evanc@isishq.fidonet.org (Evan Champion) (11/18/90)
I was planning on getting myself a US Robotics Courier HST/Dual Standard, as it was the most highly recommended by my friends. However, since buying one of those will "break the bank", I want to make sure I am buying the best for my money. I like the Dual Standard's 14.4KB, and with v.42bis (I think) that puts it around 30kb correct? Anyway, any help would be appreciated. Evan
larry@nstar (Larry Snyder) (11/20/90)
evanc@isishq.fidonet.org (Evan Champion) writes: >I was planning on getting myself a US Robotics Courier HST/Dual Standard, as >it was the most highly recommended by my friends. However, since buying one >of those will "break the bank", I want to make sure I am buying the best for >my money. I like the Dual Standard's 14.4KB, and with v.42bis (I think) that >puts it around 30kb correct? Anyway, any help would be appreciated. here on nstar we have one dual standard (without v.42bis) and another one on order (should be here later this week - but this one will have v.42bis). The V.32 throughput with uucp sessions runs right around 1150 cps, and around 800 cps when connected to another v.32 without mnp. HST to HST uucp runs at 350 cps (using "g" protocol) while zmodem transfers of compressed and/or binary data runs around 1750 cps (pure text runs around 2600 cps). The Telebits runs around 1350 cps using uucp "g" protocol, and right at 1350 cps using zmodem. The Telebit modems will hold the line better than the dual standard - -- Larry Snyder, Northern Star Communications, Notre Dame, IN USA {larry@nstar, {uunet|backbone}!nstar!larry, larry%nstar@iuvax.cs.indiana.edu} backbone usenet newsfeeds available Public Access Unix Site (219) 289-0282 (5 high speed lines)
root@zswamp.fidonet.org (Geoffrey Welsh) (11/21/90)
Evan Champion (evanc@isishq.fidonet.org ) wrote: >I like the Dual Standard's 14.4KB, and with v.42bis (I think) that >puts it around 30kb correct? Anyway, any help would be appreciated. Manufacturers like to quote *potential* data throughput using data compression. I don't buy it, since much of the marterial passed through these high-speed links is already compressed. V.42bis will not significantly improve the throughput of .ZIP or .Z (compress) files. I like to concentrate on the *raw* data throughput rate and include, where appropriate, the gains realized by synchronous framing (that's throwing away the start & stop bits). In the case of the HST 14.4 (or the Dual Standard), the raw data rate is 14400 bits/second; using MNP4, that gives theoretical maximum data transfer rates approaching 1800 CPS *without* compression (assuming a low-overhead streaming protocol). The number for 9600 bps HSTs and V.32 modems is approximately 1200 CPS. The number for Telebit TrailBlazers is somewhere around 1400 CPS, tapering off as line quality degrades. Heck, if all you're doing is sending spaces, any of these modems will transmit the data as fast as you can send it to the modem from the computer, but that's not a useful benchmark as you will rarely be sending only spaces! Geoff DISCLAIMER: I have not played with V.42bis, and cannot comment on its effectiveness beyond doubting that it will contribute *much* to precompressed files. I *have* played with MNP5 (which, as Toby has pointed out, is a vastly different technique) and usually turn it OFF as soon as I can because I get better overall performance *without* it. -- UUCP: watmath!xenitec!zswamp!root | 602-66 Mooregate Crescent Internet: root@zswamp.fidonet.org | Kitchener, Ontario FidoNet: SYSOP, 1:221/171 | N2M 5E6 CANADA Data: (519) 742-8939 | (519) 741-9553 MC Hammer, n. Device used to ensure firm seating of MicroChannel boards Try our new Bud 'C' compiler... it specializes in 'case' statements!
bruce@ccavax.camb.com (Barton F. Bruce) (11/21/90)
In article <9Naus3w163w@isishq.fidonet.org>, evanc@isishq.fidonet.org (Evan Champion) writes: > I was planning on getting myself a US Robotics Courier HST/Dual Standard, as > ... > I like the Dual Standard's 14.4KB, and with v.42bis (I think) that If all you are going to use it for is to call other such NON-STANDARDS based modems, the 14.4 mode is fine, but V.32bis is HERE even though the standard may not be 'finished'. You can get modems today that HAVE v.32bis, and you can get v.32 modems that have a simple guaranteed upgrade (proms) to v.32bis. There are many v.32s being sold that will NEVER be upgradable. If they can guarantee you that they will upgrade your modem to v.32bis for a modest price, then you *may* have the right modem. Else, I would worry some more before buying a non-standard (though popular) modem. It isn't just what will be in use in the next few months, but what will be most usable over the next few years that most people worry about when buying modems. V.32bis with v.42bis is what you will want until v.32ter (or whatever) surfaces. Another problem is that even 38.4kb is not enough now, and, unless you are using a good uart, 38.4 may not work well in many PCs. There are v.32bis w/v.42bis modems that wholesale probably well below $700. That means a good 'street' price may not be that much more. I expect the not-upgradable v.32 class modems to be super bargains in the immediate future. Some are already very inexpensive (they KNOW what their stuff is worth...). I remember paying over $2500. for 1200 baud cards when they first came out. Modems and their pricing always changes, but you can't wait forever.
bruce@ccavax.camb.com (Barton F. Bruce) (11/21/90)
In article <9Naus3w163w@isishq.fidonet.org>, evanc@isishq.fidonet.org (Evan Champion) writes: > I was planning on getting myself a US Robotics Courier HST/Dual Standard, as The DEC 11 issue of PC Magazine reviews that modem and several other popular ones. You will probably find it worth reading.
bill@unixland.uucp (Bill Heiser) (11/22/90)
In article <1990Nov19.191023.11581@nstar> larry@nstar (Larry Snyder) writes: > >The Telebit modems will hold the line better than the dual standard - > Interesting -- did you see PC Magazine's comments about the T2500? They "couldn't recommend it" because it wouldn't hold the line on noisy connections. (I don't mean to slam the T2500, I have one too ...) -- home: ...!{uunet,bloom-beacon,esegue}!world!unixland!bill bill@unixland.uucp, bill%unixland.uucp@world.std.com Public Access Unix - Esix SYSVR3 - (508) 655-3848 other: heiser@world.std.com Public Access Unix (617) 739-9753
bob@ns.UUCP (Robert J. Mathias) (11/23/90)
In article <9Naus3w163w@isishq.fidonet.org> evanc@isishq.fidonet.org (Evan Champion) writes: >I was planning on getting myself a US Robotics Courier HST/Dual Standard, as >it was the most highly recommended by my friends. However, since buying one >of those will "break the bank", I want to make sure I am buying the best for >my money. I like the Dual Standard's 14.4KB, and with v.42bis (I think) that >puts it around 30kb correct? Anyway, any help would be appreciated. > >Evan The answer to your inquire varies depending on your intended usage. If you plan to connect to PC based BBS's, you will find that most of them that support 9600 and above use USR HST (a number of these BBS's use the Dual Standard; thus support V32). If your intended use is for UUCP connections, I would recommend the TrailBlazer T2500 which supports (PEP, V32). If you plan to connect only to a system that has a V32 modem, I would go with either the Intel or Practical V32 modems. USR announced at COMDEX that they will be releasing a V32bis modem around December. It will come as V32bis only or Dual Standard modem. V32bis is the 14.4 CCITT proposed standard. One advandage to the USR version will be that when connected to another USR V32bis modem, after the modem has slowed to a lower speed (due to a bad line); the modem will attempt to go back up to the higher speed when the line improves. -- Robert J. Mathias, Jr uucp: ...!uunet!ccicpg!uis-oc!ns!bob Unisys Corporation voice: (714) 727-0323 A and V Series Systems Engineering fax: (714) 727-0350 Irvine, California
chip@chinacat.Unicom.COM (Chip Rosenthal) (11/23/90)
In article <1990Nov21.221114.11850@unixland.uucp> bill@unixland.uucp (Bill Heiser) writes: >Interesting -- did you see PC Magazine's comments about the T2500? They >"couldn't recommend it" because it wouldn't hold the line on noisy >connections. Then they did not run their tests in PEP mode. -- Chip Rosenthal 512-482-8260 | We was raising insurance premiums, ma. Unicom Systems Development | We was spreading fear of arson. <chip@chinacat.Unicom.COM> | - Michelle Shocked
larry@nstar.rn.com (Larry Snyder) (11/24/90)
bill@unixland.uucp (Bill Heiser) writes: >Interesting -- did you see PC Magazine's comments about the T2500? They >"couldn't recommend it" because it wouldn't hold the line on noisy >connections. That's not what happens in real applications - ask around - and you'll find Telebits maintaining links around the world. Did you see the review of 9600 baud modems in PC Week? The telebit's lost out primarily due to their lack of supporting a DTE speed of 38400 - while the USR V.32 and HST DS modems actually came out well.. -- Larry Snyder, Northern Star Communications, Notre Dame, IN USA {larry@nstar, {uunet|backbone}!nstar!larry, larry%nstar@iuvax.cs.indiana.edu} backbone usenet newsfeeds available Public Access Unix Site (219) 289-0282 (5 high speed lines)
tonyl@ivysoft.UUCP (Regular Tony Lin id as normal user) (11/25/90)
In article <9Naus3w163w@isishq.fidonet.org> evanc@isishq.fidonet.org (Evan Champion) writes: >I was planning on getting myself a US Robotics Courier HST/Dual Standard, as >it was the most highly recommended by my friends. However, since buying one >of those will "break the bank", I want to make sure I am buying the best for >my money. I like the Dual Standard's 14.4KB, and with v.42bis (I think) that >puts it around 30kb correct? Anyway, any help would be appreciated. > >Evan Two additional thoughts may help you buy your modem: 1. PC Magazine, December 11, 1990, has a good article comparing all high speed modems. They chose Multi-Tech's modem as the 'Editor's Choice.' 2. If you want to use your high speed modem for UUCP, wouldn't the TELEBIT be a better choice because it has PEP protocal support? But be aware of the fact that PC Magazine said it failed in their impaired line tests. Tony Lin
devil@techunix.BITNET (Gil Tene) (11/26/90)
> >Interesting -- did you see PC Magazine's comments about the T2500? They >"couldn't recommend it" because it wouldn't hold the line on noisy >connections. > Well, I wonder If they plugged it in right, or did thay plug it in at all? ;-) I don't know of PC Magazine tests, but I know this: I am using a TB2500 for an overseas connection 1/3 the way around the globe. The line is VERY noisy, it has my TB running at "only" 350-700 Bytes/sec. I can get a connection almost every time and hardly ever loose it. I have tried several other connections, including 1200,2400 (w and w/o MNP) and 9600 V.32, I have not been able to keep anything other than a TB2500 with PEP on the line for more than 30 seconds (actually, there was this one 300 baud modem...). PEP splits the line up to 512 different bands, and uses the clean ones. This makes it much more tolerant to line noise than normal modems. If TB2500 aren't good enough to hold the lines on noisy connections, then I would recommend against modems in general, since I haven't seen anything BETTER. -- Gil. -- -------------------------------------------------------------------- | Gil Tene "Some days it just doesn't pay | | devil@techunix.technion.ac.il to go to sleep in the morning." | --------------------------------------------------------------------
cpcahil@virtech.uucp (Conor P. Cahill) (11/26/90)
In article <1990Nov21.221114.11850@unixland.uucp> bill@unixland.uucp (Bill Heiser) writes: >Interesting -- did you see PC Magazine's comments about the T2500? They >"couldn't recommend it" because it wouldn't hold the line on noisy >connections. I'm not sure what mode they were using it in, but in PEP mode the T2500 holds the line better than any of the V.32s that I have tried (including the T2500 itself in v.32 mode). -- Conor P. Cahill (703)430-9247 Virtual Technologies, Inc., uunet!virtech!cpcahil 46030 Manekin Plaza, Suite 160 Sterling, VA 22170
msp33327@uxa.cso.uiuc.edu (Michael S. Pereckas) (11/26/90)
In <9842@discus.technion.ac.il> devil@techunix.BITNET (Gil Tene) writes: >> >>Interesting -- did you see PC Magazine's comments about the T2500? They >>"couldn't recommend it" because it wouldn't hold the line on noisy >>connections. >> >Well, I wonder If they plugged it in right, or did thay plug it in >at all? ;-) >I am using a TB2500 for an overseas connection 1/3 the way around >the globe. The line is VERY noisy, it has my TB running at "only" >350-700 Bytes/sec. I can get a connection almost every time >and hardly ever loose it. I have tried several other connections, >including 1200,2400 (w and w/o MNP) and 9600 V.32, I have not been >able to keep anything other than a TB2500 with PEP on the line >for more than 30 seconds (actually, there was this one 300 baud >modem...). PEP splits the line up to 512 different bands, and >uses the clean ones. This makes it much more tolerant to line noise >than normal modems. The PC Mag article is primarily about V.32. It is not clear to what extent they tested PEP. What is your experience with V.32? -- --------------------------------------------------------------------------- Michael Pereckas * InterNet: m-pereckas@uiuc.edu * just another student... (CI$: 72311,3246) Jargon Dept.: Decoupled Architecture--sounds like the aftermath of a tornado
altman@sbstaff2.cs.sunysb.edu (Jeff Altman) (11/26/90)
The PC Mag test was measuring the ability to hold the line when using V.32/V.42/V.42bis and not PEP. This makes all of the difference. - JEFF
david@twg.com (David S. Herron) (11/27/90)
In article <5@ivysoft.UUCP> tonyl@.UUCP (Regular Tony Lin id as normal user) writes: >In article <9Naus3w163w@isishq.fidonet.org> evanc@isishq.fidonet.org (Evan Champion) writes: >>I was planning on getting myself a US Robotics Courier HST/Dual Standard, as >>it was the most highly recommended by my friends. Depends.. do your friends know a lot about modems in general? > 2. If you want to use your high speed modem for UUCP, wouldn't the > TELEBIT be a better choice because it has PEP protocal support? > But be aware of the fact that PC Magazine said it failed in their > impaired line tests. Eh? How? By *ALL* reports I've heard, and *ALL* the testing I've done with telebit's they work excellently on "impaired lines". My first test was to pick up an extension line and start whistling and playing my Casio keyboard into the phone and etc. This was back in the "v4" days, surely it's gotten better with v5 ROMs? At any rate, the only effect was to slow the transmission down a bit .. *every* other modem I've worked with, admittadly only 300-2400 baud modems, would die immediately the first time a weird sound would be heard. There have been numerous reports here of Trailblazers being used to connect with *all* *sorts* of Weird Places -- Chile, India, over satellite to the Antarctica or ships at sea, etc. And it "just works" -- often these reports are accompanied by saying that some V.32 modem was tried first and couldn't even connect up. The advantages of PEP are basically that the modems are frequently sampling the line & adjusting themselves in *small* increments to fit the characteristics of the connection. This is good because the adjustments are much less drastic than on, for instance, V.32 where you go 9600->4800->dead, or even V.32bis where it will be 14.4->9.6->7.2->4.8->dead. It also helps in making and keeping connections in bad conditions. The only downfalls of PEP are 1-- it's not a "standard" .. regardless of the fact that it's a Very Good protocol. That's worth something. Especially now that V.32 chipsets are out allowing for these inexpensive V.32 modems coming on the market lately. 2-- It's half-duplex. The rapid-turnaround mode is a help, but I haven't ever used v5 ROMs so I don't know what improvements were in them. Still -- the UUCP spoofing is extra development cost & effort that wouldn't be necessary of it were full-duplex. -- <- David Herron, an MMDF & WIN/MHS guy, <david@twg.com> <- Formerly: David Herron -- NonResident E-Mail Hack <david@ms.uky.edu> <- <- Use the force Wes!
cec@cup.portal.com (Cerafin E Castillo) (11/28/90)
Just for your information, TELEBIT will be announcing a new, low-cost
V.32 modems in the next 1-2 months based on their own DSP implementation.
From what I've been told by TELEBIT, it will be a while before this new
modem has V.32bis or the V.32 Rockwell option (12 kbps) enabled in any of
its current, new, or future modems. With the T1500 currently at a
$1095 list price, and the T1000 at $795 list, I would look for the
new modem, T1600, to be at $895 list. Not much of a help to those
competing in the V.32 modem wars...
The T1600 is V.32 only, not PEP capable at all. It is the second generation
of TELEBIT technology. It makes room for some spectacular new products
that will change a lot of the way modems are used in bus-based systems.
The T1600 also features a 'true' 38.4 kbps interface.
Hope this helps.
===============================================================================
Cerafin E. Castillo || //\\ ||\\ ||
Network Consultant || //__\\ || \\ || Los Altos
Los Altos Networks || // ---\\|| \\|| Networks
340 Second St. #6 ||___// \ | \ |
Los Altos, CA 94022
(415) 941-8031 UUCP: {apple,sun,uunet}!portal!cup.portal.com!cec
INTERNET: cec@cup.portal.com
"...No hay mal que por bien no venga..."
===============================================================================mju@mudos.ann-arbor.mi.us (Marc Unangst) (11/28/90)
david@twg.com (David S. Herron) writes: > 2-- It's half-duplex. The rapid-turnaround mode is a help, but I haven't > ever used v5 ROMs so I don't know what improvements were in them. > Still -- the UUCP spoofing is extra development cost & effort that > wouldn't be necessary of it were full-duplex. Actually, I believe that each of the 512 subchannels that PEP uses can be either send or receive, so you *could* (in theory) split the line into two 6Kbps pieces and send data in both directions. The Trailblazer doesn't usually do this, because there's usually much more data flowing in one direction than the other. Also, the UUCP spoofing would be useful even if it was full-duplex. The UUCP 'g' protocol will top out at around 700cps on a 9600bps link, simply because of the protocol overhead and the time necessary for an ACK to get to the other end. With the UUCP spoofing, ACKs only have to go from the serial port to the modem, so things are much faster. -- Marc Unangst | mju@mudos.ann-arbor.mi.us | "Bus error: passengers dumped" ...!umich!leebai!mudos!mju |
john@jwt.UUCP (John Temples) (11/28/90)
In article <1990Nov26.062122.24546@sbcs.sunysb.edu> altman@sbstaff2.cs.sunysb.edu (Jeff Altman) writes: >The PC Mag test was measuring the ability to hold the line >when using V.32/V.42/V.42bis and not PEP. That's not what the article says. Quoting here: "The Telebit modem did not perform well under any of our degraded-line tests using either its V.32 or PEP mode." However, beyond this statement, there were no further references indicating that the reviewers used or tested PEP. I'd like to throw in that the article unauspiciously contained the following in its opening paragraph: "The latest modem turbochargers center on new V.32, V.42, and V.42bis error-checking and data-compression protocols, yielding throughput rates of 9,600, 19,600, [sic] and even 38,400 bits per second over dial-up telephone lines." Fortunately, they later pointed out that V.32 is not an error-checking or data-compression protocol. They also measured throughput with a nebulous "kilobits per second" rather than a more usable bytes per second. -- John W. Temples -- john@jwt.UUCP (uunet!jwt!john)
root@zswamp.fidonet.org (Geoffrey Welsh) (11/28/90)
Marc Unangst (mju@mudos.ann-arbor.mi.us ) wrote: >david@twg.com (David S. Herron) writes: > 2-- It's half-duplex. The rapid-turnaround mode is a help, but I haven't > ever used v5 ROMs so I don't know what improvements were in them. > Still -- the UUCP spoofing is extra development cost & effort that > wouldn't be necessary of it were full-duplex. >Actually, I believe that each of the 512 subchannels that PEP uses >can >be either send or receive, so you *could* (in theory) split the line >into two 6Kbps pieces and send data in both directions. The >Trailblazer doesn't usually do this, because there's usually much >more data flowing in one direction than the other. In theory, this could be done. It would probably be a vast improvement to the TB+ interactive capability if it split the channels based on dynamically monitored adta flow statistics. However, the TB does in fact send data in only one direction at a time. Some Telebit propoganda leads me to believe that there is a backchannel, but experience leads me to believe that it would be used only for PEP-related information (e.g. ACK/NAKing of PEP blocks) and not at all for user data. >Also, the UUCP spoofing would be useful even if it was full-duplex. >The UUCP 'g' protocol will top out at around 700cps on a 9600bps >link, >simply because of the protocol overhead and the time necessary for an >ACK to get to the other end. With the UUCP spoofing, ACKs only have >to go from the serial port to the modem, so things are much faster. Absolutely. I remember some old Microcom propoganda mentioning protocol spoofing, but I don't see it mentioned in their current Qx/V.32c spec sheet. -- UUCP: watmath!xenitec!zswamp!root | 602-66 Mooregate Crescent Internet: root@zswamp.fidonet.org | Kitchener, Ontario FidoNet: SYSOP, 1:221/171 | N2M 5E6 CANADA Data: (519) 742-8939 | (519) 741-9553 MC Hammer, n. Device used to ensure firm seating of MicroChannel boards Try our new Bud 'C' compiler... it specializes in 'case' statements!
hall@state.enet.dec.com (Dan Hall) (11/29/90)
In article <1990Nov28.042729.8077@jwt.UUCP>, john@jwt.UUCP (John Temples) |> From: john@jwt.UUCP (John Temples) |> Newsgroups: comp.dcom.modems |> Subject: Re: New Modems |> |> >The PC Mag test was measuring the ability to hold the line |> >when using V.32/V.42/V.42bis and not PEP. |> |> That's not what the article says. Quoting here: "The Telebit modem |> did not perform well under any of our degraded-line tests using |> either its V.32 or PEP mode." However, beyond this statement, there |> were no further references indicating that the reviewers used or |> tested PEP. |> You missed another small oblique reference. Quoting from page 329: "Unfortunately, the Telebit modems were not able to complete the data transfer tests under any of the EIA-standard impairment conditions. We saw similar problems using these modems under their native PEP mode." The article included 3-D bar graphs showing throughput of various file types, and except for the baseline tests, there were conspicuous holes where the Telebit performance should have been, because of their inability to complete any of the 1V32 through 6V32 impairment tests. BTW, someone asked recently what those tests are. They are combinational channel impairment tests established by the EIA for V.32 modems, as described in EIA/TIA standard 496-A. The baseline tests used by PC Magazine simulated a typical U.S. long-distance circuit, with the following impairment levels: Input level: -10 dBm Output level: -23 dBm IMD, 2nd-level: 52 dB IMD, 3rd-level: 50 dB Frequency shift: 1.25 Hz Phase jitter: 8 degrees No level for white noise was given (should have been about 32 dBrn), nor did they mention any gain or delay filtering (flat line?). |> John W. Temples -- john@jwt.UUCP (uunet!jwt!john) |> -Dan =_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_= Dan Hall | Telecommunications & Networks/EIC Digital Equipment Corporation | ARPAnet: hall@state.enet.dec.com Continental Blvd. | EASYnet: STATE::HALL MKO1-2/H10, PO Box 430 | Usenet : ....!decwrl!state.dec.com!hall Merrimack, NH 03054-0430 | N.E.T. : (603) 884-5879 =_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=_=
roberson@aurs01.UUCP (Charles "Chip" Roberson) (11/29/90)
In article <1990Nov26.062122.24546@sbcs.sunysb.edu> altman@sbstaff2.cs.sunysb.edu (Jeff Altman) writes: > >The PC Mag test was measuring the ability to hold the line >when using V.32/V.42/V.42bis and not PEP. > >This makes all of the difference. > >- JEFF I picked up the PC Mag article because I'm considering the purchase of a high speed modem. I was leaning towards the T2500 since I definitely want to support UUCP. However, I also want to have as much compatibility and throughput as possible with the rest of the bunch. The PC Mag article was good at clarifying the standards but it muddied the waters for me on which modem I want to purchase. In fact, I'm beginning to think there is no single modem that fits my needs. Could someone the significance of the T2500 not being able to hold the line when using V.32/V.42/V.42bis? Thanks, -chip * Work: 2912 Wake Forest Road, Raleigh, NC 27609 (919) 850-5011 * (...!mcnc!aurgate!roberson) || (roberson%aurgate@mcnc.org) || * (71500.2056@CompuServe.com) || (Chip.Roberson@f112.n151.z1.fidonet.org) #include <disclaimer.h>
max@mthvax.cs.miami.edu (Max Southall) (11/29/90)
In possible defence of Telebit and their modems' failure to measure up in the PC Mag tests, I believe it was Byte editor Phil Lemmons who wrote a special editorial a few years ago pointing out that reviews are often influenced by advertisers' input. Often, reviews are linked to advertising "spreads" by customers, and those not participating in the ad campaign often don't receive reviews at all, or ones they wouldn't want to quote to customers. At the time, Lemmons called for the computer press to clean up its dismal act, of which he gave numerous examples. At the time, Byte was a very different rag than it is now, not just a product review advertising manual, so I don't know if even Byte has succumbed. Certainly, Lemmons is no longer editor, and Steve Ciarcia with his real technical know-how was forced out too. Maybe we should ask Jerry Pournelle to do a *real* test!
pizzi@esacs.UUCP (Riccardo Pizzi) (11/29/90)
#include <std/disclaimer.h> In article <1715@chinacat.Unicom.COM> chip@chinacat.Unicom.COM (Chip Rosenthal) writes: >In article <1990Nov21.221114.11850@unixland.uucp> > bill@unixland.uucp (Bill Heiser) writes: >>Interesting -- did you see PC Magazine's comments about the T2500? They >>"couldn't recommend it" because it wouldn't hold the line on noisy >>connections. >Then they did not run their tests in PEP mode. I *never* trust magazine's tests and recensions... I am always afraid the tests could be bribed... Rick -- Riccardo Pizzi @ ESA Software, Rimini, ITALY e-mail: pizzi%esacs@relay.EU.net -or- root@xtc.sublink.org Public Access Unix @ +39-541-27858 (Telebit) << Object Oriented is an Opaque Disease >>
atman@ecst.csuchico.edu (Homeless hacker) (12/01/90)
In article <1990Nov29.040945.23924@mthvax.cs.miami.edu> max@mthvax.cs.miami.edu (Max Southall) writes: >forced out too. Maybe we should ask Jerry Pournelle to do a *real* test! What? All we would get would be 7 pages of cute anecdotes, his reasons for giving the various modems names like "Big Eunice," and no real data! (I saw no smiley, I assume you are serious.)
paul@frcs.UUCP (Paul Nash) (12/02/90)
Thus spake hall@state.enet.dec.com (Dan Hall): > You missed another small oblique reference. Quoting from page 329: > "Unfortunately, the Telebit modems were not able to complete the data > transfer tests under any of the EIA-standard impairment conditions. > We saw similar problems using these modems under their native PEP mode." While I am not a telecoms engineer, magazine scribe or salesperson, this is in direct conflict with my personal experiences. I live on a farm, so phone lines are not very good, and have three v22bis modems and a T2500. I have had various V32 modems aswell (some costing about double the price of the T2500) and have only found two reliable modems (that will connect at the highest speed possible and hang on until the end NOMATTER WHAT). These are the Racal vi2422 (v22bis) and the T2500. The others have all died on bad lines, either not connecting or losing the connection before a transfer (~20 minutes) is complete. The most reliable v32 modem (a MultiTech) had a dud connection rate of about 7 in 10 (ie: 7 duds in 10 attempts), with the best v22bis (Octocomm) getting about 5 in 10 (ie: 50% success). Your mileage might vary, but rather than relying in PC Magazine, try to get a few modems on loan, borrow, whatever and actually try them. I realise that this is an innovative concept, but it will tell you which modem will work on YOUR lines. ---=---=---=---=---=---=---=---=---=---=---=---=---=---=---=---=---=--- Paul Nash Flagship Wide Area Networks (Pty) Ltd paul@frcs.UUCP ...!uunet!ddsw1!proxima!frcs!paul
lstowell@pyrnova.pyramid.com (Lon Stowell) (12/05/90)
In article <2038@mountn.dec.com> hall@state.enet.dec.com writes: >EIA/TIA standard 496-A. The baseline tests used by PC Magazine simulated >a typical U.S. long-distance circuit, with the following impairment levels: > >Input level: -10 dBm >Output level: -23 dBm >IMD, 2nd-level: 52 dB >IMD, 3rd-level: 50 dB >Frequency shift: 1.25 Hz >Phase jitter: 8 degrees > >No level for white noise was given (should have been about 32 dBrn), nor Those are pretty benign impairment levels for a V.32 modem. Most I have tested (With the TAS 1010) tolerate CONSIDERABLY more noise than that...including single tone interference only a few dB below xmitter level at 1004, 2600, 1800 Hz... What DOES separate the world-class V.32's from the commodity products is their ability to tolerate amplitude jitter and phase roll (loosely described as an echo that is constantly being shifted in phase...) typical on satellite or transoceanic connections...although BOTH of these impairments are pretty rare in North America..and quite common on international circuits. Have yet to find a V.32 that can tolerate Gain Hits of more than a dB or so..even if the change results in staying totally within the dynamic range of the partner's receivers....
lstowell@pyrnova.pyramid.com (Lon Stowell) (12/05/90)
>In article <1990Nov29.040945.23924@mthvax.cs.miami.edu> max@mthvax.cs.miami.edu (Max Southall) writes: >>forced out too. Maybe we should ask Jerry Pournelle to do a *real* test! A major problem with typical "modem tests" is that the tester personell do not have knowledge of real world phone networks and the types of impairments that actually exist. An almost as common failure is lack of suitable protocol analysis equipment to determine whether failed connections are the fault of the modem or the DTE (and driving software) in the test. The EIA test suites address PART of the first issue...lack of Telco knowledge, but IMHO very inadequately. The second part DTE vs DCE requires expertise, as the techniques many modems use to STAY on an impaired connection will actually trigger (improperly written) disconnections on the part of the external machines used for the testing. Many so-called faulures are actually mismatches between the computer and the modems techniques. Use of a "reference" modem to guide results interpretations is HIGHLY recommended....but gaining a scientifically reliable reference point requires hundreds of man hours...as direct head-head comparisons on known impaired lines with a statistically significant sample are required....as 10 successive dial attempts between any given pair of end points are fairly likely to traverse 10 differing physical routes...and each attempt requires analysis to determine whether the modems or the DTE's caused the disconnect--and even then whether it was the modem's or DTE's (software typically) fault. In qualification testing at a prior employer for international V.32 modems, NONE of the Rockwell chipset units could compete on satellite circuits (simulated and actual) with the proprietary DSP implementations from the international vendors. This is NOT a negative comment on the Rockwell chips tho...because the variation between the best of the Rockwell based units and the proprietary DSP units was very slight...and on connections more typical of using barbed wire or railroad tracks as circuit carriers... The more non-intuitive results were between differing vendors using the SAME rev. level Rockwell chips... almost orders of magnitude difference in thru-put or connectivity on impaired ckts... Largely those vendors with strong backgrounds or engineers with real-world telco experience had far better products than those who learned phone lines by using a simulator or a textbook. ALL of the vendors used EXACTLY the same telco Simulator and series of tests...it is just that the use of a Telco network Simulator requires considerable expertise and interpretation...skills sadly lacking in many modem (AND P.C testing magazines..IMHO) vendors.
sichermn@beach.csulb.edu (Jeff Sicherman) (12/05/90)
Perhaps, some(one) ought to put together a test suite of real-world horror circuits for announced modems to be checked against for some extended period instead of going by only lab testing on simulators. Jeff Sicherman
jimmy_t@verifone.com (12/07/90)
In article <136549@pyramid.pyramid.com>, lstowell@pyrnova.pyramid.com (Lon Stowell) writes: > What DOES separate the world-class V.32's from the commodity > products is their ability to tolerate amplitude jitter and phase > roll (loosely described as an echo that is constantly being > shifted in phase...) typical on satellite or transoceanic > connections...although BOTH of these impairments are pretty rare > in North America..and quite common on international circuits. Which are the the "world-class" V.32 modems you've found? Thanks. -- +------------------------------------+--------------------------------------+ | James H. Thompson | jimmy_t@verifone.com (Internet) | | VeriFone Inc. | uunet!verifone!jimmy_t (UUCP) | | 100 Kahelu Avenue | 808-623-2911 (Phone) | | Mililani, HI 96789 | | +------------------------------------+--------------------------------------+