Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!zephyr.ens.tek.com!uw-beaver!milton!sumax!polari!rwing!pat From: pat@rwing.UUCP (Pat Myrto) Newsgroups: comp.dcom.modems Subject: Re: Overseas Connections with high-speed modems Message-ID: <242@rwing.UUCP> Date: 3 Mar 91 01:35:23 GMT References: <1018@nih-csl.nih.gov> Organization: Totally Unorganized Lines: 75 In article <1018@nih-csl.nih.gov> bert@helix.nih.gov (Bert Tyler) writes: > > ...[ deleted ] ... > >Apparantly I did not emphasize enough that these folks will be >using MS-DOS machines and MS-DOS-based software (in fact, the US >office just found out yesterday that they will be using a package >called TRAX, which is sold in the UK and is something like >Close-up/Carbon-Copy) for their file transfer. I received several >responses about the advantages of Telebit's "spoofing" capability. >While V.42bis compression looks like a very handy feature for this >application, Telebit modems won't find any protocol to "spoof". Oh, yes there ARE protocols to "spoof" in the MSDOS world - virtually all the non-streaming (ACK-NAK, or 'send and wait') file transfer protocols take a big performance hit with the hi-speed modems, even with non-PEP modem protocols, probably because of the packet-type nature of the error-correction schemes, such as MNP. The Telebits "spoofing" feature improves the performance of the XMODEM type protocols such as XMODEM, XMODEM-1K, and Batch YMODEM; the various flavors of Kermit, as well as the UUCP "spoofing". When I use the T2500 to make an interactive connection, I arrange things so it connects with XMODEM type spoofing turned on. For normal interactive use or streaming protocol usage, like ZMODEM, it has no effect that I can tell, but if I have to use XMODEM 1K, or Batch YMODEM, it helps quite a bit. If the other end only supports regular XMODEM (128 byte packets), performance is dismal without the spoofing active. While I haven't tested this, I suspect a 9600 bps link using V.32 and MNP would have poorer performance with XMODEM or YMODEM than PEP would using its spoofing, assuming the modems at each end are hooked to 9600 bps ports. On the other hand, if one isn't going to be transferring files, or only using the streaming type protocols, such as ZMODEM, then the spoofing feature won't result in any gain performance-wise. Another point to consider with PEP is its reputation to hang in there with lousy lines - it will not only reduce its speed if line conditions are bad, but if during the call the conditions improve, it will move its speed back up again - it adjusts dynamically for the line condition. Other schemes will often fall back to a lower speed if line condition deteriorates, but will not 'fall forward' when conditions improve. PEP changes in 100 bps increments, if I recall but V.32 9600 drops to 4800 in one step and stays there if it has to re-adjust. Telebits and PEP may not be what is best for your application, I don't really know what your software supports, etc. All the schemes and modems have their good and not-so-good points - personally I use the T2500 and find it quite robust at the hi speeds, with both PEP and V.32. I also notice that I can have a connection with some other make of modem over a less-than-great line, and the other end will be getting garbage characters in their reception, but my end (receiving with the T2500) will be totally clean. Once or twice one could attribute it to differences in the quality of the paths taken by each direction, but I find this to be consistant. Getting garbage characters on my screen or file xfer errors is a rare occurance for me, with or without an MNP type connection. Perhaps I have an exceptional unit, but the modem at work seems to show the same quality (though its usage over noisy lines has been much more limited). The Telebits offer great flexibility in setup, some folks find this to be a drawback, since they have difficulty in setting up the large number of S registers and options. With a little testing (use a terminal hooked directly to the modem) to get a feel how options affect things, and careful reading of the manual, I find the flexibility extremely useful in adapting to the quirks one finds in different communication packages - especially when reliable unattended operation is desired. I have no connection with Telebit, other than being a happy customer. Perhaps I am easy to please ... :-) -- pat@rwing (Pat Myrto), Seattle, WA ...!uunet!pilchuck!rwing!pat ...!uw-beaver!uw-entropy!dataio!/ WISDOM: "Travelling unarmed is like boating without a life jacket"