Received: (from ptownson@localhost) by massis.lcs.mit.edu (8.9.1/8.9.1) id DAA22472; Wed, 26 May 1999 03:07:03 -0400 (EDT) Date: Wed, 26 May 1999 03:07:03 -0400 (EDT) From: editor@telecom-digest.org Message-Id: <199905260707.DAA22472@massis.lcs.mit.edu> To: ptownson Subject: TELECOM Digest V19 #99 TELECOM Digest Wed, 26 May 99 03:07:00 EDT Volume 19 : Issue 99 Inside This Issue: Editor: Patrick A. Townson E-Biz Bucks Lost Under SSL Strain (Jody Kravitz) Data Cabling Book (Fred Atkinson) Reciprocal Compensation (was Re: Strange Problems) (Eli Mantel) Re: DNA Dragnet (Martin McCormick) Re: Do You Feel Like a Number? (Hillary Gorman) Re: Smartjack and CSU (Mel Beckman) Re: World's First Video Cell Phone Debuts in Japan (John A Beckett) Re: Ma Bell Doesn't Live There Any Longer (Arthur Ross) Re: Sex Sites Getting Screwed (Ron Bean) Re: WW-XX-YYY-ZZZ (David Jensen) Re: 10-10- Dial-Around Company List? (Richard Shima) TELECOM Digest is an electronic journal devoted mostly but not exclusively to telecommunications topics. It is circulated anywhere there is email, in addition to various telecom forums on a variety of networks such as Compuserve and America On Line, and other forums. It is also gatewayed to Usenet where it appears as the moderated newsgroup 'comp.dcom.telecom'. TELECOM Digest is a not-for-profit, mostly non-commercial educational service offered to the Internet by Patrick Townson. All the contents of the Digest are compilation-copywrited. You may reprint articles in some other media on an occassional basis, but please attribute my work and that of the original author. Contact information: Patrick Townson/TELECOM Digest Post Office Box 765 Junction City, KS 66441-0765 Phone: 415-520-9905 Email: editor@telecom-digest.org Subscribe/unsubscribe: subscriptions@telecom-digest.org This Digest is the oldest continuing e-journal on the Internet, having been founded in August, 1981 and published continuously since then. Our archives are available for your review/research. URL information: http://telecom-digest.org Anonymous FTP: hyperarchive.lcs.mit.edu/telecom-archives/archives (or use our mirror site: ftp.epix.net/pub/telecom-archives) Email <==> FTP: telecom-archives@telecom-digest.org Send a simple, one line note to that automated address for a help file on how to use the automatic retrieval system for archives files. You can get desired files in email. ************************************************************************* * TELECOM Digest is partially funded by a grant from the * * International Telecommunication Union (ITU) in Geneva, Switzerland * * under the aegis of its Telecom Information Exchange Services (TIES) * * project. Views expressed herein should not be construed as represent-* * ing views of the ITU. * ************************************************************************* In addition, a gift from Mike Sandman, Chicago's Telecom Expert has enabled me to replace some obsolete computer equipment and enter the 21st century sort of on schedule. His mail order telephone parts/supplies service based in the Chicago area has been widely recognized by Digest readers as a reliable and very inexpensive source of telecom-related equipment. Please request a free catalog today at http://www.sandman.com --------------------------------------------------------------- Finally, the Digest is funded by gifts from generous readers such as yourself who provide funding in amounts deemed appropriate. Your help is important and appreciated. A suggested donation of twenty dollars per year per reader is considered appropriate. See our address above. Please make at least a single donation to cover the cost of processing your name to the mailing list. All opinions expressed herein are deemed to be those of the author. Any organizations listed are for identification purposes only and messages should not be considered any official expression by the organization. ---------------------------------------------------------------------- Date: Wed, 26 May 1999 00:08:02 -0700 From: Jody Kravitz Subject: E-Biz Bucks Lost Under SSL Strain Thought you might be interested in what I'm up to lately... The product(s) that I've been working on have kind of eaten my life for about the last 18 months and this seemed like a good excuse to get out a short piece of mail to you. The article pointed to by the URL below was such good news that I wanted to share it. My "home" correspondence address continues to be: kravitz@foxtail.com Jody By TIM WILSON A customer stuffs his shopping cart with goodies from your Web site. Credit card in hand, he waits for a secure connection to consummate the deal. And waits. Finally, short of patience, he dumps the contents and logs off. It may sound like an e-commerce manager's nightmare, but according to the latest Web server performance statistics, it's an increasingly common phenomenon. The ghost in the machine is Secure Sockets Layer, the commonly used method of securing communications between users and Web sites. Recent tests conducted by researcher Networkshop Inc. indicate that powerful Web servers capable of handling hundreds of transactions per second may be brought to a near standstill by heavy SSL traffic. Some server configurations suffered as much as a fiftyfold degradation in performance from SSL, down to just a few transactions per second, according to analyst Alistair Croll at Networkshop. The growing problem of SSL performance has driven vendors to develop devices that can help share the Web server's processing load. IPivot Inc. next month will ship two new processors that can offload authentication and encryption on e-commerce sites. IT managers and other experts have known for years that SSL, which requires the authentication and encryption of Web server connections, can significantly slow site performance. But the problem is rapidly becoming more chronic as companies increase secured Web transactions, they said. "Our business is very seasonal, and a lot of it is concentrated in the fourth quarter. This past December, we found ourselves shuffling servers around to handle the load," said Stephen McCollum, network architect at Hewitt Associates. The $858 million company manages benefits plans for large organizations, and because Hewitt's Web traffic is personal and confidential, virtually all of it is conducted via SSL. Hewitt is far from alone in its reliance on SSL. According to a study conducted by research company Netcraft Ltd., SSL implementations doubled from 15,000 sites to more than 35,000 sites between 1998 and 1999. And many of those server sites are struggling under the load. "I'd guess that somewhere between 10 and 25 percent of [e-commerce] transactions are aborted because of slow response times," said Rodney Loges, vice president of business development at Digital Nation, a Web hosting company. That translates to as much as $1.9 billion in lost revenue, using Forrester Research numbers for 1998 of $7.8 billion in e-retail sales. According to Networkshop, even the most powerful, general-purpose Web server hardware can be dragged down by large volumes of SSL traffic. In its most recent tests, the research company found that a typical Pentium server configuration running Linux and Apache, which at full capacity can handle about 322 connections per second of standard HTTP traffic, fell to about 24 connections per second when handling a full load of SSL traffic. A similar test conducted on a Sun 450 server running Solaris and Apache experienced even more trouble. The server handled about 500 connections per second of HTTP traffic at full capacity, but only about 3 connections per second when the traffic was secured via SSL. Networkshop tests of quad-processor configurations showed that those performance ratios scale to multiserver environments as well, Croll said. A few vendors, such as Rainbow Technologies Inc., have solved the problem by offloading security processing onto a dedicated co-processor card that slips into a server. But as SSL traffic increases, adding and managing co-processor boards becomes unwieldy, IT managers said. "We found that the [co-processor] cards were kind of a kludge, because they have to be added to every server," said Digital Nation's Loges. IPivot will begin shipping two external SSL processors -- the Commerce Accelerator 1000 and the Commerce Director 8000, which includes IPivot's load-balancing system -- to help eliminate SSL bottlenecks. The Commerce Accelerator 1000 is priced at $9,995; the Commerce Director 8000 costs $39,950. Research Data: http://www.internetwk.com/numbers/numbers052099.htm What's The Price Of Security? Research shows that SSL encryption can devastate Web server performance. ------------------------------ From: Fred Atkinson Subject: Data Cabling Book Date: Tue, 25 May 1999 16:50:42 -0400 Hello, Pat, I was wondering if any of the T. D. readers would know of a good reference manual to document the data about various types of cables/interfaces? Example, look up RS232 to find the pin outs, the electrical and mechanical (type of connector, etc.) specifications, and the maximum length an RS-232 cable can reliable be and still work. This is the type of information I'm looking for, preferably a small 'quick reference' guide. Then, look up the same type of info for V.35, RS449/422, etc. If anyone has any suggestions, please email me at: 'fatkinson@loralorion.com'. Thanks, Fred ------------------------------ From: Eli Mantel Subject: Reciprocal Compensation (was "Re: Strange Problems") Date: Tue, 25 May 1999 18:31:45 GMT Cortland Richmond (crichmon@telecom-digest.org) wrote: > My ISP tells me PacBell wants this changed. PacBell wants to get out > from under reciprocal compensation to ILEC's serving ISP's, and is > asking that calls to ISP's be treated as long distance. > ... >> THIS WOULD CAUSE HIGHER RATES FOR CALIFORNIA >> INTERNET USERS AND, IN MANY RURAL AREAS, COMPLETE >> LOSS OF LOCAL INTERNET ACCESS. Let's remember that the issue of reciprocal compensation doesn't directly affect what consumers pay for service. By the somewhat arcane rules in effect, reciprocal compensation is paid only on local calls, but classifying certain calls as long distance doesn't mean that consumers will be charged for those calls. It only affects what the phone companies involved pay each other. It's probably the case that the incumbent local phone companies, such as PacBell, made a tactical error in lobbying for reciprocal compensation, with the result that the competing local phone companies took advantage of the opportunity this created by signing up businesses to use their service for phone lines that made few or no outgoing calls, hence making them eligible to receive reciprocal compensation. That mistake aside, payment of reciprocal compensation on calls included in flat-rate service is extremely arbitrary, since both phone companies involved carry the call part of the way, and neither one receives any usage-sensitive charges from the consumer. Whether or not charges for local phone service will go up if reciprocal compensation continues to be paid or ISP charges go up if reciprocal compensation is dropped is somewhat speculative, but it seems to me that reciprocal compensation is primarily a subsidy from one phone company to another, and it's difficult for me to understand how such a subsidy can be justified. ------------------------------ From: wb5agz@dc.cis.okstate.edu (Martin McCormick) Subject: Re: DNA Dragnet Date: 24 May 1999 19:06:43 GMT Organization: Oklahoma State University in Stillwater, Oklahoma In article , Mike Riddle wrote: > Perhaps someone with a longer (and better) memory than I have > can recount whether fingerprints went through a similar "growing > phase." I have read that there must be at least 10 points of positive identification between a known finger print and the one which was lifted from a crime scene. There have also been cases of police misconduct in which finger prints were simply lifted from finger print cards and then made to look like crime scene prints in order to frame a suspect. Of course, they always matched perfectly. I remember reading a magazine article many years ago in which the FBI Crime Lab was sent a finger print record of a suspect plus some crime scene prints which matched. The technician smelled a rat when a look at the crime scene prints under a microscope showed an odd thing. The finger print powder had a much different grain size than it usually does. It turned out to be Xerox toner. Somebody had just photocopied that card and then used finger print tape to lift the toner off and try to pass it off as a lifted print. It was lifted, all right. [TELECOM Digest Editor's Note: Talking about Officer Friendly, and how he and his fellow officers Serve and Protect each other with a 'code of blue silence', wasn't it simply grand to read in the papers on Tues- day that the New York City officer who liked sodomizing his prisoners was found guilty of the attack on Mr. Louima ... he is eligible for thirty years in prison without parole because of the violent nature of his crime; a punishment police officers are more than happy to recommend for everyone else of course. The other four or five officers involved are still stonewalling; refusing to testify. I think it would be a good idea to torture them until they confess; what is good enough for the people they arrest should be good enough for them as well. With a little luck, they will all get thirty year prison terms, in a place where the other inmates know what their former occupations used to be. I know, that is a terrible thing to say; sorry, it is how I feel. PAT ------------------------------ From: hillary@hillary.net (Hillary Gorman) Subject: Re: Do You Feel Like a Number? Date: 25 May 1999 13:10:21 GMT Organization: Debugging our net or deworming your pet... On Sun, 23 May 1999 05:12:34 +0100, wrote: >> TELECOM Digest Editor wrote... >> It used to just be a grim joke: one day all Americans would be >> issued numbers to use instead of names. Well, we are about half- >> way there. So far, we are still permitted to have individual >> names, but our social security numbers are rapidly becoming our >> national ID numbers. > How do (or will) these various organisations across the US (banks, > educational institutions, state governments, and airlines) cope with people > who (like myself) don't have a Social Security Number? > [TELECOM Digest Editor's Note: Then you will have to get a number or > the bank won't be able to work with you. PAT] I found this at the ssa.gov website: I do not have work permission from the INS. I need a Social Security number. Can I get one? ANSWER: Yes. The operative word is need. You can get a non-working Social Security Card when you are required to have one by law. This means that if you are in the U.S. legally and someone says you "must" have a social security number, ask for proof that you are legally required to have one. You should submit this proof along with your SS-5 Social Security Card Application. This number is not valid for employment. If you ever use this number for employment, the Social Security Administration can advise the Immigration Service. You must show Social Security a valid reason for this number. Reasons include identification for school or to open bank accounts. You can change the number to a working number when you get work permission from the INS. In practice, the Social Security Admnistration will seldom, if ever, give out non-working numbers. The SSA's position is that an Individual Taxpayer Identification Number (ITIN) takes care of the needs of any non-working Canadians. The ITIN is a nine digit number issued by the IRS to people who need a number, but who cannot get a social security number. Apply for an ITIN using form W-7, Application for IRS Individal Taxpayer Identification Number. hillary gorman...........Official Token Female..........hillary@netaxs.com "So that's 2 T-1s and a newsfeed....would you like clues with that?" hillary@hillary.net: for debugging your net or deworming your pet Net Access...The NSP for ISPs....The NOC that rocks around the clock. ------------------------------ Date: Sat, 22 May 1999 23:51:15 -0700 From: Mel Beckman Subject: Re: Smartjack and CSU At 10:37 PM -0400 5/22/99, Kevin Lundy wrote: > Is there any recommended maximum distance from the smartjack to the > CSU? I have a new T1 voice service that is giving me hell. It's an > intermittent problem - slips, dropped calles, static, etc. The line from the telco-provided smartjack to your CSU must be adequate for T1 signaling. I've installed hundreds of T1 circuits, and have learned that you can't fudge the requirements here. If you're only going a few feet, an ordinary unshielded twisted pair (UTP) Cat5 Ethernet patch cable works fine. Beyond that distance, however, you need to use shielded cable, such as D-Screen or T-Screen. The cable needs to have two pairs, each pair separately shielded. This cable can be hard to get, however, and is quite bulky, making it difficult to install. An alternative is to use shielded Cat-3. Because there is only one shield for the whole four-pair cable, you have to use two STP Cat3s to do the job. Run the send pair (pins 1 & 2 on the RJ45) on one cable, and the receive pair (pins 4 and 5 on the RJ45) on the other cable. Finally, you must ground the shielding at one and (AND ONLY AT ONE END!). Use a good-quality earth ground, such as a cold water pipe or, if you're really lucky, an established telecommunications ground strap. If you don't follow these rules, you end up with the sorts of intermittent problems you describe. Note that it's possible, with the STP Cat3 approach, to carry four T1s over one pair of cables. All the Send pairs go in one cable, and all the Receive pairs go in the other. - Mel Beckman ------------------------------ Date: Wed, 26 May 1999 05:39:50 GMT From: John A Beckett Subject: Re: World's First Video Cell Phone Debuts in Japan > TOKYO -- The world got its first glimpse at what could be the future > of mobile telephony this week when Japanese component vendor Kyocera > Corp. unveiled the first cellular phone able to transmit a caller's > picture and voice simultaneously. Would somebody kindly tell me an application for this technology? Its best use seems to be material for Jay Leno's jokewriters or "Saturday Night Live." The punch line of course would be a freeze-frame of the front of a Mack truck reflected in the caller's eyeglasses, the sound track being part of a crash followed by whatever noise is made when a cell connection is broken. /\--. John A. Beckett "Never tire of doing / \ ) Southern Adventist Univ. right." 2 Thess. 3:13 /----\---. voice: (423) 238-2701 \ / \ \ FAX: (423) 238-2431 `-' `--' home: (423) 396-2453 jbeckett@southern.edu ------------------------------ Date: Tue, 25 May 1999 08:39:13 -0700 From: Arthur Ross Subject: Re: Ma Bell Doesn't Live There Any Longer > Yes, Pat, it has been some time since I've posted to the Digest. I've > just become burned-out and dis-illusioned by what I've been seeing > going on in the telephone industry over the past year or so! :-( > mjc > [TELECOM Digest Editor's Note: Burned out and dis-illusioned? No kidding! > Nothing is the same any longer is it? I will quote the title of a book > by Thomas Wolfe: "You Can't Go Home Again" ... Ma Bell doesn't live > there any longer. PAT] Pat - This is a MARVELOUS title, a la: "Basic Basic" (book on programming), "The Umpire Strikes Back" (book by baseball umpire), etc, etc. Somebody (you?) should write the book to go with it! -- Best -- Arthur -- Dr. Arthur Ross 2325 East Orangewood Avenue Phoenix, AZ 85020-4730 Phone: 602-371-9708 Fax : 602-336-7074 [TELECOM Digest Editor's Note: "Ma Bell Doesn't Live Here Any Longer: The Decay of the USA Telephone Network Since 1984." Subtitled, "One Hundred of the Best Horror Stories About Telephone Service Which Appeared in TELECOM Digest" ... I dunno, does someone want to help me write it? PAT] ------------------------------ Subject: Re: Sex Sites Getting Screwed Date: Tue, 25 May 1999 20:35:51 CDT From: Ron Bean hapgood@pobox.com (Fred Hapgood) writes: > I disagree. These carts serve as custom catalogs and I love them. > The sole reason I don't buy at amazon.com is that their shopping cart > times out and I have found two other bookstores (altbookstore.com and > books.com) who do not impose that inconvenience on me. I take the opposite view-- when I buy books, I know what I want, and I don't need any leftover cruft getting into my current order. > The right place for a list of the books I think I might > order sometime is the website of my online bookstore. "On their website" is not necessarily the same as "in a cart". There are other ways to accomplish this. At the very least, they should give you a way to tell them "save this order for later", so they can get rid of the ones that nobody is going to come back to. I was mildly dismayed to discover that none of the "major" online booksellers (Amazon, B&N, Borders) will allow me to place an order without creating an "account", so that I have to type in the same password every time I place an order with them. I have enough passwords to remember already; I just want to order some books, should be no big deal. I don't need a password to order from a paper catalog. It doesn't bother me that they *offer* accounts to people who want them, but it seems odd to *require* something that is a convenience to some people but an inconvenience to others, and which doesn't really add anything to their ability to fill the order. If this trend continues, I'll need to remember a separate password for every vendor I do business with on the net, even though my individual orders are separate transactions that have nothing to do with each other. They can already collect marketing data about my ordering habits; they don't need a password for that. [Note to anyone who missed the word "mildly" in the above -- I'm not losing any sleep over it, but it seems totally unnecessary. I know there are lots of other online booksellers, but I haven't had time to investigate *all* of them, and so far I haven't run across any others that have a complete database of all books in print.] [TELECOM Digest Editor's Note: One of the reasons for the creation of username@telecom-digest.znn.com anonymous email accounts was to assist partly in giving people a way to transact business on the net; stick around long enough to make sure the order was handled correctly and then ditch the email address. Let it be the focal point for all the cookie collectors/processors instead of your real email address. If you want one, obtain it at http://telecom-digest.org/postoffice PAT] ------------------------------ From: djensen@madison.tds.net (David Jensen) Subject: Re: WW-XX-YYY-ZZZ Date: Tue, 25 May 1999 14:06:38 GMT Organization: At My House Reply-To: djensen@madison.tds.net On Tue, 25 May 1999 10:29:45 -0400, in comp.dcom.telecom Eric@AreaCode- Info.com (Eric B. Morson) wrote in msg : > How can a local number format YYY-ZZZ work when there are already > dozens if not hundreds of NXXs in a local area? If you reformatted the > 10 digit format from > NPA-NXX-XXXX to > WWX-XYY-YZZZ > ^ ^ > | | > 1) Consider these two digits in the sequence (I reformatted your > dialing pattern into the current 3-3-4 for the sake of comparison). > How do you propose converting NPA-NXX to WW-XX when the second "X" is > the first digit of the current NXX? the VAST majority (80-90%??) of > phone numbers in the NANP would have to change it's first digit at the > very least. MASSIVE disruption of communications would be inevitable > and the cost? Unbelievable today. Yes, there would be a one-time change from NPA to WW-XX. Aside from changes in NPA, fewer than 5% (more likely 2%) of all N's in NXX'x would have conflicts that would necessitate change in the NXX. The last six digits would remain unchanged. > 2) If you propose 2 digit sub regional codes: "XX is an optional sub > region. Larger metro areas would just use this as part of the number", > how do you suggest those Larger metro areas actually use the XX codes > WITHOUT changing every local 7 digit number's leading digit? (see #1) Metro regions would use 8 digits in their local calling area. The leading eighth digit would most commonly be the last digit of today's NPA. > 3) I won't even go into the politics and public outcry that would result > from people and businesses losing their NPA "identity" and eliminating > people's ability to know inherently that 213 is Downtown LA, 212 is NYC, > etc ... and again, the cost to every consumer ... Some will have identities, doesn't NY as the region code work? Why not LA? It costs less to make one change that will survive for decades than suffer the constant drip drip drip of splits, overlays and fights at the PUC. > People are complaining that their area code has to change, or they now > have to dial 10 or 11 digits to order a pizza ... imagine changing > every single local phone number en masse? A mess, don't you think? and People will dial no more than eight digits, not ten or eleven to get their pizza. They will keep at least the final six digits of "their" number. > no matter how you slice it, your 2 digit WW region code itself still > has a limited number of combinations within it, and there are fewer > WWs to go around than there are NPAs. How could Chicago, LA, or NYC go > to a single WW region code, since the XX sub region codes would force > the changing of all local numbers at once in addition to the loss of > their current code, and forcing 8 or 6 digit local dialing? How many > different fundamental changes in the way people dial and think of the > dialing sequences can we affect at once? Let's use Chicagoland as an example. Instead of six, seven or eight area codes, Chicagoland gets one region code. The folks in 312 will have a number that starts, say, 24 2X YYY ZZZ, the folks in 708 get 24 8X YYY ZZZ with no change in their base number except that they again have a common region code and have to dial eight digits instead of seven or ten. The shortage is not of numbers. The impending NANP shortage and the current local shortages represent a flaw in the assignment method. Grandfather current assignments, but hand out 1000 numbers at a time in the future. Reserve regional codes for wireless overlays for the five or six largest metro areas. With assignment control and overlay control, we will not require region code changes or an eleventh digit in NANP until the middle of next century. Right now, every area code split already represents a possible effect on everyone else in the phone network. Splits should not be treated as cavalierly as they are today by NANPA, the PUCs and telcos. We cannot even rely on our telco's phone books to show accurate information. Not only would a major reform stabilize the numbering system, but it would force local telcos to update their obsolete information. ------------------------------ From: Richard Shima Subject: Re: 10-10- Dial-Around Company List? Date: Wed, 26 May 1999 00:14:26 -0400 Organization: The Shima Company Barry Margolin wrote: >> This page has the definitive list: >> http://www.nanpa.com/number_resource_info/carrier_id_codes.html > What I've been wondering is if anyone has compiled pricing details for the > 10-10 codes, so that consumers can compare them easily to find the ones > that best match their calling patterns. Barry, Try looking at this interesting new site for 10-10 comparisons: http://www.10-10Phonerates.com/ Richard Shima | Internet: RShima@att.net The Shima Co. | CompuServe: 74037,2425 Mayfield Heights | FidoNet (Point on Nerd's Nook): 1:157/2.10 Ohio 44124 USA | Voice & fax: 440 461-4357 | Radio: WB8MTE ------------------------------ End of TELECOM Digest V19 #99 *****************************