kowals@studsys.mu.edu (kowalski) (02/21/89)
I am posting the following letter for my employer SST Data. Any replys will be given to them and they will reply thru this account : The U.S. Robotics Courier HST/ix modem/software package for UNIX systems has been the subject of recent controversy on USENET. The Courier HST/ix provides high speed dial-up communications for a variety of UNIX operating environments. The modem uses V.32-style trellis coded modulation at 9600 bps, which, combined with a very efficient compression scheme in the bundled datacom software package, results in error-free throughput up to 30,000 bps. U.S. Robotics apparently has struck a nerve with Telebit Corporation, whose multicarrier modem is an entrant in the UNIX market. Michael Ballard of Telebit recently issued a notice of USENET ostensibly to set the record straight on the USR Courier HST/ix modem. It's evident that Mr. Ballard is the one who needs straightening out. Ballard asserts that he helped design the datacom package bundled with the Courier HST/ix, and that is was named for the Trailblazer modem. Both of these assertions are untrue. Mike Myers, VP of Engineering for SST Data Inc., Mequon, WI, is the program's sole author. Myers says that Ballard "was not involved at all in the design of the program." The program's original brand name, "Handshake Blazer," was selected through an employee contest at SST. Myers says any similarity to Telebit's product nomenclature is coincidental. Ballard also claimed that the Courier HST/ix is not available for a number of systems, specifically Sun, 3B5/15, AIX, Pyramid, Altos and Apple. This, too, is untrue. In addition to the preceding systems Courier HST/ix is available for Fortune, AT&T 3B1/2 and 6386, SCO Xenix, Interactive UNIX, Unisys 5000/30, 50, 60, 85, 90, IBM RT and 6150, Plexus and NCR Tower 32 and XP. Additionally, Ballard tried to make an issue of the Courier HST/ix's performance with UUCP. U.S. Robotics has never claimed that the Courier HST/ix improves the performance of UUCP. The HST/ix, though, does make UUCP much easier to use. Mike Myers of SST says the HST/ix software was designed as a high speed alternative to UUCP that provides benefits such as the ability to restart a file transfer from the point of interruption (a capability not offered by UUCP). For any additional information on the Courier HST/ix, please call USR's Director of North American Sales, Bob Polychron at 312-982-5233. Or call Mike Myers, Vice President of Engineering, at SST Data, Inc. at 414-242- 3999. With a retail price of $1,295, the Courier HST/ix modem/software package delivers the best price-performance ratio available to users of UNIX systems.
karl@ddsw1.MCS.COM (Karl Denninger) (02/23/89)
In article <415@studsys.mu.edu> kowals@studsys.mu.edu (kowalski) writes: > >I am posting the following letter for my employer SST Data. Any replys >will be given to them and they will reply thru this account : Good! I'm sure they'll like this. :-) >The U.S. Robotics Courier HST/ix modem/software package for UNIX systems >has been the subject of recent controversy on USENET. > >.......The modem uses V.32-style >trellis coded modulation at 9600 bps, which, combined with a very >efficient compression scheme in the bundled datacom software package, >results in error-free throughput up to 30,000 bps. But it is not V.32 if I remember correctly (or is the modem truly V.32 capable?). This means you're not compatible with V.32, or the Telebit PEP installed base. For Unix use, the fact that every Bow-wow BBS (Fido) has one is 100% irrelavent when making a purchase decision. (advertisement and complaint/argument with Mike Ballard @ Telebit deleted) >With a retail price of $1,295, the Courier HST/ix modem/software >package delivers the best price-performance ratio available to users of >UNIX systems. Not from what I can see. Note also that the software, from what I see in the configuration info, will talk nicely to a Telebit as well! (note: Telebits run 14kbaud max uncompressed data rate, HSTs run 9.6kbaud uncompressed.... guess which modem should do better with the HST/ix software?) I can buy two Telebits as an Internet site for the cost of ONE of your Courier HST/ix's. Earlier as a Usenet site I could do the same thing. The cost of the HST/ix solution is effectively twice the price of the Telebit way. Guess how many Telebits are out there? Nearly EVERYONE we talk to has at least one; they're nearly universal in the Usenet world! HST/ix is up against the same thing here (actually worse, you need the special software too) in the Usenet universe that Telebit is in the Fidonet universe. HST/ix may do wonderful things. We have it here and were a beta site; it does work. 30kbaud is a dream though for anyone running News -- news batches are already compressed. I'd believe 30kbaud if the file is all zeros or something similar 1/2 :-). Another IMPORTANT point: the software we got was COPY PROTECTED (seriously -- on a Unix system!!!!!); I have no idea if the production version is. It's currently offline on tape; if someone wants to try it over long distance I'll reload it providing it still runs once it's "restored". In any event I would certainly ask USR or whoever about that COPY PROTECTION before ordering the product! From what I can see HST/ix (the kit) has two potentially-crippling problems: o The HST modems give some 2400 baud non-MNP capable units, and all AT&T 3b1's, fits. They simply won't connect with them AT ALL if you have high-speed enabled (that is, not &M0). Thus you are immediately incompatible with about 50% of the installed base with a Unix system if it's your only modem, unless you want to poll out only. In addition, there are some things about the HST's functionality that I would have to consider bugs, such as no 9600 baud if you disable MNP mode -- there is no "disable MNP on slow calls only". The Telebit has this, and it's a lifesaver when you need to be 100% compatible with the "slow people". For this reason alone HST/ix is not a "sole solution". o There aren't many out there, and it's expensive. $1295 is a heck of a lot of money for a special-use product; it had better do something REALLY nice. The Telebit modems are 100% compatible with the non-MNP units, no hassles there, and they talk PEP as well (for high speed). Finally, 30kbaud over HST/ix isn't even close to reality in my experience, although I'd be happy to let someone try to prove me wrong there. A note: o We only tested it locally (in-building) between two Xenix systems on two local phone lines. I have no idea how well (or if) it copes with long-distance telephone lines and/or overseas connections. USR didn't feel like giving out names of other sites across the country for us to poll..... nor did they tell anyone else to try polling us (although I did tell them we'd be more than happy to have them give our name & number out). If anyone really does have this package, and feels like beating it up to hell and back (once again, providing I can reload it!) give me a yell email or voice. Perhaps we can inject a little reality into all this marketing hype that our "friend from a not disinterested company" posted. I am specifically NOT interested in talking to anyone from SST, USR, or any other person selling these products about this. I want a few garden-variety users so we can insure the data and methodology is unbiased. We'll post some TRUE transfer rates with Usenet batches or other compressed files, hang up a few times (and see if it picks up where it left off), see how it interfaces with normal Unix(tm) software (usenet, uucp mail, our AKCS conferencing system) etc. I have no affiliation with either USR or Telebit except as a normal paying customer of both firms. We own both Telebit and HST modems. -- Karl Denninger (karl@ddsw1.MCS.COM, ddsw1!karl) Data: [+1 312 566-8912], Voice: [+1 312 566-8910] Macro Computer Solutions, Inc. "Quality solutions at a fair price"
felstein@mcnc.org (Bruce M. Felstein) (02/24/89)
I think that you are misinformed as to the way the HST operates. There is now an HST that is FULL V.32 and HST. This will be available to the public within the next several months. It is, in fact, faster than the Trailblazer. How much faster, I'm not at liberty to say. Also, the cost is much less than the Trailblazer. The HST uses class 5 MNP. If you have a V.32 modem and want to see if the Robotics is V.32 compatible, call (919) 682-4225 and see for yourself.
pete@octopus.UUCP (Pete Holzmann) (02/24/89)
In article <4077@alvin.mcnc.org> felstein@mcnc.org.UUCP (Bruce M. Felstein) writes: >There is >now an HST that is FULL V.32 and HST. This will be available to the public ^^^ >within the next several months. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Let's all learn to compare apples with apples, especially when dealing with marketing hype. 'now' == 'announced'. Availability is 'sometime in the not- too-distant future'. NEVER compare announced future products with currently shipping (or worse - old versions of) competing technologies. It just doesn't tell you anything useful. >It is, in fact, faster than the Trailblazer. >How much faster, I'm not at liberty to say. Also, the cost is much less than >the Trailblazer. 'the Trailblazer'. Which one? The original? The T1000 that you can buy for under $600? (795 list, or something like that)... how about the T2000 we've been hearing about? Or the 'new' one that has V.32, 28Kbps (uncompressed) PEP throughput, plus a kitchen sink full of other stuff that USR hasn't announced. This 'new' telebit will also be available to the public 'within the next several months'. I'm sorry, but in the Unix marketplace, to compare apples with apples, in a useful way, you've got to compare: - uucp based - compressed news batch transfers - on a variety of machines - optional in some cases: (1) compatibility with lots of normal 1200/2400 modems, with or without MNP... and (2) ability to do well over poor quality links. Under these conditions, the current telebit solutions (TB+ and T1000) beat the HST/ix all over the place. The returns can't possibly be in yet on announced but unshipped products from both companies. Clearly, USR will give TB a run for their money in the next round, but *both* companies are upping the ante a lot. It's fun to watch! Now, which one will be first to add FAX to their new modem 1/2 :-)? Pete -- Peter Holzmann, Octopus Enterprises |(if you're a techie Christian & are 19611 La Mar Ct., Cupertino, CA 95014 |interested in helping w/ the Great UUCP: {hpda,pyramid}!octopus!pete |Commission, email dsa-contact@octopus) DSA office ans mach=408/996-7746,Work (SLP) voice=408/985-7400,FAX=408/985-0859
jhood@biar.UUCP (John Hood) (02/24/89)
In article <3004@ddsw1.MCS.COM> karl@ddsw1.UUCP (Karl Denninger) writes: >In article <415@studsys.mu.edu> kowals@studsys.mu.edu (kowalski) writes: >But it is not V.32 if I remember correctly (or is the modem truly V.32 >capable?). This means you're not compatible with V.32, or the Telebit >PEP installed base. For Unix use, the fact that every Bow-wow BBS (Fido) >has one is 100% irrelavent when making a purchase decision. It's a half-duplex variant of V.32 with a slow back channel added. In other words, it's not V.32 at all. They will soon introduce a model with V.32 capability. >... >Not from what I can see. Note also that the software, from what I see in >the configuration info, will talk nicely to a Telebit as well! (note: Telebits >run 14kbaud max uncompressed data rate, HSTs run 9.6kbaud uncompressed.... >guess which modem should do better with the HST/ix software?) Currently, the Telebit, but US Robotics is just about to introduce a 14400 bps variant of their modem (vaporware, but not much longer.) >I can buy two Telebits as an Internet site for the cost of ONE of your >Courier HST/ix's. Earlier as a Usenet site I could do the same thing. The >cost of the HST/ix solution is effectively twice the price of the Telebit way. Not if you run a BBS, or (I think also) a public-access Unix machine. Both of these qualify for USR's half price sysop deal, which cuts the price of the modem to $495, and you don't get the apparently useless HST/ix software (grin). I do believe, Karl, that you fall into this category. >o The HST modems give some 2400 baud non-MNP capable units, and all AT&T > 3b1's, fits. They simply won't connect with them AT ALL if you have > high-speed enabled (that is, not &M0). This is only a problem when dialing out; it should be easy enough to stick the appropriate &M0 command into the L.sys entry for that particular machine. The Telebit has similar problems with its three little chirps and modems that attempt to recognize voice answers (including the HST ;) >A note: > >o We only tested it locally (in-building) between two Xenix systems on two > local phone lines. I have no idea how well (or if) it copes with > long-distance telephone lines and/or overseas connections. USR didn't > feel like giving out names of other sites across the country for us to > poll..... nor did they tell anyone else to try polling us (although I did > tell them we'd be more than happy to have them give our name & number out). Telebit apparently has the edge for lousy phone lines, especially overseas. For a UUCP link, dig up a copy of the pubnix public access Unix list and find oncoast, in PA. He is running an HST there. I can tell you this: you'll get about 160 cps through an HST with UUCP; the two are badly mismatched. Alternately, I could call you from my Unix machine at home, but I make no claims to having the HST/ix software either. >I have no affiliation with either USR or Telebit except as a normal paying >customer of both firms. We own both Telebit and HST modems. > >-- >Karl Denninger (karl@ddsw1.MCS.COM, ddsw1!karl) >Data: [+1 312 566-8912], Voice: [+1 312 566-8910] >Macro Computer Solutions, Inc. "Quality solutions at a fair price" I have a USR modem at home, running a BBS, and we have a Telebit on order for our UUCP feeds, both under the discount programs. I happen to think that the USR and the Telebit are both good modems, and very usable for their respective markets. The USR technically makes a terrible Unix modem (except for interactive dialin, where it would be pretty good), and the Telebit technically makes a so-so BBS modem. They're both pretty useless outside of their respective markets because no one else has one. The awfulness of the HST as a Unix modem, by the way, has more to do with the inefficiency of the 'g' protocol than anything else. Telebit implemented it in the modem, USR didn't. I think Telebit's action effectively cut off any potential development of better UUCP protocols, which would really have been the better way to go. I await the nearly-here day when both companies have a V.32 variant of their modems; then internetworking Fidonet and Usenet will be a more practical exercise than it is now. I think the HST/ix seems like an ill-conceived effort to sell modems to the wrong audience; fortunately, USR seems not to be pushing it too hard. As I recall, it wasn't even in their price list they sent me a while back. Mr. Kowalski does seem too exuberant about the modem's use for Unix. --jh -- John Hood, Biar Games snail: 10 Spruce Lane, Ithaca NY 14850 BBS: 607 257 3423 domain: jhood@biar.UUCP bang: ...!{princeton!labii,hombre!lgnp1}!biar!jhood "My mommy told me not to put beans in my ears"-- some forgotten record
len@netsys.COM (Len Rose) (02/25/89)
In article <149@biar.UUCP> jhood@biar.UUCP (John Hood) writes:
#
# Not if you run a BBS, or (I think also) a public-access Unix machine.
# Both of these qualify for USR's half price sysop deal, which cuts the
# price of the modem to $495, and you don't get the apparently useless
# HST/ix software (grin). I do believe, Karl, that you fall into this
# category.
#
That offer is just about useless on this net.. It's sure to make a
low level penetration in the Unix world if this is the best they can
do.. There are merely a handful of public access sites.. If they want
to compete with Telebit for the unix market they will have to match
their marketing efforts.
--
len@netsys.com
{ames,att,rutgers}!netsys!lenjoel@arizona.edu (Joel M. Snyder) (02/25/89)
I don't want to comment on the merits of HST vs. Telebit. However, the gentleman who is "not at liberty to say" when the HST/V.32 USR will be available is out to lunch. This modem has been available since the beginning of the year. IF you have a use for such a thing (and there are those of us out here who do), you can get it with immediate delivery. Pricing is not what I'd call competitive, but is it less than a V.32 and an HST combined. jms
felstein@mcnc.org (Bruce M. Felstein) (02/26/89)
The US Robotics with true V.32 is not yet available for shipment. It is still being beta tested to work out the kinks. I have one running on my BBs I'm running the DUAL STANDARD, which is both regular HST and a V.32 modem in one box. When they are available, the price will be 725 for sysops or I would imagine that UUCP would also qualify for sysop prices. The UUCP price for the new Telebit 2500 will be around 1200 which is about 500 more than the US Robotics.
dave@onfcanim.UUCP (Dave Martindale) (02/26/89)
In article <4077@alvin.mcnc.org> felstein@mcnc.org.UUCP (Bruce M. Felstein) writes: >I think that you are misinformed as to the way the HST operates. There is >now an HST that is FULL V.32 and HST. This will be available to the public >within the next several months. It is, in fact, faster than the Trailblazer. If it really implements V.32, it will be 9600 bps full duplex. Since uucp sends data in only one direction at a time, you will get 960 * 64/70 cps absolute maximum data throughput. (64/70 allows for packet header overhead). The Trailblazer runs 1440 cps max in one direction, and doesn't have to send the uucp packet headers. So theoretical througput is considerably higher then the HST. And I, and many others, know that the real througput is well above 960 cps. > The HST uses class 5 MNP. So what? The Trailblazer handles error control just fine. And any compression provided by MNP is useless when sending already-compressed news batches. (The Trailblazer also has built-in compression, likewise useless for news). Please explain how any V.32 compliant modem can possibly perform as well as the Trailblazer sending compressed news batches. Even if the V.32 provided uucp protocol spoofing, it would be slower.