David@cup.portal.com (08/02/88)
I certainly got a lot of reaction to my article bemoaning the Trailblazer for its lack of adherence to standards. Here's a summary of the reaction and my responding comments. Please excuse the length, but I think this is a useful dialog. In article <3408@phri.UUCP> roy@phri.UUCP (Roy Smith) writes: > The original request was for info on unsupported modems to be used >for a specific trans-Atlantic point-to-point dial-up link. Given the cost >of trans-Atlantic phone calls, the savings you realize by even slightly >better modems make it economical to buy the best you can now and toss them >in a year or two if the winds of change make them no longer the best choice. I agree with this point of view; if you save enough money, an approach that is non-standard can be justified despite the disadvantages. Business is business, after all. Thanks for the comment, Roy. Yet, he seems to agree with my contention that non-standard equipment is disposable, and is headed for the trash can in a few years. In article <5052@bigtex.uucp> james@bigtex.UUCP (James Van Artsdalen) writes: >Ventel makes a PEP modem, and I understand there are others with PEP >products already on the market. I was not aware there were any vendors second-sourcing the Trailblazer. Thanks for the info. So, there is at least one other vendor besides Telebit. >Standards are interesting things. I have never had anyone ask for a >link with bigtex that couldn't talk to my TB+ at their fastest >supported data rate (one site with a Hayes 2400 has some difficulty >with PEP-tones-first). I have never asked anyone for a link and found >that I couldn't talk to them at their highest supported data rate. "That's what I like about standards - there's so many to choose from!". Seriously, your perspective and mine are different; your situation apparently is a lot smaller in scale (I am assuming you are a netadmin type) than mine, and the environment (another assumption, news & mail access) is already pretty well permeated with Trailblazers; mine is not. I'll explain my environment, and my perspective at the end of this article. >There may be V.32, Hayes V9600 and US Robotics HST modems out there, >but I haven't seen any sign of them around here. Well, I would like to point out in the strongest possible language that I do not consider USR HST or Hayes V-series modems to be any more "standard" than the Trailblazer! When I mention V.32 modems, I mean ones compliant with CCITT Recommendation V.32, not subsets, or proprietary derivations thereof. As for not seeing V.32 modems, I assure you they are out there. I have four about two feet away from me, in fact. (Codex 2260's and GDC DC296's, for those who want to know.) In Email <8807302248.AA21314@csvax.caltech.edu> ktl@wagvax.caltech.edu (Kian- Tat Lim) writes: > One of the key features of the Trailblazer is that it contains a >high-speed digital signal processing chip. This chip allows it to encode >and decode the special communications protocol it uses. However, the protocol >is not a function of the chip, which is quite general-purpose, but of the >software that drives it. As a result, Telebit, by "merely" changing the >EPROMs in the unit, can have the Trailblazer use any given protocol. In fact, >Telebit has stated that if a true international standard for high-baud-rate >communications is adopted, they will add it to the Trailblazer ROMs. Thus, >a Trailblazer is a pretty safe choice and won't have to be tossed in the trash >after a couple of years. Thanks for the info. Sounds like Telebit has some product engineers who know their stuff. If and when they are V.32 compliant, then my arguments disappear. But I won't hold my breath. In email <8807310205.AA22406@csvax.caltech.edu> mangler@csvax.caltech.edu (Don Speck) writes: >Ventel also makes Trailblazer compatibles under license. > >When our lab manager said she wanted to buy another 2400 bps dialin >for cit-vax, the brand I suggested was Telebit. That's right, our >Trailblazer is used almost entirely at 2400 bps. It cost 60% more >than the Vadic VA4224 that she was going to get, but it doesn't get >stuck like our 4224's do. Well, I can use my V.32 Codex 2260's in a similar way. But it wouldn't be their primary application. (My silly management expects me to implement cost- effective solutions.) I do find the V.32/V.22bis combination useful for combining dial-in interactive access with dial backup for dedicated lines, however. In Email <8807310548.AA14883@octopus.UUCP> sun!octopus!pete (Pete Holzmann) writes: >How many V.32 modems talk to each other? The last tests I saw, NONE. This >could have changed by now, but several months ago, the various PC magazines >tested them. They liked the speed, but didn't find any more inter- >compatibility than Telebit has. I haven't seen the articles you are referring to, so I have to guess a bit as to what modems were reviewed. If you are referring to modems such as the USR HST and Hayes V-Series, you're right, they don't interoperate. But, to repeat myself, these modems aren't any more standard than the Trailblazer. And if you are referring to modems the bill themselves as fully V.32 compliant, remember how some V.22bis modems wouldn't talk to each other until the modem engineers got more experience? Probably a similar situation. >I hate tossing stuff too. Which is *exactly* the reason I went with the >Telebit. It is so smart, they can add nifty things like V.32 simply by >changing the PROMs! I don't know if you saw the survey they sent out on >the net. But they were looking to find out which new features people >want. Right now, V.32 is not at the top of the list. I think FAX support >was higher, for example (they could add that as well). > >Telebit will add V.32 when the market demands it. For now, the PEP protocol >is a lot nicer than V.32, so they aren't giving it up. Nor are they wasting >time on something slow like V.32. > >Personally, I'd much rather see them add the Fax support, the new MNP level >7+ ultra-compression support (if a 2400 baud modem can get 9600 on binary, >think of what a 14K modem can get! 38.4 support, here we come!!!) To repeat myself again, if and when the Trailblazer supports V.32, my arguments disappear. Thanks for the comments, Pete. In Email <8807310823.AA00975@stanton.TCC.COM> uunet!stanton.TCC.COM!donegan writes: >I have several rules in implementing data communications products for my >company: > >1) You NEVER throw away something that works, weather it is standard or not. > >2) You implement whatever it takes to get the job done, even if it is NOT > standard. > >Given those two rules of thumb, the Trailblazer works for me in environments >that choke V.32 modems (or any others for that matter). One really ugly >environment to do dial-ups to is from the US to Singapore, nasty line hits >and echo/noise constantly. Only the Trailblazer worked in this application. >If your application is a leased line one, and you are using an error >correction (such as X.25 or whatever) then use whatever standard modem you >choose and you'll be as happy as the unreliable nature of the 'last mile' in >a foreign country will allow you to be. For nasty environment dial-ups use a >Trailblazer, or if you find something better (real world testing, not bench >testing) please let me know. I don't think we disagree on the "rules" here. In fact, I too hate throwing things away, which was the whole point of my original posting. And if I have to use non-standard equipment to get the job done, so be it. But I don't have to think it is a good long-term solution. Having had personal experience with communications to foreign countries, I can agree that sometimes it seems that an ultra-high-noise-immunity (but proprietary) modem like the Trailblazer is the only solution. If I absolutely had no other choice, I would use it. But the reality is that there is almost always another choice. Now, my general comments as to my environment and perspective, to amplify my previous posting: The environment I must implement communications systems in is a worldwide corporation with a heavy emphasis on IBM/SNA and X.25 architectures. I am responsible for the planning and implementation of new networks and the upgrade and fine-tuning of established networks. So, I have a very large, global view as opposed to some of the readers, who seem to range upward from the very small and very domestic U.S.A. outlook. If I have a choice, my choice will *always* be to utilize a standards-oriented approach. The benefits are: - I don't have the disposable-equipment syndrome, or support at a given vendor's whim, because there are many vendors competing for my business. - I have equipment I can keep for five years and re-use in applications that weren't even thought of when the equipment was originally purchased. - Standards-oriented equipment is much cheaper to operate over a five year period; non-standard equipment will have to be replaced before then. - Foreign PTT's and domestic carriers are much easier to deal with when troubles arise if I use standards-compliant equipment. - Accessory systems such as network management can be added with far greater ease to a network based on standards rather than one based on one vendor's proprietary schemes. And, as my last thought of the day on this topic, consider *why* vendors such as Telebit are offering price incentives to USENET sites. It is not out of the goodness of their hearts. They are doing it in an attempt to seduce this particular influential market sector away from a standards-based approach, which is of course to their benefit over the long term, and to the general consumer's detriment. An analogy is they want you netadmin types to quit using rectangular bricks and start using spherical ones, so they can dominate the spherical brick market and make lots of money. It's not a conspiracy, but don't believe for a moment that Telebit puts being nice to USENET sites ahead of making money. Just take a more consumeristic attitude - is the short-term benefit worth the long-term problems? I don't think so. I say, "The Emperor has no clothes!", when it comes to Telebit, that's all. David@cup.portal.com David McCord 415/424-5644 voice
roy@phri.UUCP (Roy Smith) (08/03/88)
In article <3408@phri.UUCP> I said: > buy the best you can now and toss them in a year or two if the winds > of change make them no longer the best choice. In article <7805@cup.portal.com> David@cup.portal.com replies: > he seems to agree with my contention that non-standard equipment is > disposable, and is headed for the trash can in a few years. David misses my point, which is that *ALL* computer equipment is disposable and is headed for the trash can in a few years. It wasn't too many years ago when I was agonizing over whether I should go with Bell 212A (i.e. standard) modems or some other manufacturer's (was it Ventel?) almost standard 1200 baud modems which were supposed to be more resistant to line noise. Guess what? 5 years later, it makes precious little difference which decision I made then because we almost never use 1200 baud any more and I fully expect that within a year or two we won't support it any more. Yes, I'm convinced that Trailblazers are the best 9600 baud dial-up modems made today and that people would have to be nuts to buy anything else (in fact, I refused the gift of a MicroComm 9624/X not long ago). On the other hand, I'm also sure that 5 years from now my Trailblazers will be sitting in that pile of "broken but not worth repairing" equipment. Why am I going to need 9600 bps modems when I've got 64 kbps digital service on my home phone line? -- Roy Smith, System Administrator Public Health Research Institute {allegra,philabs,cmcl2,rutgers}!phri!roy -or- phri!roy@uunet.uu.net "The connector is the network"
john@synsys.UUCP (John C. Rossmann) (08/04/88)
In article <7805@cup.portal.com> David@cup.portal.com writes: > >And, as my last thought of the day on this topic, consider *why* vendors such >as Telebit are offering price incentives to USENET sites. It is not out of >the goodness of their hearts. They are doing it in an attempt to seduce this >particular influential market sector away from a standards-based approach, >which is of course to their benefit over the long term, and to the general >consumer's detriment. >David McCord David's comment here is interesting, but I think he has the implications a bit skewed. Although I'm possibly wrong here, what little I've read seems to say that V.32 isn't really that much of a standard yet -- or at least that V.32 doesn't have significant market penetration yet. I don't think that Telebit is trying to seduce anyone AWAY from an established solid standard. I think that, given the current state of V.32, Telebit is attempting to CREATE a new and better standard. If Telebit can get thousands of modems out there in the Usenet world, and if Telebit can establish a reputation for greater reliability and cost effectiveness, they will have established a new de-facto STANDARD. Then it may very well be the CCITT's turn to adapt, not Telebit's! John Rossmann (uucp: uunet!synsys!john) (CIS: 70701,3125) Burke, VA
james@bigtex.uucp (James Van Artsdalen) (08/04/88)
In article <7805@cup.portal.com>, David@cup.portal.com wrote: > [...] They are doing it in > an attempt to seduce this particular influential market sector away > from a standards-based approach, which is of course to their benefit > over the long term, and to the general consumer's detriment. Standards are NOT set by people sitting on international settings with huge amounts of money and resources from their companies backing their efforts. Standards ARE set by large numbers of people buying compatible hardware. If everyone buys a TB+ and no one buys V.32, then V.32 is not a "standard" in any useful definition of the word, and the TB+ is clearly a standard. The only thing all those nice people with their fancy committees do is generate lots of publicity and some monentum - they have little say in what really happens except as in they buy and use the equipment they advocate, and that they generate publicity. The good thing about committees is that a standard emerges before it otherwise would. The bad thing about them is that they generate standards that are useful for vendors, not ones that are useful for users. > [...] > but don't believe for a moment that Telebit puts being nice to USENET > sites ahead of making money. I don't. Telebit is playing the standards game, just the V.32 committee did, but Telebit is going to play strictly in the marketing area and skip the fancy committees. Thus far, Telebit is half the price and higher performance. Has anyone seen any market statistics as to how many PEP modems there are in end-user hands as opposed to V.32, USR HST or Hayes V series? > Just take a more consumeristic attitude - is the short-term benefit > worth the long-term problems? I don't think so. I say, "The Emperor > has no clothes!", when it comes to Telebit, that's all. Feel free to point out any long term problems. Without a sudden drop in V.32 prices, that in a mysterious way does not affect TB+ prices, I can't see how the situation will change. And even if it does, I don't expect ANY of my equipment to have more than a 5 year lifetime, nor do I expect any of it to be current for more than 3 years at best. Certainly none of your V.32 modems will be current technology in five years (ISDN perhaps?)... PS. Yes, I do believe that V.32 will be better than the existing TB+ in the long term. However, that's another day, and V.32 is not the answer for usenet sites today. I haven't been bean counting, but I suspect that my TB+ has ALREADY paid for itself, rendering the issue of throwing it away in the future moot - the TB+ did the job when V.32 couldn't. Two or three years from now V.32 will probably be the best thing, but that doesn't help next month's phone bill. -- James R. Van Artsdalen ...!ut-sally!utastro!bigtex!james "Live Free or Die" Home: 512-346-2444 Work: 328-0282; 110 Wild Basin Rd. Ste #230, Austin TX 78746
dewey@execu.UUCP (Dewey Henize) (08/04/88)
In his comments about Telebit, James made some good points about how they (Telebit) are basically bypassing the 'normal' committee route for setting standards and are instead letting the marketplace 'vote' with it's bucks. I couldn't agree more that that is what is going on in this case. I would like to make a friendly note here, though, that this is not the way it will always go. I believe that we have a situation here were technology and pricing simply got way ahead of the committee types, so we get a chance to vote-by-buck. (I happen to like this particular case). I don't think you will be able to generalize this, though, since there are an awfully lot of situations where the standards committees don't have the market pressure to make an end-run around them, and thus the committee types get a chance to make up stuff, write many and varied learned papers about each others learned papers, and basically do a million things to make it appear the 'standard' is an accomplished fact - when it sometimes no longer even applies to the current world. Look at all the huhu about standard COBOL (ugh) and FORTRAN. These folks are still fighting it out, over and over. There is hardly any COBOL or FORTRAN out there though is simply meets standards - because that isn't enough. Telebit might have it right, in the ad I saw recently. Something like 'the only Unix (tm) modem'.... Dewey P. S. - we have ordered a couple, haven't got em yet. I am NOT an employee of Telebit, or anyone other than Execucom..... Just adding an observation. -- =============================================================================== | execu!dewey Dewey Henize @ Execucom Systems Corp 512/346-3008 | | You don't think my employer APPROVES of these ideas, do you?? Sheesh! | ===============================================================================
smb@ulysses.homer.nj.att.com (Steven Bellovin) (08/04/88)
Of course, there are countries in the world where it's illegal to connect non-standard modems to the phone line. Those are markets that are closed to Trailblazer at this time.
rls@telebit.UUCP (Richard Siegel) (08/04/88)
With regard to some of the articles that have been circulating in this newsgroup
about standards, here is Telebit's position:
1. The TrailBlazer Plus currently has an estimated 45-60% of the installed base
of HIGH-SPEED (greater than 9600 bps) modems. With the inclusion of such
features as our protocol support for UUCP, we have listened to the Unix
community, and delivered performance in a way that saves users money
with every use.
2. V.32, the CCITT recommendation is indeed a standard, in the sense that if
you purchase a TRUE V.32 modem from one manufacturer (such as UDS), and
try and talk to one of anothe manufacturer (such as Concord), they will
actually converse. Just like V.22 bis modems (2400 bps). This recommendation
does not include any specification for error control, so then you are in
the proprietary technology again (until V.42 finally arrives).
3. Telebit is both aware of and participating in the standards committees in
Europe. We have been working with the V.34 working group for nearly two
years, and are one of the two final proposals for an ultra-high-speed
asymmetrical modem. Our proposal is for a 28 Kbps version of PEP, and we
have demonstrated a working prototype for evaluation to the CCITT. But
this "standard" is not expected to be finalized until next spring, at the
earliest.
4. Currently, the only protocols for communicating at above 9600 bps that are
supported by more than one manufacturer are V.32 (there are several makers
of V.32), and the Telebit PEP technology. In addition to Telebit, several
other manufacturers have licensed our technology. This list includes Ven-Tel,
Racal-Milgo, GTE, Digital Pathways, and others. No other modem manufacturer
can make that claim.
5. As stated previously, Telebit is very interested in the Unix marketplace,
for obviously financial reasons. But our modem would not have become the
standard so quickly if the product was junk. This is our most technically
sophisticated and vocal market, and if we were not selling the best product,
everyone on the net would hear about it! We are committed to this and other
major markets, and are constantly devising new ways of improving our product
for higher performance, at the same or lower price.
Thanks for your continued support and encouragement, and keep up the
discussions; this is one of the best forums anywhere for intelligent modem
discussion, and Telebit is committed to participation in it, and learning from
the many different points of view.
Keep up the discussions!
Regards,
================================================================================
Richard Siegel Phone: (415) 969-3800
Product Manager UUCP: {sun,uunet,ames,hoptoad}!telebit!rls
Telebit Corporation ARPA: telebit!rls@ames.ARPA
"We are, after all, professionals"...HST
================================================================================David@cup.portal.com (08/06/88)
In article <3419@phri.UUCP> roy@phri.UUCP (Roy Smith) writes: > David misses my point, which is that *ALL* computer equipment is >disposable and is headed for the trash can in a few years. It wasn't too >many years ago when I was agonizing over whether I should go with Bell 212A >(i.e. standard) modems or some other manufacturer's (was it Ventel?) almost >standard 1200 baud modems which were supposed to be more resistant to line >noise. Guess what? 5 years later, it makes precious little difference >which decision I made then because we almost never use 1200 baud any more >and I fully expect that within a year or two we won't support it any more. Well, nothing lasts forever, I can agree with you there. In fact, most network planners other than myself use a three year period as the time period a given piece of equipment will be in service, and do their financial analyses accordingly. But I prefer a five year plan, as it is more appropriate for my current environment. By the way, my company has scads of 212A modems still in use, many only used in 103 mode. Many 208s also. So, you can justify the non-standard approach. Fine. The trash-can syndrome doesn't bother you. But, your environment is different than mine; I *know* that the gear I install today is going to be staring me in the face four years from now, either still serving the same application or with me wanting to re-use it in applications I can't possibly predict now. I believe that any network planner in a Fortune 100 corporation is in the same situation. David@cup.portal.com David McCord 415/424-5644
David@cup.portal.com (08/06/88)
In article <309@synsys.UUCP> john@synsys.UUCP (John C. Rossmann) writes: > Although I'm possibly wrong here, what little I've read seems to say >that V.32 isn't really that much of a standard yet -- or at least that V.32 >doesn't have significant market penetration yet. Well, this argument greatly resembles the great 802.3 vs. Ethernet debate back when the IEEE decided to not simply endorse standard Ethernet. They made some changes to keep all possible vendors at a disadvantage (approving Ethernet with no changes would have given certain vendors a huge advantage) and came out with 802.3. Standards are the consumer's only protection from markets which have little or no competition. Unless there's a dozen or so vendors in a given market, it doesn't matter how big the market is; you don't have any protection if that vendor goes bankrupt, is bought out by competitors and shut down, or if any number of other possible events occurs. The V.32 standard exists. Arguing that it can be ignored because it doesn't have x percent of the market is arguing against all of the communications standards ever formulated, because at one time or another they all lacked x percent of the market. David@cup.portal.com David McCord
james@bigtex.uucp (James Van Artsdalen) (08/06/88)
In article <291@telebit.UUCP>, rls@telebit.UUCP (Richard Siegel) wrote: > 2. V.32, the CCITT recommendation is indeed a standard, [...] > This recommendation does not include any specification for error control, > so then you are in the proprietary technology again (until V.42 finally > arrives). Does this imply that if I call someone's brand X V.32 with my brand Y V.32, that there isn't necessarily any error detection or correction going on between the modems? This doesn't make V.32 sound terribly useful to me. Does the PEP "standard" (ie, what is licensed to other manufacturers) specify the modem-to-modem error correction protocol? -- James R. Van Artsdalen ...!ut-sally!utastro!bigtex!james "Live Free or Die" Home: 512-346-2444 Work: 328-0282; 110 Wild Basin Rd. Ste #230, Austin TX 78746
chris@mimsy.UUCP (Chris Torek) (08/07/88)
>In article <309@synsys.UUCP> john@synsys.UUCP (John C. Rossmann) writes: >> Although I'm possibly wrong here, what little I've read seems to say >>that V.32 isn't really that much of a standard yet -- or at least that V.32 >>doesn't have significant market penetration yet. In article <7897@cup.portal.com> David@cup.portal.com writes: >Well, this argument greatly resembles the great 802.3 vs. Ethernet debate >back when the IEEE decided to not simply endorse standard Ethernet. They >made some changes ... and came out with 802.3. Indeed. And now many Ethernet manufacturers provide two (or more) modes: `useful' mode and `useless' mode. `Useful' is Ethernet 2; you can guess which the other is. (Obviously I am biased since we use Ethernet 2 :-) .) (A third mode is Ethernet 1, which is like 2 but is DC coupled.) >The V.32 standard exists. Arguing that it can be ignored because it >doesn't have x percent of the market is arguing against all of the >communications standards ever formulated, because at one time or another >they all lacked x percent of the market. Many of them still do. Numerous standards were never much used at all. It is impossible to predict whether new standards such as V.32 will become popular, although manufacturers can get rich if they guess right. (Users get poor either way :-) .) -- In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7163) Domain: chris@mimsy.umd.edu Path: uunet!mimsy!chris
wayne@teemc.UUCP (//ichael R. //ayne) (08/07/88)
I just had a realization that I thought I would share with the net. Since Telebit started with the half-price deal, I have not seen a single posting of someone selling their TB+. This tells me that almost every customer must be satisified with their purchase. I know I am (actually, I wish that usenet sites could purchase 3-4 at half price...) /\/\ \/\/ -- Michael R. Wayne --- TMC & Associates --- wayne@teemc.uucp INTERNET: wayne%teemc.uucp@umix.cc.umich.edu uunet!umix!teemc!wayne
dave@onfcanim.UUCP (Dave Martindale) (08/07/88)
Telebit had a booth at SIGGRAPH, pushing Trailblazers for use in transmitting images over phone lines. I stopped and talked to one of the people for a while (don't remember his name though). He mentioned that Telebit had demonstrated a 28Kbps version of PEP, and I asked if the existing modems would be upgradable to do this. The answer was no - the current DSP chip doesn't have enough power, and they switched to a faster one to get the needed crunching power. I wonder if the current modems don't have the power to do V.32, but the upgraded hardware would?
rusty@cadnetix.COM (Rusty) (08/09/88)
In article <309@synsys.UUCP> john@synsys.UUCP (John C. Rossmann) writes: ... >that V.32 isn't really that much of a standard yet -- or at least that V.32 >doesn't have significant market penetration yet. I don't think that Telebit >is trying to seduce anyone AWAY from an established solid standard.... >... > they will have established a new de-facto STANDARD. I bought a non-Epson-compatible printer many years ago, just before Epson(tm) became the standard. Guess what? I plan to wait for a standard to emerge before I even CONSIDER buying a faster than 2400baud modem! But, then I dont run megabytes of info per day. Not even kbytes of info per day. And rarely, if ever, long distance. So I can afford to wait. And, on another (slightly-related) subject: Aren't you glad, those of you in the USA, that the government does not own the communication network? And that, so far, you can use whatever modem speed you want? (But watch out, this surcharge they plan to add (Have added?) for modem communications is a step in a BAD direction!) ----- Rusty Carruth \ Cadnetix \ 5775 Flatiron Pkwy. \ Boulder CO 80301 \ (303) 444-8075 'HOME': P.O. Box 461 \ Lafayette, CO 80026 RADIO: N7IKQ UUCP:cadnetix!rusty DOMAIN:rusty@cadnetix.com PATH?:{uunet,boulder}!cadnetix!rusty