I's Project- Recreating the Dial in ISP Experience
On Tue, May 6, 2014 at 10:57 PM, Steven Landon <slandon110 at gmail.com> wrote:
We want to recreate the fun of the unix dial in days,
and dialing in via
PPP to the internet. The goal eventually will be to have 12 lines serving
up Shell and PPP access via VOIP inbound at speeds around 14.4.
Cool project! I'm interested in hearing how well V.32bis QAM modulation
(14.4 Kbps) works over VOIP, since a lot of VOIP applies compression
schemes intended specifically for voice. On uncompressed PCM, or PCM with
only basic companding (mu-law or A-law) it should work fine, since that's
what has most commonly been used by the phone companies for carrying
domestic POTS calls. International voice calls often used more
compression, which at the very least could be expected to increase the
error rate.
Eric
7 May
7 May
2:01 a.m.
New subject: Son and I's Project- Recreating the Dial in ISP Experience
On Wed, May 07, 2014 at 12:55:45AM -0600, Eric Smith wrote:
[...]
Cool project! I'm interested in hearing how well
V.32bis QAM modulation (14.4
Kbps) works over VOIP, since a lot of VOIP applies compression schemes
intended specifically for voice. On uncompressed PCM, or PCM with only basic
companding (mu-law or A-law) it should work fine, since that's what has most
commonly been used by the phone companies for carrying domestic POTS calls.
International voice calls often used more compression, which at the very
least could be expected to increase the error rate.
My experiments on this front involved CSD calls to my VoIP number, which was
delivered over a 2M/256k ADSL to an ATA into which I'd plugged my old USR
modem.
(CSD is https://en.wikipedia.org/wiki/Circuit_Switched_Data, which is basically
dialup over mobile. It was reasonably popular for mobile Internet access until
the telcos offered cheap(er) packet-switched services.)
I never did get it to work reliably even though it should have been a shoo-in
for a solid connection. After all, the only analogue hop was the couple of feet
of phone cable between the ATA and the modem. I suspect I was being nobbled by
jitter and/or packet loss and the ATA wasn't re-clocking the signal properly.
Yes, I was using the G.711a "codec" so it wasn't compression that was
killing
it.
Curiously, the telco's modem bank tried to make V.34 connections even though
CSD can only do 9.6kb/s. Turning off the advanced modulation schemes in my
modem got me a passable connection at 7.2kb/s with periodic retrains.
I'd try this again now I'm on 17M/1M ADSL, but I'm not sure I could assemble
the technology any more. My old phones have packed in, and I'm unconvinced that
modern handsets implement this somewhat obsolete standard. (And also, my telco
doesn't have any GSM infrastructure, although it's not as if I can't get hold
of a SIM for one that does.)
8:31 a.m.
New subject: Son and I's Project- Recreating the Dial in ISP Experience
> I'm interested in hearing how well V.32bis QAM
modulation (14.4
> Kbps) works over VOIP, since a lot of VOIP applies compression
> schemes intended specifically for voice. On uncompressed PCM, or
> PCM with only basic companding (mu-law or A-law) it should work
> fine, since that's what has most commonly been used by the phone
> companies for carrying domestic POTS calls.
Well...it should work fine if there is absolutely zero packet loss.
Modem codecs designed to handle POTS-generated errors usually do not
perform well in the face of VoI-generated errors. I ran into this with
faxes (which are layered atop one of the modem codecs, I forget which
one); I found that faxing over VoI worked well only with zero ploss.
Even so much as one lost packet caused rather unpleasant behaviour.
(There is a fax codec - T.38 is the number my memory is giving me -
that is designed to deal with VoI-style errors rather than POTS-style
errors; my limited impression is that support for it is uncommon but
works well when present.)
My experiments on this front involved CSD calls to my
VoIP number,
[...]
I never did get it to work reliably even though it
should have been a
shoo-in for a solid connection. After all, the only analogue hop was
the couple of feet of phone cable between the ATA and the modem.
Data ploss was probably the culprit. It's rare for a long-haul link on
the open Internet to run with absolutely no ploss.
I suspect I was being nobbled by jitter and/or packet
loss and the
ATA wasn't re-clocking the signal properly.
And/or you were running into designed-for-POTS codecs getting hit with
non-POTS errors.
Yes, I was using the G.711a "codec" so it
wasn't compression that was
killing it.
My experiments also used G.711. I worked, at the time, for a provider
that did VoI telephony and had an interest in knowing how well faxing
could be expected to work....
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
11:05 a.m.
New subject: Son and I's Project- Recreating the Dial in ISP Experience
On 05/07/2014 08:31 AM, Mouse wrote:
But, but--I thought the point of this was to re-create the experience!
Say, like calling Palo Alto from Los Gatos in 1975....
(Those who have actually had the experience will remember it vividly.
Los Gatos was GTE; Palo Alto was Pac Bell and where the twain met made
for some interesting experiences).
--Chuck
> I'm
interested in hearing how well V.32bis QAM modulation (14.4
> Kbps) works over VOIP, since a lot of VOIP applies compression
> schemes intended specifically for voice. On uncompressed PCM, or
> PCM with only basic companding (mu-law or A-law) it should work
> fine, since that's what has most commonly been used by the phone
> companies for carrying domestic POTS calls.
Well...it should work fine if there is absolutely zero packet loss.
3655
days inactive
3655
days old
4 comments
5 participants
participants (5)
-
abuse＠cabal.org.uk -
cclist＠sydex.com -
mouse＠Rodents-Montreal.ORG -
slandon110＠gmail.com -
spacewar＠gmail.com