Received: from eff.org by kragar.eff.org with SMTP id AA21273 (5.65c/IDA-1.4.4 for ); Mon, 13 Apr 1992 12:51:54 -0400 Received: by eff.org id AA20460 (5.65c/IDA-1.4.4 for pub-infra-exploder); Mon, 13 Apr 1992 12:34:26 -0400 Date: Mon, 13 Apr 1992 12:34:21 -0400 Message-Id: <199204131634.AA20454@eff.org> From: goldstein@carafe.tay2.dec.com (k1io, FN42jk) Subject: ISDN backbone costs, prices To: pub-infra@eff.org (pub-infra mailing list) Bob Frankston's comments, and Mitch's earlier comments from another bulletin board, both point out the importance of ISDN usage pricing. How much does switched data differ from voice? New England Telephone's new Mass. tariff assigns measured voice usage rates to switched intra-office ISDN switched 64 kbps. Most businesses have to pay measured usage anyway. Residential customers in Mass. are entitled to pay for flat-rate local voice calling, with various radius options in the Boston area. But that's not to be applied to ISDN clear- channel 64 kbps. Some other telephone companies have offered voice/data parity for ISDN. I believe that's the case with Southwestern Bell and Pacific Bell, as well as BT (aka British Telecom). Of course, BT doesn't offer any flat-rate local calling to anyone; their local message rates are higher than anywhere in the US. Within a CO, voice = data is a good assumption. Everything's handled at 64 kbps. So parity is logical. But Mass. DPU has a veneer of justification for requiring this to go at message rate: The flat rates were set based upon the typical voice user's traffic. While modems benefit from that (we don't face the odious "modem tax" here the way some SWBell and Moscow Tel customers may), it's still a distortion of the "intent" of flat rates: If ISDN were offered for flat-rate usage, then a single ISDN access line could make _two_ simultaneous calls for "free", one on each B channel! The fix to that gets confusing. The new Mass. tariff charges a fixed surcharge, atop your voice line rate, for ISDN. It doesn't matter if you're Metropolitan or Measured service. If you could make two calls at a time over one Metro-rated ISDN line, they'd be "losing" the revenue that they'd get today by selling you a second Metro line, which costs quite a bit more than a Measured line. The rational-user's response is also interesting. I think that it's possible to make two "speech" calls at a time over an ISDN line. A "speech" call falls under the flat rate. I assume NET doesn't figure that this will be a big problem. To be sure, the "problem" is not one of "losing money" (rate lower than cost), but of "opportunity cost" (lower profit margin). The 1.6c/minute measured rate (over a nickle beyond 8 miles) is way, way above their true cost. It subsidizes the low basic residential rates and cheap installation rates. Crocodile tears come to mind. Now, the rational-user's response is applied to data. If the call is the same voice or data, why even tell the network that it's data? If you originate a local call bearer service = speech, it'll still carry data at 56 kbps, if not the full 64 kbps (American T1 being an issue). So you may still get the flat rate for data. Is this "cheating"? No, because the network is only delivering the grade of service that you contracted for, which is speech, and any ability to send data is without warranty of any kind. Furthermore, you _have to_ do this in order to go between COs. Without SS7, all interoffice calls default to speech, and data calls fail. Since the interoffice trunks are (almost?) all digital now, why not use them for data anyway, over the Speech bearer service. That's actually an official feature of the Digiboard IMAC remote ISDN to Ethernet bridge. It will do 56 kbps over Speech calls, complete with echo suppressor cancellation tone (needed for long haul inter-LATA use). And it's a capability that some telcos actually tell us to use -- Pac Bell mentioned it to customers as a work-around for the lack of SS7. I hope to "test" it over NET's network this year. So the data price is a compromise between different theories, and you don't really have to pay it anyway. What a country! :-) This, btw, really bewilders ISDN users in Europe. They don't even dream of trying this "speech" hack. They don't usually need to, anyway, and they worry that they'd not get PTT certification for equipment that ran that way. And they're too timid. And they don't have flat rate local calls to begin with. And it's not in the CCITT Recommendations. It's not even called out explicitly in the ANSI Standards. So it's a mandatory part of ISDN "folklore", which implementors have to know, but nobody wants to write it down. fred