NetNews Usenet Archive 1992 #27/NN_1992_27.iso/spool/comp/dcom/fax/2137

Text File | 1992-11-17 | 7.8 KB | 169 lines

Newsgroups: comp.dcom.fax
Path: sparky!uunet!rde!ksmith!keith
From: [e-mail] (Keith Smith)
Subject: Re: MultiTech Fax Modems could run Gnu NetFax if you call
Organization: Keith's Computer, Hope Mills, NC
Date: Mon, 16 Nov 92 20:36:49 GMT
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]> <[e-mail]>
Lines: 158

In article <[e-mail]> [e-mail] (Stephen Friedl) writes:
>I write UNIX fax software for a living, so I have some expertise
>in this area.  I believe that the UNIX and DOS fax worlds are
>fundamentally different markets, and they share surprisingly few
>characteristics.  Please note, however, that I do not presume to
>speak for anybody but me.

I will agree here, but the issue is *STANDARDS*.  Just like V.32 & V.42
and ieee 802.3 and TCP/IP and ...

>
>I suggest that moving to the Real Class 2 standard is going to be
>a lot slower than many of us would like, and that the fax
>software vendors (both UNIX and otherwise) will have a MUCH
>higher impact on this than will individuals.  After my posting,
>we see this response putting words in the mouths of the fax
>vendors:
>
>> Yea,  f*ck it.  I could care less whether or not my software will
>> run on more than one fax modem.  Screw the standards.
>
^^^^^^^^^^^^^^^^
This was my statement & not attributed & quoted out of context.  It
should be heavy with :) :) :) :)!

>Well, in a way, yes.  Most of the UNIX fax software vendors only
>support a couple of modems -- V-Systems ONLY supports those from
>Multi-Tech -- and I invite you to ask them why.  Most of them

Probably because There *ARE NO DECENT STANDARDS* right now, and it is
extremely difficult to support 80 bazillion different commands for each
modem to do what you want.

>undoubtedly work on other modems, they may admit this, but they
>simply won't support them.  Why might this be?

See above.

>
>The answer is that providing real live commercial support for a
>lot of modems is a huge headache with very little benefit, and

BECAUSE of the lack of a standard.  i.e.  You can't say, Because that
modem does not adhere properly to the standard specification,
specifically command XXXX, Tell the modem manufacturer to fix his
firmware.

>nobody wants to deal with it.  Unlike data modems where there is
>fairly little conversation between the host software and the
>modem itself (most data is passed onto the other end), fax modem
>firmware is quite complicated and always has bugs.

No more so than V.32bis, V.42bis, or just interpeting the AT commands
properly.  C'mon, be realistic.  The difference is that again, there was
no standard, so who is to say what the "proper" behavior should be?

>V-Systems has been fortunate that Multi-Tech has provided us very
>high quality firmware and unbelievably good support, but we still
>have to keep track of firmware revisions for the handful of their
>modems that know about fax.  It is not a free ride even if you go
>with the best.

I agree.  I really like the MT products.  I'd also bet that they will
adopt most of these standards.  But the point is that no software folks
will write software for a standard if no modem will support it.  That's
dumb to write software you can't use.  And there are always bugs in this
new technology, give it a year.

>I write:
>>We have invested a lot of time getting the software to work with
>> these modems in a commercial environment, and we cannot for the
>> life of us think of any reason to want to start over.
>
>our poster responds:
>> start over what? How about *ADD* adherance to the standard instead of
>> starting over.  Then your crummy over-priced commercial products will
>> run on more than one modem, broadening your market for more sales.  Wow,
>> what a concept!
>
>Obviously, a well-designed facsimile system will not have to be
>rewritten from scratch, but the modem driver certainly will be: I
>believe that that Real Class 2 standard is sufficiently different
>that simply modifying the current driver won't work.  I do
>believe, though, that this new driver will be much easier to
>write than the first one, because the experience of the first all
>applies to the second.

Gee, so your modem driver will first do a class 2 check, and send
commands as appropriate.  Not a problem at all.  Even make it
configurable (Try class 2 first (Y/N)?) setup.

>So, we are sitting here with solid modem firmware, a good host
>modem driver, and happy customers: why would we want to do this
>rewrite?  Not only will we have to do our stuff again, we have

You just said the only stuff that would change would be the modem
driver!

>to watch while the modem vendor does the same thing.  What is
>means is that for a while, everything is shaky again until all
>the bugs are worked out, but what do we have?  Support for a
>standard modem that allows us to plug+play other standard
>modems.  Why would we want this?

So you can sell your software to people who didn't buy the single type
of modem that you have.  For example, suppose Telebit gets off it's ass
and builds fax into it's Worldblazer that is class 2 compatable.  Now
LOTS of unix folks have WB modems, and Company M has software that
supports class 2 fax modems, not as good as yours but similarly priced
and it works with these 10 class 2 modems you already have.  Lessee I
want fax, and I can trash all my Super-PEP screamer modems and buy
MultiTech's at $600 a pop + your software, or I can buy this other
software and a firmware upgrade, spend half the money and still have the
same capability.  Now you tell me why.

>As I said, supporting fax modems under UNIX is a huge headache.
>You cannot believe how much hassle it is trying to figure out how
>to build the proper cable for an XYZ serial card (that the
>customer may not have documentation for), spending hours tracking
>down a bug and finding that it's in the kernel's tty driver, or
>telling a customer that the only thing they can do is reboot
>every so often to clear a locked port.  V-Systems supports dozens
>of machines, and as the one who did the majority of the ports and
>cable specifications, I know a lot about this.

What do cableing, OS differences, and Customer stupidity have to do with
adhereance to Open Standards?

>Basically, the UNIX fax customers are not clammoring for support
>for cheap modems (though they would not object to them if they
>worked well), and Multi-Tech is working very well.  They are right
>to see if the marketplace demands Real Class 2 before they do it,
>and they are sure not getting pushed on this from us.  I hope we
>don't have to think about it for a long time.  What Hayes does
>will have a big impact on this (again, I hope I don't have to
>think about it).

Nope, but Unix folks aren't any different than any others, in that if
they have already invested heavily in a different brand modem that WILL
support the standard they damn sure ain't gonna throw it away just for
the ability to run one particular vendors software

>This is a lot of stuff, and the folks used to DOS faxing will
>surely think this is arrogant or lazy or something like that.
>I hope that the other UNIX fax vendors will kick in their two
>cents worth (especially if they agree with me).

On the DOS side of the coin, most FAX modems *COME* with DOS faxing
software, so the standard is not as big of an issue.  The whole Open
Systems concept is based on standards,  and the more hardware your
software supports, the bigger your market.  The more standardized the
hardware in your proposed enviroments the easier it is to support more
different vendors products.  Plain and simple, the standard broadens the
market for products that use it.
--
Keith Smith          uunet!ksmith!keith            5719 Archer Rd.
Digital Designs      BBS 1-919-423-4216            Hope Mills, NC 28348-2201
Somewhere in the Styx of North Carolina ...