otto@tukki.jyu.fi (Otto J. Makela) (05/13/91)
In article <yy5s21w164w@halcyon.uucp> halcyon!ralphs@seattleu.edu (Ralph Sims) writes:
On a compressed feed, V.42bis is overkill, and may not provide any increase
in throughput, as it incorporates MNP-5 with an almost 4-1 compression of
data. It would smoke if you were to use uncompressed large files, but I
doubt that there'd be any increase if the file were compressed and you
used V.32. Compressing data which is already compressed is not a Good
Thing, and often results in files 'larger' than you started with. If
you're using V.42bis with compression turned on and if your files are
already compressed, try turning it off and see what happens (this reverts
you to 'standard' V.42 LAP-M).
This is fortunately incorrect. One of the best features of V.42bis is that it
can (unlike MNP5) disable the compression if it would make things worse. I'm
not sure if it can do this when connected to a MNP5 modem, but I doubt it.
I believe the comp.dcom.modems folks can clear things up here ?
--
/* * * Otto J. Makela <otto@jyu.fi> * * * * * * * * * * * * * * * * * * */
/* Phone: +358 41 613 847, BBS: +358 41 211 562 (USR HST/V.32, 24h/d) */
/* Mail: Kauppakatu 1 B 18, SF-40100 Jyvaskyla, Finland, EUROPE */
/* * * Computers Rule 01001111 01001011 * * * * * * * * * * * * * * * * */dalew@twiki.PDX.COM (Dale A. Weber) (05/13/91)
otto@tukki.jyu.fi (Otto J. Makela) writes: > This is fortunately incorrect. One of the best features of V.42bis is that i > can (unlike MNP5) disable the compression if it would make things worse. I'm > not sure if it can do this when connected to a MNP5 modem, but I doubt it. > I believe the comp.dcom.modems folks can clear things up here ? You get either MNP (with or without level 5) or V.42bis but not both. Yes, V.42bis is smart enough to turn itself OFF if compression would create larger transfers where MNP is not. If your modem has both V.42bis and MMP, that it will try to negotiate V.42bis first and then fall back and try MNP if that fails, and finally to a standard connect if both fail if you are using autoreliable mode. -- Internet: dalew@pdx.com OR dalew@twiki.pdx.com UUCP: ..!{ogicse, sun!nosun, tektronix}!tessi!twiki!dalew BBS: +1(503)239-4960 1200/2400 Bps [MNP5], 24 hours, PCPable via ORPOR WORK: Northwest Analytical, Inc. Voice: +1(503)224-7727
root@zswamp.uucp (Geoffrey Welsh) (05/13/91)
In a letter to All, Otto J. Makela (otto@tukki.jyu.fi ) wrote: >One of the best features of V.42bis is that it >can (unlike MNP5) disable the compression if it would make >things worse. I'm not sure if it can do this when >connected to a MNP5 modem, but I doubt it. There's a bit of confusion here. The V.42 spec includes, as an annex, MNP up to service class 4. The V42bis spec, as far as I know, doesn't necessarily imply MNP5 compatibility, though most V.42bis modems do offer it. Naturally, when a V.42bis modem calls an MNP-only modem, it must fall back to MNP protocols, including MNP data compression... and therefore loses the advantages of V.42bis data compression. -- Geoffrey Welsh - Operator, Izot's Swamp BBS (FidoNet 1:221/171) root@zswamp.uucp or ..uunet!watmath!xenitec!zswamp!root 602-66 Mooregate Crescent, Kitchener, ON, N2M 5E6 Canada (519)741-9553 "He who claims to know everything can't possibly know much" -me
bobl@graphics.rent.com (Bob Lindabury - SysAdm) (05/15/91)
otto@tukki.jyu.fi (Otto J. Makela) writes: > This is fortunately incorrect. One of the best features of V.42bis is that i > can (unlike MNP5) disable the compression if it would make things worse. I'm > not sure if it can do this when connected to a MNP5 modem, but I doubt it. > I believe the comp.dcom.modems folks can clear things up here ? If you connect to another V.42bis modem, it will disable compression if a file can't be compressed thusly avoiding the MNP5 compression of a file that is actually a little larger than the original compressed file. If a V.42bis modem connects to a MNP5 (remember, we are only talking compression schemes here) then the V.42bis modem *can* (some manufacturers didn't build in a fallback to an MNP5-like compression scheme) fallback to MNP5-type compression. My Hayes Ultra 96 does this with the Alternate compression protocol that is MNP-5 compatible. However, you are once again stuck with 2:1 compression and no checking for pre-compressed files sizes. In most cases, I would imagine that the V.42bis should be on and active as all sites transfer some files that are not compressed. News may be compressed but mail isn't. You might as well get that compression savings on the mail files. As for the news, I don't believe you will lose much if anything using a V.32 modem with V.42bis compression enabled. I haven't tested out the actual cps rates in different modes because I haven't felt that it was worth the time and effort since I am getting reasonable rates as it is. -- Bob The Graphics BBS 908/469-0049 "It's better than a sharp stick in the eye!" ============================================================================ InterNet: bobl@graphics.rent.com | Raven Enterprises UUCP: ...rutgers!bobsbox!graphics!bobl | 25 Raven Avenue BitNet: bobl%graphics.rent.com@pucc | Piscataway, NJ 08854 Home #: 908/560-7353 | 908/271-8878
admiral@admiral.UUCP (Doug Fields) (05/17/91)
bobl@graphics.rent.com (Bob Lindabury - SysAdm) writes: > file. If a V.42bis modem connects to a MNP5 (remember, we are only > talking compression schemes here) then the V.42bis modem *can* (some > manufacturers didn't build in a fallback to an MNP5-like > compression scheme) fallback to MNP5-type compression. My Hayes Not quite. While V.42 does include MNP1-4, V.42bis does NOT include MNP5. It is usually an additional feature of V.42bis modems to include MNP5 but that is NOT part of the V.42bis spec from the CCITT. > compression savings on the mail files. As for the news, I don't > believe you will lose much if anything using a V.32 modem with > V.42bis compression enabled. I haven't tested out the actual cps > rates in different modes because I haven't felt that it was worth the > time and effort since I am getting reasonable rates as it is. You will not get a slowdown... This is from a test with two V.42bis HST's using Zmodem with V.42bis on and off on a .Zipped file. The modem will keep on chugging at around 1650 cps either way. With the standard unix compress, which uses almost the exact same algorithm, it would be close to impossible to get any further compression. Doug Fields -- 100 Midwood Road, Greenwich, CT 06830 --- (FAX) +1 203 661 2996 uucp: uunet!sir-alan!admiral!doug --- Thank you sir-alan/mail and wizkid/news! Internet: fields-doug@cs.yale.edu --------------- (Voice@Yale) +1 203 436 0184 BBS: (HST/V32) +1 203 661 1279; (MNP6) -2967; (PEP/V32) -2873; (V32/V42) -0450
bobl@graphics.rent.com (Bob Lindabury - SysAdm) (05/19/91)
admiral@admiral.UUCP (Doug Fields) writes: > bobl@graphics.rent.com (Bob Lindabury - SysAdm) writes: > > > file. If a V.42bis modem connects to a MNP5 (remember, we are only > > talking compression schemes here) then the V.42bis modem *can* (some > > manufacturers didn't build in a fallback to an MNP5-like > > compression scheme) fallback to MNP5-type compression. My Hayes > > Not quite. While V.42 does include MNP1-4, V.42bis does NOT include MNP5. > It is usually an additional feature of V.42bis modems to include MNP5 but > that is NOT part of the V.42bis spec from the CCITT. Ah...let's look at the above paragraphs. Don't I clearly state that a V.42bis modem ***** can ***** fallback to MNP-5 if the modem manufacturer built in this ***** special ***** addition to the modem? I specifically didn't state that it was a feature of V.42bis because it isn't. It is the modem manufacturers option to provide a fallback mode. I think we are in total agreement here and I think my paragraph states that. -- Bob The Graphics BBS 908/469-0049 "It's better than a sharp stick in the eye!" ============================================================================ InterNet: bobl@graphics.rent.com | Raven Enterprises UUCP: ...rutgers!bobsbox!graphics!bobl | 25 Raven Avenue BitNet: bobl%graphics.rent.com@pucc | Piscataway, NJ 08854 Home #: 908/560-7353 | 908/271-8878
lancelot@UUCP (Thor Lancelot Simon) (05/25/91)
In article <1991May22.145955.20039@bluemoon.uucp> grant@bluemoon.uucp (Grant DeLorean) writes: >bobl@graphics.rent.com (Bob Lindabury - SysAdm) writes: > >>There are several levels of MNP above 5 all the way up to 9. These >>are not generally supported by consumer modems and are hard to come >>by. > > There is no MNP 6 or MNP 8, there are MNP 7 and MNP 9 though. The >reason the latter two are so unusual is that Microcom wants >exhorbitant amounts of money to license them, so very few modems >but their own use them (and none of the ones that do have them are >inexpensive consumer grade modems). This is *absolutely* false. I type this while logged in using a pair of MNP 6 modems. In fact, MNP 6 is a modulation scheme, not an error-control protocol. I am fairly sure it's one of the many variants on v.29ft (or is that 27ft?) a 9600 bps main channel with a 300 bps backchannel. File transfer performance is about like the old HST protocol, which figures. I am pretty sure MNP 8 exists. I've also heard of MNP 10 being used in Microcom's specialized cellmodems, but I don't have any confirmation for this. Followup to comp.dcom.modems. -- ******************************************************************************* *Thor Simon * Okay, just a little pin-prick...There'll be no more-* *lancelot@spock.UUCP * Aieeeeaaaugh!-but you may feel a little _sick_. * *decwrl!spock!lancelot * ---Pink Floyd *
tnixon@hayes.uucp (05/28/91)
In article <1991May25.022520.4101@spock.UUCP>, lancelot@UUCP (Thor Lancelot Simon) writes: > In article <1991May22.145955.20039@bluemoon.uucp> grant@bluemoon.uucp (Grant DeLorean) writes: >>bobl@graphics.rent.com (Bob Lindabury - SysAdm) writes: >> >> [various contentions about MNP classes above 5] > > This is *absolutely* false. I type this while logged in using a pair of MNP 6 > modems. In fact, MNP 6 is a modulation scheme, not an error-control protocol. > I am fairly sure it's one of the many variants on v.29ft (or is that 27ft?) a > 9600 bps main channel with a 300 bps backchannel. File transfer performance > is about like the old HST protocol, which figures. I am pretty sure MNP 8 > exists. I've also heard of MNP 10 being used in Microcom's specialized > cellmodems, but I don't have any confirmation for this. Followup to > comp.dcom.modems. Mr. Simon is basically right. Microcom has defined up through Class 10. Here is a brief summary of the MNP Classes above 5: MNP6: V.29-based fast-turnaround ping-pong modulation scheme. MNP7: Adaptive Huffman compression scheme based on character pairs rather than single characters for which throughput is claimed up to 3-to-1. MNP8: Announced, but never shipped. MNP9: Several minor protocol enhancements, including piggybacked acknowledgements and multiple selective reject. MNP10: "Adverse Link" enhancements, including repacketizing of data into smaller frames during retransmission, higher number of retries before giving up, automatic retraining including changes in modulation scheme, and automatic adjustable carrier level. I'm only aware of one other company that ever made an MNP6 modem (Multitech), and don't know of ANY others that have done MNP7 or MNP9. Microcom has licensed MNP10 for cellular telephone modems to Rockwell to make chipsets and to a couple of other companies, and intends to license it widely. The CCITT is studying data transmission on cellular networks, and may adopt similar techniques for enhancement of LAPM (but won't adopt MNP10 because the V.42 "alternative protocol" is officially "frozen"). -- Toby Nixon, Principal Engineer | Voice +1-404-840-9200 Telex 151243420 Hayes Microcomputer Products Inc. | Fax +1-404-447-0178 CIS 70271,404 P.O. Box 105203 | UUCP uunet!hayes!tnixon AT&T !tnixon Atlanta, Georgia 30348 USA | Internet hayes!tnixon@uunet.uu.net