SB> IS> It depends on the modems and chipsets/firmware used, some use quite
 SB> IS> different algorithms for line quality versus channel bitrate
 SB> IS> adjustments.  Some are more or less adjustable by the user, and some
 SB> IS> are not.

 SB> Another thing I didnt mention is the protocol. I understand that V34
 SB> ALWAYS has retraining enabled (?) and that its only V32/bis that can
 SB> have it enabled/disabled.

No, you're still talking about some specific manufacturer's option sets in
different revisions, not the ITU-T standards as such, which are very
open-ended about how (and which) options may be implemented.  I'm not
really clear which modem brands / models / chipsets we're talking about
here?

 SB> IS> In Rockwell-chip modems %E0 both disables local and rejects remote
 SB> IS> retrains, as I understand it.  Noone seems to use
 SB> that, except perhaps
 SB> IS> on leased lines.

 SB> I have seen a number of modems, including my old 14.4K Cirrus-Logic based
 SB> Dynalink that actually have %E0 in the factory defaults. (AT&F) DOH!

Yes, I once had to play with an earlier 14400 Dynalink for a friend. 
Dreadful defaults.  No, not dreadful - near unusable.  Once properly setup
it went ok, but most people don't expect to have to learn esoteric commands
to change the out-of-the-box setup .. and some relatively mediocre modems
have sold well on reputation, just by having reasonable turn-it-on-and-go
defaults.

 SB> IS> %E1 accepts and responds to both fallback and
 SB> IS> fallforward requests from the remote modem, but only initiates
 SB> IS> fallback requests when detecting unacceptable signal quality (at a
 SB> IS> desired bit error rate), while %E2 allows the modem to
 SB> both accept and
 SB> IS> initiate fallforward and fallback requests as necessary.

 SB> My old modem didnt have %E2 as a valid setting, but it did have %G1 to
 SB> enable fallforward/fallback.

The %E commands I mentioned above were specifically for Rockwell-chip
modems; I only chose these because they're so common, so took a punt.  Very
different commands to my Dataplex, too, as well as your Cirrus-chip
Dynalink.  And thanks for that info - I knew it wasn't Rockwell, but
thought maybe AT&T ..

 SB> etc.  Differing
 SB> IS> revisions of some modems/chipsets may also behave quite differently
 SB> IS> with different types of line impairments.

 SB> Yup, there are definately no clear cut answer with anything
 SB> to do with modems,

Indeed, yet more and more people expect them all to talk the same lingo.

 SB> people saying things like "oh its a Rockwell modem, it will
 SB> do such and such" is generalising somewhat...

True, though it's likelier to be close than one (say, like mine) that uses
%E to set its Compromise Equaliser, and %G to set Signal Quality Thresholds
:)

 SB> IS> (understandably, given the usual poor standard of technical
 SB> IS> documentation in modem manuals) seem to think that %E1
 SB> will stop their
 SB> IS> modem falling forward when requested to do so by a
 SB> remote (%E2-using,
 SB> IS> in Rockwell parlance) modem, and perhaps avoid using it due to that
 SB> IS> misunderstanding.  It's demonstrably not so.

 SB> Interesting. Yes, the manual for my Dyanlink was pretty sparse, and was
 SB> basically just a glorified command listing. The modem I had before that,
 SB> which was an old Rockwell-based GVC 14.4 Fax/Data had an
 SB> absolutely excellent
 SB> book. It literally had chapters dedicated to explaining
 SB> things like retraining,
 SB> error correction, compression, etc...but I dont have that book anymore :(

I wouldn't buy any modem without a decent manual, or so I used to say ..
but I've recently obtained a second-hand Netcomm M7F without one, and I'm
hoping it'll work on pretty poor lines, where you need to fiddle .. see
next msg.

 SB> IS> full retrains), moving next to no data meanwhile.  Many failed
 SB> IS> connects, much cash being wasted.
 SB> IS> Using %E1 on the 'problem' modem allowed my modem to control rate
 SB> IS> adjustment attempts, keeping line speed down and
 SB> successful throughput
 SB> IS> rates up, while allowing fallforward when, as often
 SB> happened, the line
 SB> IS> was seen to improve later on in the call, allowing

 SB> Hmm, so what setting do you recommend in general? Or is there no
"Best"
 SB> setting?

No, it depends on brand/model/chipset and in some cases, firmware revision.
 I've since seen your later message where you mention the USR / Dynalink,
and I can't help with setup for those at all.

 SB> IS> Keeping a (local, at least) connection healthy matters
 SB> much more than
 SB> IS> top speed, at least when you're interested in
 SB> minimising cost.  Other
 SB> IS> people out on poorer lines have found %E1 at their end greatly
 SB> IS> improving connectivity, even with some assorted more recent Rockwell
 SB> IS> gadgets, to here anyway ..

 SB> Thanks for your informative reply, Ian.

Just a rave .. people whose lines are only 'marginal' with respect to
whether they get 33600 connects both ways all/most/some of the time haven't
lived! :)

Cheers, Ian

--- MaltEd 1.0.b5