Re: [hylafax-users] rockforce octo and multitech MultiModemISI

["Andrew Rinaldi" <andrew.rinaldi@xxxxxxxxxxxx>]

Hi Uwe,

I'll do my best to answer your questions.

'Wedging' is a generic term which is used in the HylaFAX world to describe a
state in which the application and modem can no longer communicate.  We have
seen many reasons for this over the years having built products with or
evaluated almost every modem chipset.  Here are a few that spring to mind;

DSP hardware or firmware crash
controller hardware or firmware crash
'modem' hardware bugs (Venus Interrupt register bug is a good example of
this)
serial driver bugs (the Linux serial driver is in a constant state of flux)

As a modem developer we can fix many of the DSP and controller firmware
issues, and during the last 10 years we have fixed most of them on Venus.
Obviously, we cannot fix the chipset hardware bugs although we do provide
workarounds in our hardware (SideBand reset and SideBand Interrupt Status)
and we do actively work with the chipset manufacturers to improve their
products (we are heavily involved in the LSI [Agere] replacement for Venus -
CFAX).  The serial driver issues are obviously fixable but a significant
burden with so much kernel development taking place....

I agree that wedging should not exist.  Indeed on our new PCI Express
products we have not seen any of the chipset related issues that we had with
Venus.

Fax reliability should not be a function of internal or external but in
reality it comes down to who's developing products in the fax market.  Most
USB modems are 'soft' and none of the modem chipset vendors are still
developing or releasing fax firmware for this market.  USB also introduces
some timing constraints which can become difficult when you scale to large
numbers of lines.  Historically, we developed a range of USB products but
never released them because we have very fixed ideas about what we want in
terms of reliability - we are trying to achieve fax failure rates in the
'real' world below 0.1%.

Please remember that fax is a niche market so most of the chipset vendors
are not interested.

The IQ Express family includes a 1 port product.  The whole family is half
height, half length, PCI Express, that will fit into some incredibly small
systems.  It's also based on the very latest CFAX chipset so you can count
on many years availability and support (unlike Venus which is close to EOL).

http://www.mainpine.com/products_IQE.php

The price you pay for a Mainpine product includes all of the development
expertise that we bring to the fax market, it includes support for almost
every fax package, including HylaFAX and HylaFAX+.  You say that paying 4
times as much as the competitor is not worth the money... but I ask you how
much is your time worth and how much do lost fax's cost you or your customer
?  I also suspect that Multitech's response is telling you something....

I don't know the answer to your question about OpenBSD because my expertise
is much more on the hardware and firmware side of things.... did you raise a
mainpine support ticket ?  I'd be very surprised if you didn't get a
response to that sort of question....

Regards

ANDREW RINALDI
Mainpine Limited - Support
USA +1 503 822 9944 | Asia/Europe +44 1225 869439  
andrew.rinaldi@xxxxxxxxxxxx | www.mainpine.com

-----Original Message-----
From: Uwe Dippel [mailto:udippel@xxxxxxxxx] 
Sent: 15 June 2007 09:10
To: andrew.rinaldi@xxxxxxxxxxxx
Cc: Yohan Yudanara; hylafax-users@xxxxxxxxxxx
Subject: Re: [hylafax-users] rockforce octo and multitech MultiModemISI

On 6/13/07, Andrew Rinaldi <andrew.rinaldi@xxxxxxxxxxxx> wrote:

[...]

Though I can partially agree with you, I am still amazed about this
'wedgeing' problem. Yes, it exists. But it does not for cards like VGA
or NICs that transport a thousand times more data. My suspicion is
still on hardware or software ...

Also, (I did ask for external, USB modems a few weeks ago, without
answer), I can't make out why an external modem should be less
reliable than an internal one. We want to go for embedded systems; and
then PCI-slots are not always the preferred choice. Actually, the
combined embedded system and the external modem can be smaller than
the PCI in the casing. Plus, with USB one should be able to be more
flexible w.r.t. the number of lines.

To be honest, and here Yohan might be able to help, I have tried in
vain over the last weeks to buy a few samples from Multitech;
contacting them through their site, a few times, no answer, sending an
e-mail, no answer, finally sending a fax, still no answer. I wonder
about their business model ... ;)
Would you have any e-mail address there, where someone is actually
willing to sell a modem or two ?

I do appreciate the presence of Mainpine in here; and the support of
Hylafax. But the restriction to PCI and a minimal number of two lines
is an obstacle as well. If, if, I could get a single line V34 from
that other company, the difference in price would be about 1:4; also
because of the one line (that is sufficient for us) versus 2 lines
(minimum).

Oh, and I still did not get an answer for my question, per e-mail, if
mainpine supports OpenBSD out of the box ?

Uwe



____________________ HylaFAX(tm) Users Mailing List _______________________
  To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
 On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
  *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*