s142075@fred.ucdavis.edu (Brewski Rogers) (06/11/91)
dan the man zerkle says: >o There's a catch. No, I don't know what the catch is. However, > there must be some reason that this thing sells for a third of > the price of comparable equipment. If the "standards" people > could have come up with something that offers the same perfformance > at a considerably lower cost, why didn't they? Same reason people buy MACS: becuase they're stupid. The catch is that they are non-standard and the company that makes them is slow in delivering. full-duplex 9600 bps is just about the dumbest thing I ever heard of. Anyone who NEEDS to send AND receive at full speed (as opposed to 1 9600 bps and 1 450 bps channel) please respond. I'd be glad to hear what all that extra technology is being used for. -bruce
rkushner@sycom.UUCP (Ronald Kushner) (06/12/91)
s142075@fred.ucdavis.edu (Brewski Rogers) writes: >dan the man zerkle says: >>o There's a catch. No, I don't know what the catch is. However, >> there must be some reason that this thing sells for a third of >> the price of comparable equipment. If the "standards" people >> could have come up with something that offers the same perfformance >> at a considerably lower cost, why didn't they? > >Same reason people buy MACS: becuase they're stupid. > >The catch is that they are non-standard and the company that makes >them is slow in delivering. > >full-duplex 9600 bps is just about the dumbest thing I ever heard of. >Anyone who NEEDS to send AND receive at full speed (as opposed to >1 9600 bps and 1 450 bps channel) please respond. I'd be glad to hear >what all that extra technology is being used for. For one thing, the HST doesn't see to do very well with protocals like the G protocal used for UUCP where it sends an ack and nak after every packet sent. Its always trying to allocate the high speed channel back and forth. If they would have added support for these protocals in the firmware like the Telebit did, they might have sold more modems to people who are UUCPing..I don't think the Telebit is full duplex, but it sure works alot better than the HST. Another problem is ASCII sending to something that echos it back to you, your only getting 450CPS, or a bottleneck somewhere. Plus its getting so cheap, why not have full duplex 14400. USR mailed me the new sysop package, and v.32bis wasn't much more than the HST($50 I believe). -- C-UseNet V0.42d Ronald Kushner Life in Hell BBS +1 (313) 939-6666 P.O. Box 353 14400 USR HST V.42 & V.42bis Sterling Heights, MI 48311-0353 Complete Amiga Support UUCP: uunet!umich!vela!sycom!rkushner (We are not satanic, just NUTS!) DISCLAIMER: I say what I mean, and mean what I say.
ddj@zardoz.club.cc.cmu.edu (Doug DeJulio) (06/14/91)
In article <9154@ucdavis.ucdavis.edu> s142075@fred.ucdavis.edu (Brewski Rogers) writes: >full-duplex 9600 bps is just about the dumbest thing I ever heard of. >Anyone who NEEDS to send AND receive at full speed (as opposed to >1 9600 bps and 1 450 bps channel) please respond. I'd be glad to hear >what all that extra technology is being used for. Sure. We're using our full-duplex 19.2kbaud modem for our house's lifeline to the internet. While I'm telnetted in to send this message, another user could be telnetted out to do a compile on one of the schools computers, and two other users could have FTP sessions going, in both directions. -- Doug DeJulio ddj@zardoz.club.cc.cmu.edu
rhealey@kas.helios.mn.org (Rob Healey) (06/14/91)
In article <9154@ucdavis.ucdavis.edu> s142075@fred.ucdavis.edu (Brewski Rogers) writes: >dan the man zerkle says: >>o There's a catch. No, I don't know what the catch is. However, >> there must be some reason that this thing sells for a third of >> the price of comparable equipment. If the "standards" people >> could have come up with something that offers the same perfformance >> at a considerably lower cost, why didn't they? > They get 9600 baud on a good day with the wind blowing in the right direction... I.e. they are 2400 baud modems with V.42bis compression that can get 4:1 compression when compressing a file of zero's... VERY misleading advertising. Most of the time you'd be lucky to get 4800 baud throughput on these things and when you did the interactive performance would be VERY jerky. >full-duplex 9600 bps is just about the dumbest thing I ever heard of. >Anyone who NEEDS to send AND receive at full speed (as opposed to >1 9600 bps and 1 450 bps channel) please respond. I'd be glad to hear >what all that extra technology is being used for. > You must not work at a place that needs to transfer alot of data quickly over common carrier phone lines... V.32 allows for bidirectional transfer at 9600 baud. Now personally, I prefer my T2500 that does 19K or so in one direction and <110 on the ACK channel. But there are many reasons in business to transfer 9600+ baud in both directions at once, i.e. updating databases while loading the receipts from a days sales. Full-duplex 9600 isn't dumb, it's just that most hobbiests and students/academics wouldn't see any obvious use. -Rob
matt@vrtwo.UUCP (Matt Buford) (06/18/91)
writes: >In article <9154@ucdavis.ucdavis.edu> s142075@fred.ucdavis.edu (Brewski Rogers >>full-duplex 9600 bps is just about the dumbest thing I ever heard of. >>Anyone who NEEDS to send AND receive at full speed (as opposed to >>1 9600 bps and 1 450 bps channel) please respond. I'd be glad to hear >>what all that extra technology is being used for. > >Sure. We're using our full-duplex 19.2kbaud modem for our house's >lifeline to the internet. While I'm telnetted in to send this >message, another user could be telnetted out to do a compile on one of >the schools computers, and two other users could have FTP sessions >going, in both directions. I have a USR 14.4K DS coming in the mail and my feed (sycom) has a 14.4 HST (half-duplex). Well, he tells me that when our modems connect at HST that because of the echo checksums that UUCP uses my throughput will be UNDER 2400 BAUD on my 14400 baud modem... NOT good. Also, ever heard of bimodem? -- .____________________________________________________________________________. | | | | Please send all | Sysop of the Virtual Reality BBS | | complaints to | uunet.uu.net!umich!vela!sycom!vrtwo!matt | | | 808-337-1560 * 2400 baud * C-Net Amiga | | >NIL: | USR 14400 baud DS on order - 40 megs storage | |__________________________|_________________________________________________|
dillon@overload.Berkeley.CA.US (Matthew Dillon) (06/21/91)
In article <matt.0215@vrtwo.UUCP> matt@vrtwo.UUCP (Matt Buford) writes: > >I have a USR 14.4K DS coming in the mail and my feed (sycom) has a 14.4 HST >(half-duplex). Well, he tells me that when our modems connect at HST that >because of the echo checksums that UUCP uses my throughput will be UNDER 2400 >BAUD on my 14400 baud modem... NOT good. This is a well known bug with the HST, it tries to turn its half duplex connection around on even DNet's 3 BYTE ACKS!! Talk about screwups... The only thing that works well is X,Y,ZMODEM. >Also, ever heard of bimodem? >-- >.____________________________________________________________________________. >| | | >| Please send all | Sysop of the Virtual Reality BBS | >| complaints to | uunet.uu.net!umich!vela!sycom!vrtwo!matt | >| | 808-337-1560 * 2400 baud * C-Net Amiga | >| >NIL: | USR 14400 baud DS on order - 40 megs storage | >|__________________________|_________________________________________________| -Matt -- Matthew Dillon dillon@Overload.Berkeley.CA.US 891 Regal Rd. uunet.uu.net!overload!dillon Berkeley, Ca. 94708 USA
bruce@zuhause.MN.ORG (Bruce Albrecht) (06/22/91)
>full-duplex 9600 bps is just about the dumbest thing I ever heard of. >Anyone who NEEDS to send AND receive at full speed (as opposed to >1 9600 bps and 1 450 bps channel) please respond. I'd be glad to hear >what all that extra technology is being used for. If UUCP and other communications programs where smart enough, you'd be sending data both directions at 9600 baud, and include packet acknowledgements as part the packet headers. Actually, since V.32 modems set to synchronous mode, they can be used with X.25 to do exactly that. Tymnet (now BT Tymnet) developed a protocol based on X.25 called X.PC around 1983 which allows up to 15 independent channels over a serial line. If people supported this for BBSs, you'd be able to upload, download, and read mail simulatneously with only a slight amount of overhead. Although most traffic seems to be one-way, I'm sure people would be willing to change if something like X.PC became widely used. -- bruce@zuhause.mn.org
dillon@overload.Berkeley.CA.US (Matthew Dillon) (06/24/91)
In article <bruce.3753@zuhause.MN.ORG> bruce@zuhause.MN.ORG (Bruce Albrecht) writes: >>full-duplex 9600 bps is just about the dumbest thing I ever heard of. >>Anyone who NEEDS to send AND receive at full speed (as opposed to >>1 9600 bps and 1 450 bps channel) please respond. I'd be glad to hear >>what all that extra technology is being used for. > >If UUCP and other communications programs where smart enough, you'd be sending >data both directions at 9600 baud, and include packet acknowledgements as part >the packet headers. Actually, since V.32 modems set to synchronous mode, they can >be used with X.25 to do exactly that. Actually, UUCP does use piggy-backed ACKs, but since file transfers only occur in one direction at once they are pretty useless, only used during a turn-around, and not implemented properly in ports of UUCP G (specifically, they blow up a trailblazer in spoofing mode and most PC UUCPs) >Tymnet (now BT Tymnet) developed a protocol based on X.25 called X.PC around >1983 which allows up to 15 independent channels over a serial line. If people >supported this for BBSs, you'd be able to upload, download, and read mail >simulatneously with only a slight amount of overhead. Although most traffic >seems to be one-way, I'm sure people would be willing to change if something >like X.PC became widely used. > >-- > >bruce@zuhause.mn.org I'd have to look at the X.PC spec, if they put it through the grinder just to make it work on PCs without multitasking it probably isn't worth it. If it sticks to X.25 it probably is worth it. -Matt -- Matthew Dillon dillon@Overload.Berkeley.CA.US 891 Regal Rd. uunet.uu.net!overload!dillon Berkeley, Ca. 94708 USA
matt@vrtwo.UUCP (Matt Buford) (06/24/91)
writes: >In article <matt.0215@vrtwo.UUCP> matt@vrtwo.UUCP (Matt Buford) writes: >> >>I have a USR 14.4K DS coming in the mail and my feed (sycom) has a 14.4 HST >>(half-duplex). Well, he tells me that when our modems connect at HST that >>because of the echo checksums that UUCP uses my throughput will be UNDER 2400 >>BAUD on my 14400 baud modem... NOT good. > > This is a well known bug with the HST, it tries to turn its half > duplex connection around on even DNet's 3 BYTE ACKS!! Talk about > screwups... The only thing that works well is X,Y,ZMODEM. Well I already sent you mail about this, but I'll put it here so others can see it too... Will increasing the packet size fix or at least help this? -- .____________________________________________________________________________. | | | | Please send all | Sysop of the Virtual Reality BBS | | complaints to | uunet.uu.net!umich!vela!sycom!vrtwo!matt | | | 808-337-1560 * 2400 baud * C-Net Amiga | | >NIL: | USR 14400 baud DS on order - 40 megs storage | |__________________________|_________________________________________________|
billsey@agora.UUCP (Bill Seymour) (06/25/91)
In article <9154@ucdavis.ucdavis.edu>, Brewski Rogers writes: > > full-duplex 9600 bps is just about the dumbest thing I ever heard of. > Anyone who NEEDS to send AND receive at full speed (as opposed to > 1 9600 bps and 1 450 bps channel) please respond. I'd be glad to hear > what all that extra technology is being used for. > I think anyone who's using a 9600 baud modem for Kermit or XModem transfers will say they NEED full duplex at full speed. It's only when you're in ZModem or WXmodem type transmissions that you don't lose performance with the half duplex modems. > -bruce BTW, I believe the modem the original article was refering to (a '9600' baud modem for $169) was really a 2400 baud modem with V.42bis. That means you can use a 9600 baud DTE rate, and theoretical throughput can be as high as 9600 baud. Fits the price range anyway... -Bill Seymour nesbbx!billsey@agora.uucp or nesbbx!billsey@agora.rain.com ***** American People/Link Amiga Zone Hardware Specialist NES*BILL ***** Bejed, Inc. NES, Inc. NAG BBS NES BBX BBS Home Sometimes (503)281-8153 (503)246-9311 (503)656-7393 (503)640-9337 (503) 640-0842
dillon@overload.Berkeley.CA.US (Matthew Dillon) (06/26/91)
In article <matt.0287@vrtwo.UUCP> matt@vrtwo.UUCP (Matt Buford) writes: >> >> This is a well known bug with the HST, it tries to turn its half >> duplex connection around on even DNet's 3 BYTE ACKS!! Talk about >> screwups... The only thing that works well is X,Y,ZMODEM. > >Well I already sent you mail about this, but I'll put it here so others can >see it too... > >Will increasing the packet size fix or at least help this? Did I miss the mail? In anycase, increasing the packet size would help transfers a little but no matter what you do the windowing gets f#@$d up. As I said, I would pound on HST to fix *their* problem, it's a major deficiency in their product. -Matt -- Matthew Dillon dillon@Overload.Berkeley.CA.US 891 Regal Rd. uunet.uu.net!overload!dillon Berkeley, Ca. 94708 USA
lron@easy.lrcd.com (Dwight Hubbard) (06/26/91)
In article <matt.0287@vrtwo.UUCP> matt@vrtwo.UUCP (Matt Buford) writes: > > writes: >>In article <matt.0215@vrtwo.UUCP> matt@vrtwo.UUCP (Matt Buford) writes: >>> >>>I have a USR 14.4K DS coming in the mail and my feed (sycom) has a 14.4 HST >>>(half-duplex). Well, he tells me that when our modems connect at HST that >>>because of the echo checksums that UUCP uses my throughput will be UNDER 2400 >>>BAUD on my 14400 baud modem... NOT good. >> >> This is a well known bug with the HST, it tries to turn its half >> duplex connection around on even DNet's 3 BYTE ACKS!! Talk about >> screwups... The only thing that works well is X,Y,ZMODEM. > >Well I already sent you mail about this, but I'll put it here so others can >see it too... > >Will increasing the packet size fix or at least help this? Yes, assuming of course the other end supports the new packet size (generally not to likely). Even with the larger packet sizes though throughput doesn't generally exceed the speeds of a V.32 modem, so if your going to modify the protocol you would be best off to get Zmodem running between the two systems and let the modem run without using the back channel at all.
lron@easy.lrcd.com (Dwight Hubbard) (06/26/91)
In article <1959c56e.ARN01fc@nesbbx.UUCP> nesbbx!billsey@agora.UUCP (Bill Seymour) writes: >In article <9154@ucdavis.ucdavis.edu>, Brewski Rogers writes: >> >> full-duplex 9600 bps is just about the dumbest thing I ever heard of. >> Anyone who NEEDS to send AND receive at full speed (as opposed to >> 1 9600 bps and 1 450 bps channel) please respond. I'd be glad to hear >> what all that extra technology is being used for. >> > I think anyone who's using a 9600 baud modem for Kermit or XModem >transfers will say they NEED full duplex at full speed. It's only when >you're in ZModem or WXmodem type transmissions that you don't lose performance ^^^^^^^ WXmodem is a windowed protocol and it uses the back channel as much as Xmodem does, the protocol I think your talking about is YModem-G which is a streaming protocol. > BTW, I believe the modem the original article was refering to >(a '9600' baud modem for $169) was really a 2400 baud modem with V.42bis. >That means you can use a 9600 baud DTE rate, and theoretical throughput >can be as high as 9600 baud. Fits the price range anyway... Nope, Compucom makes a 9600 asymetric modem with MNP-5 and I think V.42bis as well and it has a price of $169.
civir1070@ucsvax.sdsu.edu (FURRY R) (06/27/91)
In article <1959c56e.ARN01fc@nesbbx.UUCP>, nesbbx!billsey@agora.UUCP (Bill Seymour) writes... > > BTW, I believe the modem the original article was refering to >(a '9600' baud modem for $169) was really a 2400 baud modem with V.42bis. >That means you can use a 9600 baud DTE rate, and theoretical throughput >can be as high as 9600 baud. Fits the price range anyway... Hate to disagree, but... The ad in BYTE (ok,ok, I got 3 free issues... then I canceled...honest!) says: 9600-38400 bps modem...$169 "NOW you can afford a _speedmodem_. The _champ_ has a raw speed of 300-9600 bps and 4:1 data compression for throughput up to 38,400 bps." Seems like it is a 9600 baud modem with compression. Too bad it's only an IBM internal. :( ___________________________________________________________________ / |CIVIR1070@ucsvax.sdsu.edu| Q: Is there a UNIX FORTRAN optomizer? \ \ | Scott Ellis | A: Yeah, "rm *.f" _ / / |_________________________| _ // Amiga \ \_______________________________________________\X/_________________/
lron@easy.lrcd.com (Dwight Hubbard) (06/27/91)
In article <dillon.9021@overload.Berkeley.CA.US> dillon@overload.Berkeley.CA.US (Matthew Dillon) writes: >In article <matt.0287@vrtwo.UUCP> matt@vrtwo.UUCP (Matt Buford) writes: >>> >>> This is a well known bug with the HST, it tries to turn its half >>> duplex connection around on even DNet's 3 BYTE ACKS!! Talk about >>> screwups... The only thing that works well is X,Y,ZMODEM. >> >>Well I already sent you mail about this, but I'll put it here so others can >>see it too... >> >>Will increasing the packet size fix or at least help this? > > Did I miss the mail? > > In anycase, increasing the packet size would help transfers a little > but no matter what you do the windowing gets f#@$d up. > > As I said, I would pound on HST to fix *their* problem, it's a major > deficiency in their product. Not really, it was marketed to produce high speed transfers for BBS users at a time when the echo cancelation for V.32 was to expensive. US Robotics wasn't the only company to do this either, Telebit, Hayes and a few other companies implemented either asymetric or half duplex protocols similar to HST for the same reasons. Now US Robotics is pushing the V.32bis modems out the door at only slightly higher prices than the HST modems in an effort to phase them out of the market. -- ---------------------------------------------------------------------- -Dwight Hubbard INTERNET: lron@easy.lrcd.com - -Kaneohe, Hawaii USENET : ...!uunet!easy!lron - - BIX : lron - ----------------------------------------------------------------------
dillon@overload.Berkeley.CA.US (Matthew Dillon) (06/29/91)
In article <lron.5492@easy.lrcd.com> lron@easy.lrcd.com (Dwight Hubbard) writes: >In article <dillon.9021@overload.Berkeley.CA.US> dillon@overload.Berkeley.CA.US (Matthew Dillon) writes: >> >> As I said, I would pound on HST to fix *their* problem, it's a major >> deficiency in their product. > >Not really, it was marketed to produce high speed transfers for BBS users >at a time when the echo cancelation for V.32 was to expensive. US Robotics >wasn't the only company to do this either, Telebit, Hayes and a few other I was refering to the algorithm they chose to switch the high-speed direction.... it's way too sensitive. All that really needs to be done is to make it a bit less sensitive. -Matt -- Matthew Dillon dillon@Overload.Berkeley.CA.US 891 Regal Rd. uunet.uu.net!overload!dillon Berkeley, Ca. 94708 USA