💾 Archived View for gemini.spam.works › mirrors › textfiles › messages › BACKWATER › bw830203.txt captured on 2022-06-12 at 15:41:48.
-=-=-=-=-=-=-
NUMBER OF LINES: 625
1 If you are in need of help, you need but ask....
2 ********************** REMOVED: 3 FEB 83 ***************************
3 Welcome to BWMS (BackWater Message System) Mike Day System operator
4 ************************************************************
5 GENERAL DISCLAIMER: BWMS IS NOT RESPONSIBLE FOR ANY INFORMATION PLACED ON
6 THIS SYSTEM.
7 BWMS was created as an electronic bill board. BWMS is a privatly owned
8 and operated system which is currently open for use by the general public.
9 No restrictions are placed on the use of the system.
10 It is intended that the system be normally used for messages and
11 advertisments by the users. As the system is privatly owned, I retain the
12 right to remove any and all messages from the system which are offensive
13 to me. Additionally because of the limited size of the system, it will be
14 periodically purged of older messages. (only 629 lines of data can be saved)
15 The saved information will be cycled to drive 'B' while the information on
16 drive 'B' will be archived, and a fresh disk will be installed in drive 'A'.
17 To leave a message, type 'ENTER' and use ctrl/C or break to get out
18 of the enter mode. The message is automaticly stored.
19 If after entering the message you find you made a mistake,
20 use the replace command to replace the line.
21 To exit from the system, type 'OFF' then hang up.
22 Type 'HELP' to see other commands that are available on the system.
23 ========================================
24 *******************************************************
25 NOTICE.NOTICE.NOTICE.NOTICE.NOTICE.NOTICE.NOTICE
26 O C
27 T P. C. S. MEETING LOCATION CHANGE I
28 I T
29 C The Portland Computer Society will not O
30 E be meeting at the Far West Federal any N
31 . more. .
32 N E
33 O We will now be meeting at the P.C.C. C
34 T Sylvania campus at 12000 SW 49th. I
35 I The meeting time is 15 Jan 83 at 1:15 PM T
36 C Doors open at 1:00 PM. Room ST-B14 O
37 E Everyone is Welcome! N
38 . .
39 N P.S. The Feb meeting is the 19th, and the E
40 O March meeting is the 19th also. C
41 T I
42 I The Sylvania campus is just off I-5 T
43 C about 1/4 mile past the Mcdonalds. O
44 E N
45 . .
46 NOTICE.NOTICE.NOTICE.NOTICE.NOTICE.NOTICE.NOTICE
47
48 ****************************************************************
49 Actually Price is the determinator of which processor
50 gets used in a particular computer.
51 1) price of hardware - sometimes cheapest chip is chosen.
52 2) price of software - maybe manuf. does'nt
53 want to go to the expense of writing new software.
54 3) gains from exclusivity - Atari and TI do this:
55 by making their machines incompatable with almost everything
56 else, they then control the market for software
57 and hardware. (at least they think they can)
58 **************************************************************
59 I HAVE A MOTORCYCLE FOR SALE OR PARTIAL TRADE FOR
60 COMPUTER GOODIES. IT'S AN 1980 SUZUKI GS1100 E. GREAT
61 BIKE, BUT I RARELY RIDE IT. LOW BLUE BOOK IS 2500. I'D
62 TAKE A DAISYWHEEL PRINTER OR AN EPSON 100 IN PART TRADE.
63 I ALSO NEED A TRAILER FOR A 23' BOAT OR WHATEVER.
64
65 =====================================================
66 LEAVE ME A MESSAGE HERE OR CALL BILL BABCOOCK xxx xxxx EVENINGS
67 *-*-*-*-*-*-*-*-*-*-*-* FOR SALE -*-*-*-:=*-*-*-*-*-*-*-*-*-*
68
69 1967 PONTIAC FIREBIRD CONVERTIBLE.
70
71 - POWER STEERING
72 - POWER BRAKES
73 - 400 CUBIC INCH, 325 HORESPOWER
74 - TURBO 400 AUTOMATIC W/CONSOLE SHIFTER
75 - 5 MONTH OLD RED ACRYLIC ENAMEL PAINT, CAR GARAGED SINCE
76 PAINTED. BODY VERY STRAIGHT.
77 - NEW (<200 MI.) RADIALS ON PONTIAC STOCK MAGS.
78 - NEW HEATER CORE AND CONTROLS
79 - TOP IS ORIGINAL, IN GOOD CONDITION - NO LEAKS/TEARS.
80
81 ASKING $4300.00 - HOWEVER, PRICE IS NEGOTIABLE.
82 IF INTERESTED, LEAVE A MESSAGE ON 'BIT BUCKET' (xxx-xxxx),
83 THIS SYSTEM (BWMS) OR CBBS/NW.
84
85 *-*-*-*-*-*-*-*- RICK BENSENE 1/09/83 11:13PM -*-*-*-*-*-*-*
86 have atari 2600
87 for sale
88 in v.g cond. asking $90.
89 if interested
90 call xxx-xxxx
91 ask for tom
92 would like to trade that
93 atari + cash for VIC-20
94 compatable printer
95 or 16K RAM card - TOM
96 %@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%
97
98 off
99 the sooner you fall behind
100 - the more time you
101 have catch up !!
102 @@@@@@@@@@@@@@@@@@@@
103 pat m. - give me a call
104 -t.m. @ S.W.
105 TOM M.-LEAVE MSG ON THIS SERVICE
106 PAT M.
107 EXIT
108 LIST
109
110
111 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
112 Question time;
113 Why is the 16 bit technology better then the 8? Most of the jobs
114 running on microcomputers seem to be business programs or word processing.
115 When I am running WordStar, I am storing 7 bit ASCII codes in a 8 bit
116 cell, if I run WordStar on CP/M-86 I store the same 7 bit code in a 16 bit
117 cell. This is progress?
118 Does anybody have a real reason to buy a 16 bit machine? By the way,
119 I do not class the IBM-PC as a 16 bit machine, it's a 8 bit running 16 bit
120 software. (in my book a 16 bit machine better have a 16 bit data bus!)
121 Can anyone think of a reason to buy a 16 bit machine that does not
122 apply equally to the coming 32 bit technology?
123
124 the Phantom Glitch.
125 1/10/82 20:30
126 -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
127 Does anyone have a copy of FIG-Forth Ver. 1.4V for the Atari?
128 Leave a message on any of the BBS's in town for Pann McCuaig.
129 Thanks.
130
131 -=-=-=-=-=-=-
132 P.G. -- no, not really, for most folks it's part of the me too syndrome.
133 I feel that current MICRO 16-bit technology is an intermediate step, and
134 that a discrete jump in User-available computing power awaits 32-bit CPU's,
135 various dedicated co-processors, and even cheaper and denser memory than is
136 currently available. Of course, a REAL conceptual gain in software
137 could obviate all hardware-oriented arguments. --Pann--
138
139 ???????????????????
140 While 16 bit machines offer no clear-cut advantages over current 8
141 bit machines, there is a reason to buy them: They will become
142 obsolete somewhat later than 8 bit ones. If you have no overriding
143 reason one way or another, go with the more recent technology.
144 I realize that's only a marginal reason, but for a unsophisticated
145 user, may be the best course.
146 For nearly all word-processing applications, an 8-bit
147 system works just as well as a 16 bit machine. That's why I may
148 just continue using my Z80, skip the 16 bit revolution, and buy
149 again when 32 bits is king. For someone entering the field just
150 now, and has no previous investment in hardware, 16 bits may be
151 as good or even better.
152 ??????????????????????
153 If you are waiting for the next better thing, you had better set in for
154 a long time: there will always be a newer and better just over the hill.
155 The advantage of a 16 bit machine is, at the machine code level: more
156 flexible instruction set, bigger address range, higher power arithmetic
157 and logic functions; at the high level code level, bigger addresses, maybe
158 speed from the things mentioned earlier; at the user level, well?? maybe..
159 If 8086 Wordstar wastes 1/2 the 16 bit word to store a letter, it is
160 because it isn't a 16 bit program, just a converted 8 bit program. This is
161 why most current 16 bit programs aren't much better: they aren't 16 bit
162 programs. (Of course, most z-80 programs are really 8080 jobs, too.)
163 The question of what is an 8 or a 16 bit processor is not so simple. The
164 data bus to memory is important only if the memory is the slowest thing
165 in the system. For the 8088 it may or it may not be, and their memory
166 manager helps make it less important. An 8 bit machine of current vintage
167 must wait for memory because it can't do anything else. Lots of 16 and
168 bigger machines use half word memory access, like PDP-11's, and IBM 360's
169 (of course only in some cases, the bigger ones in each series don't.)
170 As has been noted, anything that works is 'good enough', but a computer
171 is ususally considered "general purpose", and therefore will be used for
172 things it will do well, as well as being pushed to do more; when a bigger
173 machine is better. However, I have a bigger abacus than you have, so it is
174 obviously better. 29 rows of beads. It adds faster than I can shuffle em,
175 but is very intolerant of being jiggled.
176 I don't reccommend waiting for HAL, he'll be here too soon anyway.
177 ......................................B.A.D....................
178 %@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%
179 PAT M. - GLAD TO SEE THAT U GOT ON... Y A WED NITE MEETING ???
180 NEED TO GET TOGETHER SOMETIME SOON AND COPY SOFTWARE. TALK JOE INTO A MODEM ??
181
182
183 maybe we shud use BIT BUCKET so can access msg w/o going
184 thru lots of bs. ??(t.m.)
185 Anyone interested in observing a PMS system locally, contact me for phone
186 demo. By appointment only. Dennis Work Phone (xxx-xxxx)
187
188 *******************************************************
189 WHAT IS ALL THIS DISCUSSION ABOUT 8 BIT SOFTWRE RUNNING ON 16 BIT MACHINES
190 THE ADVANTAGE OF 16 BITS LIES IN THE POTENTIAL TO ADDRESS A VERY
191 LARGE BUFFER AND THEREFORE RUN SOFTWARE THHAT TALKS TO COMPUTER ILLITERATES
192 THEY ARE NOT FOR YOU GUYS TO BEGIN WITH! SURE, THE SOFTWARE DOESN'T
193 EXIST YET. WHAT WAS AVAILABLE FOR THE ALTAIRS AND SWTP 6800.
194 OBVIOUSLY THAT COMPARISON IS LIMITED, THE EIGHT BIT SOFTWARE MAKES THE 16
200 COMING TO A THEATER NEAR YOU.....................
201
202 D. T. -- THE SHAKY TERRESTRIAL
203
204 A YOUNG BOY BEFRIENDS A SKID ROW WINO IN ORDER TO HELP
205 THE BUM GET BACK TO 3RD AND BURNSIDE. IN THE COURSE OF
206 GETTING HIM THERE, THE FEDS ARE HOT ON THEIR TAILS SINCE
207 THE WINO IS AN EX-PRESIDENT........
208
209 -------------------------------------------------------------
210
211 COMING SOON TO A THEATER NEAR YOU.............
212
213 THE REVENGE OF SPOCK
214
215 MR. SPOCK AND DARTH VADER GET IT ON IN THE BIGGEST MOST
216 EXPENSIVE SPACE DUEL EVER. THE SPECIAL EFFECTS FOR THIS ONE
217 HAVEN'T EVEN BEEN CREATED YET.
218
219 --------------------------------------------------------------
220
221 NEED A GREAT HEADACHE REMEDY....COME TO THE FOLKS AT
222 JOHNSON & JOHNSON. WE HAVE A NEW PRODUCT....
223
224
225
226 IT'S CALLED......
227
228 D E A T H D R U G
229
230 IF IT DOESN'T GET RID OF YOUR HEADACHE YOU NEED NOT RETURN
231 FOR YOUR MONEY BACK SINCE YOU WON'T BE AROUND TO COLLECT.
232
233 ---------------------
234
235 THE ABOVE MESSAGE WERE BROUGHT TO YOU BY THE GREATER PORTLAND
236 AREA WACCO CLUB....MAKING THE
237 WORLD BETTER THROUGH STUPID
238 HUMOR........
239
240
241
242 END OF LINE
243 *****************
244 The fastest computer in the world is now a CDC something-or-other
245 with a throughput of very close to 800 Megaflops (floating point
246 operations per second). Wowwwwwwwwwweeeeeeeeeeeeeeeeeeeee!
247 I want one................
248 ******************
249
250 whats all the excitement about andway, isn't an atari 2600 good enough for
251 you guys? If not , then I don't know what to tell you.
252 SLOTH
253 ************************************************************
254 Talking about 8-bit micros:
255 Well, the shovel still isn't obsolete even though
256 there are now machines to dig holes faster.
257 ****************************************
258
259 The 'Fastest' computer in the world alluded to above is the
260 CDC Cyber 206. And although they claim that is is the fastest computer
261 in the world, the 'ole' Cray 1 still can do some things faster. Operations
262 per second isn't the only way to judge a computer's speed!
263
264 _________________________
265 The Cyber mentioned above is a machine with a parallel architecture,
266 essentially meaning that a large number of separate operations can
267 be done simultaneously. Some people object to giving them the title
268 of 'fastest computer' because they aren't strictly a 'von neumann'
269 architecture. The Cray is essentially a serial machine, similar to
270 most microprocessors.
271 As it turns out, many tasks which require a large amount of
272 computer time can use parallelism, as in the Cyber. So which is
273 faster? I'm not really sure.
274 -------------------------
275 ARF.
276
277 `````````````````@@@@@@@@@@@@@@
278 HEY GUYS !!! IT S. FAYE. ANY
279 BODY (GUYS) HAVE A APPLE COMP.
280 IT BEST MACHINE. OLD WIVES TALE
281 S ON BURNSIDE USE APPLE. 16 BIT
282 COMING SOON. BYE MEN.
283 @@@@@@@@@@@@@@@@```````````````
284 ****************************************PANN: I HAVE fig_FORTH FOR ATARI
285 BY PAT MULLAKEY (WHO NOW IS A LOCAL RESIDENT).
286 I ALSO HAVE DOS-COMPATIBLE FORTH. YOU'RE WELCOME TO EITHER.
287 CALL ME AT xxx-xxxx.
288 ...CU...PAT
289
290 ****************************************
291
292 So the CDC is a parallel machine? So what, most computers are, at least
293 4 or 8 bits parallel. However, I was just perusing my new Motorola catalog
294 and saw just what I was looking for. No silly 16 bit machine, not even 4;
295 A 1 (ONE) BIT COMPUTER!! You got it. Now, would somebody tell me what
296 we can do with a one bit computer?????
297
298 Speaking of speed, the latest Byte had an article on benchmark speed,
299 and the fastest machine they listed was an IBM 3033. The Cray was way
300 down, but of course the different computers weren't always running the
301 same kind of code.
302 Assembly coded programs ran fastest, the various compilers went from almost
303 as fast to s l o w w w,, and the interpreters like basic and Pascal were
304 v e r y ss ll oo w wwww . I think the range was something like .003 seconds
305 to 3000 or so. The program type (assem vs comp vs interp) made more differ
306 than the machine it ran on. In other words you speed freaks, learn to
307 program for speed, don't ask for a faster toy.
308 .....................................B.A.D........1-14-83/20:54 (p.m.)...
309 **************************
310 The problem with 'programming for speed' is that this is used as an
311 excuse for poor programming. Example: Basic programs run faster if
312 spaces are removed. Result? People write unreadable programs, and
313 counter deserved criticism with 'I'm just trying to make it run
314 faster.' Also, in some earlier Basics, frequently used subroutines
315 if placed in the beginning of the program, ran faster.
316 This is a poor state of affairs, since a lot of effort is
317 wasted on such mistakes. It's a lot easier, in the long run, to
318 make a computer that's twice as fast than to try to rewrite an
319 already-close-to-optimum program written in most any compiled
320 language.
321 I'm not saying to ignore the software. But if it can't
322 be made readable and logical, accept the slower speed and don't
323 try to hack a program to death.
324 *****************************
325 Micro-80 had an article on proggramming for speed, that changed the time
326 of a simple demo program from minutes to a few seconds. There is no way
327 we will see a small computer that is 100 times faster than what we have
328 now. A program adjusted for speed is not necessarily a bad program, or
329 an example of bad programming style. They may often have things about
330 them that is not (by themselves) good, like no spaces between words,
331 or no remarks, but so what? You can't read the runnable code for a Pascal
332 or FORTRAN program, either. The remarks can be in a separate file, or
333 maybe the instruction book where they belong. The difference between good
334 and bad programs is far greater than the diference between the best and
335 the worst computer you cas buy. (And this is far greater than the
336 difference between the CPU chips in them.) The ability to speed up, or
337 slow down a program by rewriting it, or rearranging it is not a bug, but
338 a characteristic of any language. A point to be aware of, though, is that
339 the same tricks do not always work in some other language or dialect.
340 If removing the spaces doesn't change the speed, using integers, or
341 for-next loops probably will.
342 I don't believe the program should contain comments, anyhow. Put these
343 in the documentation. This is of course something most programmers don't
344 seem to be able to do: document your code. In fact I sometimes wonder if
345 most of them can even speak english, let alone write.
346 But then, I hear a scream that this will make code unreadable, but
347 again so what? There is no programming language that is easy to read,
348 because it wasn't meant to be, except for a computer. People speak Human
349 languages, so describe what you did in a human language, if you can.
350 Good lick there........................B.A.D.....1-16-83..18:08..........
351 For once I tend to agree with you-
352 However some BASICs allow the use of long
353 names for variables -- these can make reading
354 the program easier. But as far as that goes
355 with (a lot of) practice you can read even
356 disassembled object code.
357 *****************************************************
358 B.A.D. you are out to lunch! There is nothing wrong with separate documentation
359 files, but the code itself MUST be documented if it is to be maintained!
360 There are many ways to do this, and appropriate use of PASCAL or Forth (yes,
361 Forth!) can be self-documenting to a great degree. But if the language
362 (or implementation) doesn't allow for easy self documenting code, then you
363 MUST add comments to the extent necessary that a competent programmer in
364 the language/system you are using can figure out what's going on -- and you
365 must use the "community standard" definition of competent, not some idealized
366 definition you might have.
367 As to the problem with spaceless BASIC programs with cryptic variable names,
368 the appropriate way to deal with this was implied in your message: To wit,
369 write the original code using spaces (TABs even!) and long, meaningful
370 variable names, and include adequate comments. Then run this program source
371 through a utility which truncates variable names, and throws out comments
372 and spaces. Give yourself the best of both worlds.
373 ---------------------------------------- Pann 1/16/83 21:30 ------------
374
375
376
377
378 +++{
379 **************************************
380 FOR SALE---CENTRONICS PRINTER---
381
382 MODEL 737-1 PARALLEL INPUT
383
384 80 COL/132 COL/PROPORTIONAL
385
386 LIKE NEW CONDITION W/MANUALS
387
388 PAID $800 IN 1981
389
390 WILL SELL FOR $350.00
391
392 CONTACT: GLENN AT
393 xxx-xxxx EXT. xx DAYS
394 xxx-xxxx EVES. AFTER 6:00 PM
395
396 **************************************
397 I would have to disagree with you that small computers will not see any
398 great improvement in speed. Accually they are currently horibly slow
399 The reason for this is that they use the next to slowest type of technology
400 to build them (NMOS). Just going to TTL could boost the speed factor by
401 5 to 10, and going to ECL can boost it to that 50 to 100 mark. Of course
402 we haven't even touched Johsephine junctions (and please don't!), accually
403 I have my doubts as to the viability of this particular technology as it is
404 rather difficult to work with and expensive. Other technolgies in the wings
405 could easily come close to doing the same thing without all that cyrogenics!
406 ECL and GASP both aproach this speed. Another aproach is to build a base
407 machine that is more efficeint. Going to a wider buss and reducing the state
408 cycles that the machine goes through for each instruction can easily boost
409 the speed by a factor of 10. Of course with all this super speed we do
410 have a problem, and that is feeding our poor hungery beast fast enough.
411 some of this can be handled with newer and faster memories, and lookahead
412 addressing, but I think that you will find that the memory access and I/O
413 access will always be the limiting factor in computers. After all, you got
414 to get the data to begin with in order to crunch it. and a very large amount
415 of computer work these days is just moveing data from one place to another
416 with a little massaging in the middle.
417 I do however agree with you that a little improvement in programing speed
418 can do wounders too, but I do at times wonder what all the fuss is about.
419 generally I write in BASIC because it is easy not because it is fast. If
420 I need fast I run the program thru a complier and to heck with the spaces!
421 Or if I realy need fast I convert it to assembler which sometimes improves
422 the speed (But not always) which is not that difficult once you have the
423 basic structures layed down. Usally the speed is picked up because the
424 code is written differently in assembler as you have full control of
425 the machine there are many things you can do that can't be done in basic
426 because it doesn't know how to allow them (thus the often used USR fuction
427 is born.)
428 Although writting two sets of code sounds nice on the surface, I find it
429 extremly hard to believe that anyone would accualy do it. There usualy is
430 barely enough reason to write the original code to begin with! And expecting
431 a hobiest to do it is about as usfull as praying for snow in the Sahara.
432 *** THE EVER RANTING CISTOP MIKEY *** 17 JAN 83 ************************
433 IT WOULD APPEAR THAT 'CISTOP MIKEY' REEKS OF BOVINE FECAL MATTER !!!
434 IF YOU PERSIST IN FLASHING BUZZWORDS AT LEAST USE THEM IN THE
435 PROPER CONTEXT. ALSO A DICTIONARY COULD GREATLY IMPROVE YOUR SPELLING.
436
437 *****POOF!*****
438 MESSAGE
439
440 MESSAGE
441
442 AN EXPERIENCED VIDEO TECHNICIAN IS NEEDED BY THE CITY OF PORTLAND...
443 MESSAGD
444 TEE, SEE MSG ON BIT BUCKET, LOST UR PHONE NO. PLSE CALL BACK OR LEAVE IT ON
445 BIT BUCKET - TOM xxx-xxxx
446 -----------------------------------
447 Why are personal computers slow: Part 1.
448 A bit (pardon the pun) of History. Floppy disks weren't
449 invented as an on-line storage medium. IBM made them as a
450 small, easily transported medium of data exchange, especially where
451 an entire operating system for a big machine was to be reloaded
452 by a field service person. Another application was key-to-disk, a
453 replacement for keypunching. The floppies weren't ever supposed
454 to be used as direct disk devices. Unfortunately (or fortunately,
455 depending on whether you think the glass is half full or empty)
456 we're relatively stuck with them.
457 Question: Does the average person NEED processing capability
458 significantly more powerful than the current personal computer???
459 Let's hear your ideas on the subject.
460 ----------------------------------------
461 %@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%@%@
462 HEY SYSOPR !!!! HOW CAN I GET TO BOTTOM 25 LINES W/_O READING ALL THE DIATRIBE
463 EVERY TIME ?? BOTTOM COMMAND DOESNT
464 SEEM TO DO WHAT I WANT....
465 MIKE, YOU ARE RIGHT: WE CAN GET FASTER MACHINES, BUT YOU HAVE
466 JUST EDSCRIBED THE REASON LARGER MINIS AN MAINFRAMES ARE FASTER,
467 AND ALSO WHY THE COST $100,000 AND UP, AND WHY MICROS ARE $100
468 AND UP (ALSO DOWN) S TO THE BIT? ABOUT FLOPPIES, NEITHER
469 WAS THE MICRO INVENTED AS A COMPUTER, BUT AS LOGIC RELPLACEMENT
470 FOR A CALCULATOR OR TRMINAL. E THANKFUL FOR THINGS INVENTED
471 FOR OTHER THINGS: THAT' ALL WE'VE GOT. ND THERY ARE BOTH
472 BETTER THAN WHAT THEY REPLACED. UNTIL TOMMORROW, AT LEAST.
473 ............................B.A.D.....1-19-83...20:36
474
475 GEEZ! Come on guys! You can type better than that. Mr. Day, your diatribe
476 would be much more effective if you would read it after typing it.
477 And Mr. D., is your machine dropping charactors again? An error every once
478 in a while is OK, but the rate has been increasing rapidly of late.
479 Please excuse the typos,
480 the Phantom Glitch.
481 (((((((((((((((((((((((((((((((((((((((((((((((((01/19/83))))))))))))))))))
482 No, the average person certainly doesn't need more processing power than
483 is currently available from various "personal computers." But then again,
484 so what?
485 ................................................. pann 1/19/83 21:20.........
486
487 THIS HAS CERTAINLY BEEN INTERESTING. YES, MORE DICTIONARIES, NOT MORE SPEED
488 WILL SERVE THE PERSONAL COMPUTER USER MOST EFFECTIVELY.
489 EPS/MMP 1/19/83 2207
490 =======================
491 About a year ago I read an article about direct digital transfer
492 through the phone lines. In effect, the phone system will bypass the
493 A/D step, and the subsequent D/A step, and act like a 56K baud modem.
494 Whaddaya think, how's this gonna change the world. Will modem
495 manufacturers go the way of the buggywhip manufacturers?
496 :::::::::::::::::::::::::::
497 READ WHAT I WRITE?!!! HEVEN FORBID THAT SUCH A THING WOULD COME TO PASS!!
498 WHY I MIGHT FIND OUT WHAT I WAS TRYING TO SAY AND I JUST NEVER COULD
499 LIVE WITH THAT!! AS FOR THE DICTIONARY, WELL, MAYBE ONE OF THESE DAYS
500 I'LL INVEST IN ONE, BUT I REALLY DON'T THINK IT WILL HELP SINCE I CAN'T
501 READ. AND IT IS OBVIOUS FROM MY RAVINGS THAT I CAN'T WRITE. THE REST, WELL..
502 **** THE NON-FLAMBOYANT CISTOP MIKEY *** 19 JAN 83 *****************
503 P.S. TRY TO FIND SAHARA IN THE DICTIONARY (WOULD YOU BELIEVE IT IS NOT
504 IN THE WEBSTERS UNABRIDGED!)
505 ...................................................
506 If you want to read the bottom 25 lines, just type 'BOTTOM', then 'UP 25',
507 and 'PR 25'. Nothing to it.
508 **************************************************************************
509 With regard to "direct digital transfer": the terminology is somewhat
510 misleading. Ma Bell and other suppliers have, of course been digitizing
511 your voice (and modem tones) for many years now when going between central
512 offices. That doesn't mean that there isn't a significant amount of
513 massaging of the "digital" information prior to placing it on a twisted
514 pair. This is necessary to allow for the attenuation and phase charact-
515 eristic of the medium and to prevent undesirable crosstalk.
516 The point is, you're not going to get rid of the modem (or some other
517 piece of conversion hardware) because it is still an analog world out
518 there, despite all the flap about the information age megatrend. There
519 are any number of high speed limited distance modems available, but
520 all of them that I've seen require a DC circuit between the modems,
521 a commodity you will be increasingly hard pressed to find these days,
522 let alone in the future (unless you will be willing to pay dearly!!)
523 {}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{} Analog Alan {}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}
524 In regard to some of the above comments:
525 1. At work we have several pieces of automated equipment worth
526 several tens of thousands of dollars. They do a number of
527 testing and sorting steps in a matter of seconds that would
528 take someone with regular measuring gear several MINUTES to do.
529 They operate fast enough that the operations can be considered
530 simultaneous. What miracle of modern technology controls these
531 monsters? Why, its a "lowly" 4004! Like the man said, it's not
532 obselete, it has fewer uses.
533 2. Enough of arguing speed, on micros we haven't reached the
534 "ultimate" limit (speed of light lag) yet, but a simple
535 calculation shows that it requires about 2/3 of a nanosecond
536 for a signal to travel (1-way) a 20 cm path. Since the signals
537 are actually travelling at less than light-speed (I'm NOT
538 kidding!), sometimes much less, there is an inherent limit
539 in the speed of MOST systems that is much closer than some
540 of you think! I get a factor of about 1 Ghz as a "maximum"
541 clock speed. I doubt that S-100 and similar systems can get
542 that high. Also as the clock speed goes up circuit design is
543 going to get WEIRD! Ask any radio tech about what kind of weird-
544 ness occurs above 100 Mhz!
545 I'm sure there are some errors in specific details of the
546 above, but I'm just as sure that the general principles hold.
547 ......................................................................
548 So what about speed! Do it right the first time and then speed doesn't
549 matter....enter the dictionary. Besides, even if it took twice as long,
550 I bet that you'd still get a kick out of it! You call this work!
551 EPS 1/20/83 20:00
552
553 ////////////////////////////////////////////////////////////////////
554 YES INDEEDY THE PRINCIPLES TO HOLD, IT DOES NOT MEAN THAT A COMPUTER CAN
555 NOT OPERATE FASTER THEN 1GHZ ONLY THAT IT CAN'T USE EXISTING TECHNICS TO
556 DO IT. THERE HAS BEEN SOME FOOLING AROUND WITH A OPTICAL COMPUTER WHICH
557 OF COURSE BY ITS NATURE OPERATES AT LIGHT SPEED, BUT I EXPECT THAT OTHER
558 THEN SOME VERY SPECIAL APPLICATIONS THIS WILL NOT SEE MUCH USE IN COMPUTERS
559 AS WE KNOW AND LOVE THEM. I DON'T EXPECT TO SEE COMMON COMPUTERS RUNNING
560 MUCH ABOVE 20-50 MHZ BECAUSE OF THE PROBLEMS WITH CONTROLING SIGNALS AT
561 THOSE SPEEDS. BUT THERE WILL DEFINATLY BE COMPUTERS RUNNING FASTER THEN
562 100MHZ THESE WILL BE THE CRAYS AND CYBERS OF THE FUTURE. WHAT YOU WILL
563 SEE MORE OF IN THE FUTURE AS FAR AS LOW LIFE COMPUTERS IS A WHOLE LOT
564 MORE PARALLEL COMPUTING AND DISTRIBUTED COMPUTING. IT IS INTERESTING
565 THAT EVERYONE KEEPS SEEING THE FUTURE WITH A GREAT BIG CENTRAL COMPUTER.
566 YET THIS DEFEATS ITS SELF. THE PROBLEM HERE IS THAT IF THE BIG HONKER
567 GOES DOWN THE WHOLE SYSTEM DIES. WHAT IS REALLY NEEDED IS LOTS OF LITTLE
568 SYSTEMS THAT SHARE THEIR DATA WITH EACH OTHER AND SPECIALIZED CENTRAL
569 UNITS THAT PERFORM CERTAIN JOBS WITH ALTERNATES IN CASE OF FAILURE.
570 THIS IS THE WAY LIFE WORKS, AND IT IS THE WAY OUR COMPUTER SYSTEMS
571 WILL HAVE TO WORK IF THEY ARE TO SURVIVE.
572 The speed of light is important because the address needs to go out, and then
573 the data returns, and this is the limit in addressing memory; however, this
574 is not the final limit. Oneway processors, called pipelined architecture, is
575 not unusual in systems designed for speed, the program and data enter at one
576 end, the prodessor massages them, and spits them out the other end. All
577 current micro's (and most mainframes) only do one thing at a time, so the
578 delay in any stage (usually memory) slows them down. If you call for a piece
579 of data, and use it immediately you have to wait until it is available, but
580 if you can call for it and then do something else until it becomes available,
581 you can run faster than memory. As yet, this is only used for very slow
582 stuff like discs, but we aren't designing much for speed yet. A state
583 machine processor like the 8080 could be implemented so each clock cycle
584 a new instruction is read from memory, and the prior one is moved to the next
585 layer in the processor. The time to execute an instruction does not change,
586 but the average time does, because several are executing at the same time.
587 Some early computers did this, because they were inherently so slow, such
588 as the Bell labs #1 ESS that connected you with BWMS. It needs all the
589 funny speedups because it's memory access time is 5 or 6 microseconds.
590 With 3 instructions running at the same time, though it handles about 30k
591 time share terminals (ie telephones). Personal computers of the 16 bit sort
592 are about as powerful at arithmetic, but would trip over eth special jobs
593 involved, like scanning all those lines to see who needs dial tone.
594 As to errors, I can't type, and when I use that !@#$%^ silent 700 through
595 the computer at work to access, things get lost. Unlike the sysop, I can
596 read, but don't like to (never my own).
597 W e do need more computer power in home computers, but not speed, mostly
598 more memory.
599 ..................................B.A.D............21:40/jan 21 83............
600 AMEN *** CISTOP MIKEY *** 22 JAN 83 ***********
601 It's funny that the entire tone of the conversation on here changes so
602 drastically from one disk to the next. It must be associated with the
603 orientation of the magnetic domains on the disk when it is made.
604 |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
605 I've always wondered about those pipelined machines...if they have several
606 instructions simultaneously executing, what happens when one of them is a
607 conditional branch??? Do they just throw out the following steps and
608 hiccup for a while?
609 ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
610 YES, THE PIPELINED PROCESSORS ARE AT A DISADVANTAGE ON CONDITIONAL
611 JUMPS;THEY USUALLY DO JUST THROW AWAY THE NEST INSTRUCTIONS, HOWEVER,
612 ONLY THE INSTRUCTIONS IN THE INPUT QUEUE ARE DISCARDED, AS THE JUMP IS
613 TESTED EARLY. JUMPS TAKE LONGER THAN SAY REGISTER SWAPS, SO MAY USE MORE
614 TIME AND SEEM LIKE SEVERAL INSTRUCTIONS. E.G. 8080 4 CLOCKS VS 12 TO 17.
615 THE FULL DIGITAL ACCESS TO THE PHONE LINES EXISTS, DIRECT FROM
616 THE SPECIAL ADAPTER ON THE SUB('S PREMESSES. IT IS CALLED (FROM
617 MA BELL) DIGITAL DATA SYSTEM, OR DDS,; UNFORTUNATELY THE FCC AND
618 PUC WILL NOT ALLOW BELL (YET) TO OFFER THIS AS A SWITCHED SERVICE,
619 AND IT IS AVAILABLE ONLY POINT TO POINT. SOME PBX MAKERS LIKE
620 NORTH ELECTRC AND ROHM DIGITIZE THE VOICE AT THE PHONE, AND
621 SEND BINARY DATA OVER THE LINE: ANY DIGITAL DEVICE THAT CAN MATCH
622 THEIR DATA FORMAT CAN ALSO WORK AS PURE DATA. THE PRIMARY REASON
623 IT IS NOT TO BE SEEN ELSEWHERE IS INERTIA..
624 ......................................B.A.D................
625