#2 33 07 Aug 85 06:32:49 $0.00 To: Sysop And All On on Fido #100 in Net #115 Subject: Long

---
Master Index Current Directory Index Go to SkepticTank Go to Human Rights activist Keith Henson Go to Scientology cult

Skeptic Tank!

#2 33 07 Aug 85 06:32:49 $0.00 From: Donald Larson on Fido #333 in Net #115 Attache_Node, Schaumburg...........IL To: Sysop And All On on Fido #100 in Net #115 Subject: Long Distance Services Everyone has asked at one time or another which of the long distance services are the best for data transmission. This is especially true for the person who calls all over the country in search of that one program that will make their life easier. Here's the story, according to an article in the August, 1984 issue of Data Communications. The editors of the magazine made the following tests and observations. Carrier Connect Attempts 1st Try BER Error BLER Signal Time Required Burst Loss Allnet 17.86 1.07 93 0 0 0.0 -21.5 AT&T 14.16 1.07 93 105 0.99 1.0 -18.7 ITT-USTS 19.68 2.93 40 2535 9.02 72.1 -20.6 MCI 14.96 1.27 73 427 1.82 14.4 -24.2 SBS 21.72 1.13 87 64 0.50 1.5 -18.9 SPRINT 16.40 1.67 56 96 0.50 4.5 -18.3 TELESAVER 26.67 1.40 80 4314 23.3 109.0 -19.5 WU 19.78 2.60 47 221 1.64 7.4 -24.8 Connect Time - Seconds from last digit to answerback from distant modem Attempts - Average Number of calls required to complete call to other end 1st Try - Percent of calls completed on first try BER - Bit error rate, number of bits in error per million bits Error Burst - number of bursts where 80 consecutive bits in error per million bits transmitted BLER - Block error rate, number of 1,000 bit blocks per 1,000 blocks Signal loss - loss of signal level (in db) at mid frequency (1004 hz) This is installment 1 of 3 installments. More later. Don Read Command [2] 1 - 26, * : #3 30 07 Aug 85 21:39:21 $0.00 From: Donald Larson on Fido #333 in Net #115 Attache_Node, Schaumburg...........IL To: Sysop And All on Fido #100 in Net #115 Subject: Long Distance Services (Part 2) OK folks, here's installment 2 on long distance services for data transmission. Again, this information is from an article in Data Communications, August 1984. Here are some additional pieces of data: Carrier Slope S/N P/AR Jitter Billing Allnet 5.4 27.8 57.5 5.8 8.4 $3.52 AT&T 1.7 31.0 71.3 4.5 9.0 $5.06 ITT-USTS 6.3 25.4 55.3 10.0 9.0 $3.95 MCI 7.8 27.9 58.4 6.7 10.0 $4.31 SBS 5.6 30.4 50.0 4.1 9.4 $3.67 SPRINT 7.4 28.7 63.6 6.1 10.0 $5.10 TELESAVER 7.3 26.1 41.9 8.5 11.0 $4.91 WU 8.6 29.0 62.8 5.5 11.0 $4.73 Slope - Difference in attenuation (or loss) between middle frequency (1004) and the high and low levels (averaged). (Bell Spec < 14) S/N (Signal to Noise) - Measurement of signal strength over noise (static) background. (Bell Spec > 24 db) PAR (Peak to Average Ratio) - Measurement of velocity distortion according to frequency. Medium speed spec from Bell is > 48. Jitter (Phase Jitter) - Measurement of AC influence upon signals, measured in degrees. Maximum permitted by Bell is 10 degrees. Especially rough on phase modulated modems (212A is Phase shift keyed). Billing - Actual time was 8.5 minutes. Amount shown is the amount the carrier billed. This may be due to billing errors and/or minimum connect times (particularly important for Fido's short connects). Dollar amounts are for coast to coast tests and may not be accurate for other comparisons. In the next and final installment, I'll discuss these stats in a management overview style (excerpted from the article and annotated by my experience) along with personal observations. Stay tuned for chapter 3 (Summary and Recommendations along with a discussion of equal access and the national toll network). Don Read Command [3] 1 - 26, * : 3 4 #5 37 13 Aug 85 16:18:05 $0.00 From: Donald Larson on Fido #333 in Net #115 Attache_Node, Schaumburg...........IL To: Sysop And All on Fido #100 in Net #115 Subject: Long Distance Services - Part 3 This is the executive summary of the previous two articles. Allnet - Best cost and bit error rates. Suffered slightly in analog line measurements. Data applications are highly ranked but the voice levels may suffer. Excellent prices and billing accuracy. AT&T - Best overall quality for mixed voice and data applications. In general, the rates they charge are not bad for the quality compared to the batting averages, rate accuracy, etc. Still the standard to compare against. GTE Sprint - Good quality for data, less for voice. The general call "batting average" for first try call completion. Ended up as the highest call cost on timed calls. Their actual rates are lower than AT&T but the billing accuracy was poor. ITT USTS - Poor by almost all standards, except for price. Recommended for patient voice callers with plenty of time. MCI - Mediocre for data. Voice quality probably acceptable but data performance not as good as others available. SBS - Second best in savings and surprisingly good quality. A little long in call set up time. May be some problems with over 1200 bps data due to relatively poor P/AR rating. Heavily recommended by Data Communications's magazine tests. Telesaver - Worst bit and block error rates overall. Minimal savings over AT&T does not justify for data or even voice. Western Union - Poor call completion first attempt average; often required more than six attempts to connect. Passable data performance but marginally. Not recommended for short voice and data calls but savings may be worth it on long length rates. I personally have been using SBS for several months with a great deal of success. I highly recommend it. I have also used MCI but had a greater problem with error rates. My best advice is to try out the carrier and see what you see. Don PS - Let me know your experiences; I'll try to collect them and publish them occasionally.

---

E-Mail Fredric L. Rice / The Skeptic Tank