jwb@uunet.uu.net (Jim Breen) (12/15/89)
In article <2108@accuvax.nwu.edu>, euatdt@euas17c10.ericsson.se (Torsten Dahlkvist) writes: > Hello again! > In article <2023@accuvax.nwu.edu> jwb@cit5.cit.oz (Jim Breen) writes: > >What you need to solve your problems are some ISDN Terminal Adaptors > >(TA) of various flavors. The problem is they haven't been developed > >yet! ............. > Funny you should ask... > I spent five years (83 - 88) as part of Ericsson's ISDN terminal > project. We did produce a feature-phone and a range of TA:s which > conform very closely to the "official" ISDN spec. The deviations were [etc.] > Ericsson's terminals are available NOW for Ericsson customers. The > TA:s handle V24, X21 and X25 to mention the more popular protocols. > The only problems are availability and the prices... > You see, we started that project way back before any VLSI:s had > appeared on the market (actually, we cooperated closely with AMD in > their work with their chipset) and the custom-circuits used in that > first generation of terminals are *expensive* and hard to get. This > puts the prices of the terminals at a level few customers can handle > and in reality all sales so far have been to Telcos using Ericsson > equipment who want to set up field-trials for ISDN. I feel somewhat humbled talking about ISDN with someone like Torsten, who is clearly well on top of both the technology and the pit-falls. It is also fascinating for me, an Australian, to be discussing topics with a Swede via a newsgroup moderated in the US. What is also fascinating is the apparent low level of understanding of ISDN in the US. Why is this? Is it the state of fragmentation in the US telecommunications industry? Does a country need monopoly suppliers like Televerket or Telecom Australia in order to have a working ISDN? Getting back to Torsten's reply to my reply to Dr Weber's questions, clearly he is right; there *ARE* TAs around. The trouble is you can scour the trade press and not find a single advertisement for them. You can ask Ericsson's Australian subsidiary for details of Torsten's TAs and be told that they can only supply bog-standard adaptors for their MD110 PABXs (and this from the company which supplies the AXE exchanges (COs) on which our ISDN is built!) As a Volvo owner (a Japanese speaking one at that) I am saddened that Ericssons aren't taking on the world with their TAs. You can only take corporate conservatism so far. _______ Jim Breen (jwb@cit5.cit.oz) Department of Robotics & /o\----\\ \O Digital Technology. Chisholm Inst. of Technology RDT\ /|\ \/| -:O____/ PO Box 197 Caulfield East 3145 O------O _/_\ /\ /\ (p) 03-573 2552 (fax) 572 1298
Torsten Dahlkvist <euatdt@euas17c10.ericsson.se> (12/21/89)
In article <2267@accuvax.nwu.edu> munnari!cit5.cit.oz.au!jwb@uunet.uu.net (Jim Breen) writes: >I feel somewhat humbled talking about ISDN with someone like Torsten, >who is clearly well on top of both the technology and the pit-falls. My irony-detector started trembling on that paragraph. I'm not used to such praise. Must be my old inferiority-complex acting up again :-) >It is also fascinating for me, an Australian, to be discussing topics >with a Swede via a newsgroup moderated in the US. What is also >fascinating is the apparent low level of understanding of ISDN in the >US. Uh oh, asbestos suit on quick... >Why is this? Is it the state of fragmentation in the US telecommunications >industry? Does a country need monopoly suppliers like Televerket or >Telecom Australia in order to have a working ISDN? I can't wait to see the flames this is coing to cause! >Getting back to Torsten's reply to my reply to Dr Weber's questions, >clearly he is right; there *ARE* TAs around. The trouble is you can >scour the trade press and not find a single advertisement for them. >You can ask Ericsson's Australian subsidiary for details of Torsten's >TAs and be told that they can only supply bog-standard adaptors for >their MD110 PABXs (and this from the company which supplies the AXE >exchanges (COs) on which our ISDN is built!) Let me put it this way: I work at the company which did virtually ALL the R&D on both the AXE COs and the MD110 PABXs. *I* don't have even an MD digital line, much less an ISDN one. >As a Volvo owner (a Japanese speaking one at that) I am saddened that >Ericssons aren't taking on the world with their TAs. You can only take >corporate conservatism so far. Of course Jim is right. There's something saddening about the way big companies are going about this business, Ericsson not excluded. (Don't let anyone at Management see this...) You see, while small compared to the dragons of the telecom world, Ericsson is still in many aspects an uncomfortably big corporation when it comes to introducing changes. The MD110 PABX is manufactured by one company division while the AXE exchanges are made by another one. Unfortunately, phones and TAs are considered "office equipment" and thus fall under the same division as the PABXes. The sales people over there are utterly convinced that "ISDN can never be a success because they use 4-wire installations which must obviously be twice as expensive as ordinary ones". Needless to say, the Ericsson-specific (digital) feature-phones and TAs available for MD110 use 2-wire installations... ("bog-standard" to use Jim's words.) But all is not lost yet. I forwarded the message that "there are no TAs for sale in the US" to some appropriate people who immediately started sprouting little $$-symbols in their eyes in the best possible Scrooge MacDuck-style. We've made a HUGE investment in R&D on these buggers and every sale would be a help in cutting the losses! The problem preventing an immediate introduction is that the entire U.S. ISDN spec is "bog-standard", i.e. U.S.A. has chosen to specify a different method of rate adaption than the rest of the world and there are other "sublte" differences too. Some of them are fairly easily overcome - like the protocol differences I mentioned in my previous posting - while others would need major re-designs (or at least program changes in several micro- processors). I wasn't aware of that last part until I asked around a bit. Sorry. But I must ask a question to the net, U.S. readers in particular: I know for a fact (= I was there when it happened!) that Ericsson has been approached by one of *the* major Japanese electronics manufacturers (no, I'm NOT going to say which one - I'm far out on a limb as it is already!) which wanted to sell a line of ISDN phones with built-in TA functions. Essentially a small feature-phone with a V.24, X.21 or (I think, but memory is vague) X.25 connector. We tried it and found that with minor changes in CO software it could be used with our CO (remember I said our implementation is *very* close to the international standard. We like to think it's *the* closest one on the market at present!). Now, the price tags on these phones were rather persuasive - I think they'd be a success on the U.S. market at once if they were released. My question: Does any net.reader know if they're available in the U.S.? If not: Does anybody know why? My own hypothesis so far is that they're slowed down by those very same "oddities" of the U.S. ISDN spec as have deterred us. Knowing the capacity of these Japs I'd expect US-type TAs shortly unless your own import restrictions prevent them :-) /Torsten P.S. Jim: has Telecom Australia made any introduction of BRI yet? As far as I know, Ericsson has only sold PRI to them; hence the MD110 stuff. P.P.S. MD110 = Ericsson's modular PABX. I'm not sure if Marketing has used the same name in the U.S, but it's currently beeing installed at MIT, so I know it's available there. It uses a proprietary (sp?) 2B+D 2-wire interface for feature-phones which provides "ISDN-like" facilities but (unfortunately) so far in a non-standard way. Torsten Dahlkvist ELLEMTEL Telecommunication Laboratories P.O. Box 1505, S-125 25 ALVSJO, SWEDEN Tel: +46 8 727 3788
"Fred R. Goldstein" <goldstein@carafe.enet.dec.com> (12/22/89)
In article <2353@accuvax.nwu.edu>, euatdt@euas17c10.ericsson.se (Torsten Dahlkvist) writes... >In article <2267@accuvax.nwu.edu> munnari!cit5.cit.oz.au!jwb@uunet.uu.net >(Jim Breen) writes: >>Why is this? Is it the state of fragmentation in the US telecommunications >>industry? Does a country need monopoly suppliers like Televerket or >>Telecom Australia in order to have a working ISDN? >I can't wait to see the flames this is coing to cause! I was gonna say, "deliver me from temptation" but instead I'll take the bait. >But all is not lost yet. I forwarded the message that "there are no >TAs for sale in the US" to some appropriate people who immediately >started sprouting little $$-symbols in their eyes in the best possible >Scrooge MacDuck-style. We've made a HUGE investment in R&D on these >buggers and every sale would be a help in cutting the losses! >The problem preventing an immediate introduction is that the entire >U.S. ISDN spec is "bog-standard", i.e. U.S.A. has chosen to specify a >different method of rate adaption than the rest of the world and there >are other "sublte" differences too. There are lots of differences between US POTS and Euro-aussie POTS. For example we use 1.544 Mbps T1 and mu-255 PCM encoding instead of 2.048 Mbps E1 and A-low encoding. ISDN is a lot less different. There is no US rule for rate adaptation, since that's entirely transparent to the network. We have this system here, "competition", which basically says that the telco's fist stops at the customer's face, at the demarcation jack (reference point U). Rate adaptation occurs at a higher layer, so the telco has no say. The manufacturers here (especially AT&T and IBM) tend to favor the HDLC-based rate adaptation technique found in CCITT V.120, which was developed here. Some European-based vendors prefer the older bit-bashed technique found in V.110. And Northern Telecom has its own non-standard T-link which is fairly well established in the field. As long as the customer matches both ends, it'll work. The middle (network) needn't care. Hayes uses V.120, if I recall. >But I must ask a question to the net, U.S. readers in particular: I >know for a fact (= I was there when it happened!) that Ericsson has >been approached by one of *the* major Japanese electronics >manufacturers (no, I'm NOT going to say which one - I'm far out on a >limb as it is already!) which wanted to sell a line of ISDN phones >with built-in TA functions. Essentially a small feature-phone with a >V.24, X.21 or (I think, but memory is vague) X.25 connector. We tried >it and found that with minor changes in CO software it could be used >with our CO (remember I said our implementation is *very* close to the >international standard. We like to think it's *the* closest one on the >market at present!). "ISDN Phones" are a classic example of technology missing the market. ISDN makes a terrible desktop data solution. Most terminal-host connectivity is within the building (local area). ISDN, no matter how you slice it, costs more for this than a LAN with terminal servers. And the terminal servers tend to offer more flexibility, features, etc. Better, faster, cheaper. Pick three. So why waste ISDN by putting it on the desktop? Yes, I remember the early "integrated voice/data PBX" days, and helped put a rack of PBX data modules next to a VAX. Ugly, costly, and happily abandoned to the dregs of history. ISDN's strength in the data world is when you go beyond the LAN. ISDN to the home, or ISDN between locations. It makes a great modem replacement. But who in their right mind uses modems to dial down the hall? ISDN voice/data phones are about as useful as phones with built-in V.22bis modems: No reason not to put a phone on a modem, but not a mass market item. Fred R. Goldstein goldstein@carafe.enet.dec.com or goldstein@delni.enet.dec.com voice: +1 508 486 7388 Do you think anyone shares my opinion, let a long a big corporation?
jwb@cit5.cit.oz (Jim Breen) (01/03/90)
In article <2353@accuvax.nwu.edu>, euatdt@euas17c10.ericsson.se (Torsten Dahlkvist) asks: > P.S. Jim: has Telecom Australia made any introduction of BRI yet? As > far as I know, Ericsson has only sold PRI to them; hence the MD110 > stuff. The short answer is no, not yet. PRI was under trial for a year or so (it kept being delayed by you-know-who's AXE software), and was formally introduced last August. Telecom has announced "BRI" for mid-1990, as a "supplementary service" wherein a PRI will be demuxed into 14 2B+Ds in a local exchange. Just how much this will differ from true BRI, I don't know, but I have to spend some time soon finding out. I guess the mux will have to sort out the SAPIs, TEPIs, etc. In article <2361@accuvax.nwu.edu>, goldstein@carafe.enet.dec.com (Fred R. Goldstein) writes: > "ISDN Phones" are a classic example of technology missing the market. > ISDN makes a terrible desktop data solution. Most terminal-host > connectivity is within the building (local area). ISDN, no matter how > you slice it, costs more for this than a LAN with terminal servers. [etc] ... and > ISDN's strength in the data world is when you go beyond the LAN. ISDN > to the home, or ISDN between locations. It makes a great modem > replacement. But who in their right mind uses modems to dial down the > hall? [etc] Exactly. The applications I am looking at for ISDN involve using it for WANs, as either a replacement for modems and DDS NTUs, or as a standby to back-up our 2.048 M services. This is why I would like some TAs, either to go on separate BRIs or to go on our P(A)BX. There *is* a demand for ISDN phones, provided they supply this sort of wide area data connectivity. _______ Jim Breen (jwb@cit5.cit.oz) Department of Robotics & /o\----\\ \O Digital Technology. Chisholm Inst. of Technology /RDT\ /|\ \/| -:O____/ PO Box 197 Caulfield East 3145 O-----O _/_\ /\ /\ (p) 03-573 2552 (fax) 572 1298
Peter Desnoyers <desnoyer@apple.com> (01/10/90)
In article <2353@accuvax.nwu.edu> euatdt@euas17c10.ericsson.se (Torsten Dahlkvist) writes: > But I must ask a question to the net, U.S. readers in particular: I > know for a fact (= I was there when it happened!) that Ericsson has > been approached by one of *the* major Japanese electronics > manufacturers (no, I'm NOT going to say which one - I'm far out on a > limb as it is already!) which wanted to sell a line of ISDN phones > with built-in TA functions. Essentially a small feature-phone with a > V.24, X.21 or (I think, but memory is vague) X.25 connector. We tried > it and found that with minor changes in CO software it could be used > with our CO (remember I said our implementation is *very* close to the > international standard. We like to think it's *the* closest one on the > market at present!). Sorry to come into this discussion so late (I had a rush project before Christmas and am just catching up on news). Anyway, this sounds similar to the AT&T 7500 set with the optional rate adaption card. I imagine that it would cost lots-o-bux, but I know they are selling them to real people (i.e. they're not just engineering samples) because I saw one in a hotel in Key West. It'll handle 5ESS BRI (of course) at the S interface and seems relatively reliable. I've never used the rate-adaption card - I think the physical interfaces available are RS232 and V.34. Peter Desnoyers Apple ATG (408) 974-4469