“I can imagine people having as many as twenty or thirty windows ready

to call up with notes or working papers,” says John Butler, a product

manager with Microsoft.

Also, windows software may let you switch noticeably faster from one

program to another. And with a RAM (temporary computer memory) above

500K, you may even be able to do so instantly.

The plus of this? You won’t need to return as often to your computer’s

operating system and feed the programs one by one into the RAM.

So the machine may seem to impose itself less between you and your work.

When “windows shopping,” however, you should ask these questions and

more:[105]

1. Is the convenience worth the extra several hundred dollars you’ll be

paying for the window software alone if it doesn’t come with your

computer?

2. Up to how many windows can you see on your screen at once?

3. How do the windows look alongside each other? Do they =overlap=, just

like papers atop each other? Or do they =tile=? That is, if you select

more windows, the existing ones shrink, and you view less from them.

4. How about =data transfer=? If you move information from one

electronic file to another, will important details remain? For

instance, the way rows and columns coexist with each other on a

spreadsheet? Can you cut a graph out of a chart and insert it in the

text of a report without it shrinking or ending up distorted? Will it

reproduce in as much detail as it would without your reaching it

through a window?

5. What kind of graphics—=bit mapped= or =character based=? The bit

mapping means sharper images. Your computer keeps track of each little

dot, each pixel, on the screen. That hogs memory and may rule out

color. =Character-based= systems, though, don’t let you make your

lines and curves as smooth as bit mapping does. They must work with

already-shaped letters, numbers, and other visual forms.

6. Will the window program work with ordinary software or just products

written for it? And how many of the windows’ special features will do

work when you use regular programs?

7. Will the windows at least slightly slow down some programs? A word

processor may take longer for you to get from one part of your report

to another.

8. Is the program picky about the computers it’ll work with? A window

system _may_ need over half a million bytes of RAM and a hard-disk

drive storing 5 megabytes. Also, as of this writing, windows seemed

geared more to the IBM-style MS-DOS computers than to the older but

cheaper CP/M ones. Besides, some companies may not sell windows

programs directly to ordinary buyers. Microsoft got various micro

manufacturers to bundle the program with their products.

9. Does the program require a mouse—the gadget you roll on your disk to

move the cursor? How easily can you control the program without a

mouse?

Footnote 105:

I’m basing the “windows-shopping” questions on points raised by

various micro publications, especially _Popular Computing_, May 1984,

p. 96.

BACKUP X ❑ Of Mice and Men—and Touch Pads, Touch Screens, Etc.[106]

“If you’re a trained, high-volume production typist,” asked Seymour

Rubinstein, the WordStar developer, “what are you going to do with a

mouse except feed it cheese?”[107]

Footnote 106:

For background in researching this appendix—but not for the opinions,

strictly mine—I’m grateful to Forrest M. Mims’s excellent article “A

Few Quick Pointers.” It appeared in May 1984 _Computers and

Electronics_.

Footnote 107:

Rubinstein’s “cheese” quote is from _InfoWorld_, May 14, 1984, p. 57.

Score one for Rubinstein. He says mice are great—if you have three

hands.

Doing graphics? A mouse, maybe. But damned if I’m going to take my hands

off the keyboard to push the cursor from one spot on the screen to the

next.

It’s simply too much wasted motion. I instead just press the cursor keys

right above the main keyboard. Or I use WordStar’s cursor-moving

commands. And even if I hadn’t learned touch typing a quarter century

ago, I’d still wonder if a mouse for word processing wasn’t the Silicon

Valley version of _The Emperor’s New Clothes_. Next time you’re in

California, maybe you’ll see Apple execs naked in the streets as well as

their hot tubs. Well, maybe not. The mouse could be a great marketing

tool for sales reps peddling Macs or Apple IIc’s to people hoping to do

word processing. But experienced typists? Many would probably groan over

all the excursions that the mouse forced them to take from the main

keyboard.

Some enemies of mice—cats?—also wonder if jockeying the cursor this way

couldn’t be a little tiring for people writing or typing. Think of the

hand-eye coordination required. You’re rolling a palm-sized gizmo on

your desk to position the cursor on a single letter at times, and that

might not wear too well if you‘re working for hours on end.

Mind you, the rodents have their friends, especially at Apple, where,

inconsiderately, the hardware wizards didn’t even favor the Macintosh

with cursor keys.

Joe Shelton, the Apple products manager mentioned in Chapter 7, says he

does most of his writing with a mouse. He suggests the mouse-equipped

Macintosh for “naïve users.” Your term, Joe. Now I’ll slightly water

down my jeremiads. Macintosh-style computers offer nifty graphics and

nice offbeat typefaces. So some trendy writers may want one—rodent or

not.

And if you’re an executive or someone else not doing heavy word

processing? Then maybe, just maybe, a mouse is for you. Perhaps you’re

working with spreadsheets, a number of programs in fact, and you write

only a small fraction of the time.

Richard Webb—a partner at Peat, Marwick and Mitchell, the big accounting

firm that advised Apple during development of the Macintosh and ordered

thousands of them—swears by the mouse for spreadsheets. He says that

alone would justify the mouse.

I’m not a spreadsheet artist and will take a pro’s word. But at least

for heavy-duty writing and typing, the old cursor keys are my best bet.

Graphics is different. There, the cursor keys are more cumbersome.

One artist, however, wanted not a mouse but an electronic “pad and a

stylus”; he might be happy with a =digitizing tablet=—also known as a

=graphics tablet=—like the well-known KoalaPad. You can write on this

surface with a stylus or your finger and the computer will display the

lines on its screen.

A =touch-sensitive screen= is still another possibility—for some

people—both in word processing and graphics. You point your finger at a

spot on the screen. Bingo! You can start moving a paragraph or perform

graphics magic. But touch typists may face the same problem as with the

mouse—wasted motion—and some people may tire of reaching up to the

screen again and again. Also, touch-sensitive screens may not be precise

enough for you to pick out just one number or letter.

Hugh Hunt raised an interesting issue.

“What happens if a fly lands on the Hewlett-Packard screen?” he asked

someone about a computer with a touch screen.

Well, I hear, the HP 150’s screen uses infrared touch sensors that are

more than fly length from the glass. “Debugging” the 150,

Hewlett-Packard must have thought of _everything_.

Yet other pointing devices are:

● The =joystick=. Moving the stick around, you move the cursor. A neat

idea. But it’s more fit for video games than word processing and many

other business programs—you just can’t point exactly.

● The =trackball=. You move the cursor by rotating a ball inside an

enclosure. Want the cursor to go faster? Then rotate the ball more

energetically. The trackball is found most often in arcade-style

systems. It’s great for chasing aliens and may have uses in

spreadsheets and data-base management, but some people say it’s an

abomination for word processing.

● The =light pen=, with which you could electronically “draw” on the

screen. _Draw?_ Okay. _Write?_ Well, it’s “wasted motion” time again

here, as with the other items on this list.

As computers climb the executive ladders—as more nontypists use

them—these alternatives may grow in importance. And what about people on

the factory floor? They may use such gizmos to make new inventory

entries or machine adjustments. Partly the world is learning to type,

and partly the computer is learning to understand devices other than the

keyboard.

What cursor-control gizmos are ahead? I’ll keep reading the _National

Enquirer_ articles about people moving objects with their minds. Maybe,

the hard-core hackers are thinking, the next gimmick will feature some

user-friendly ESP.

BACKUP XI ❑ The Micro Connection: Some Critical Explanations

When John Fuller hooked up his Heath micro over the phone lines to the

Wang at his office, he had a head start on many other do-it-yourself

telecommuters.

He wasn’t a computer or electronics rookie. For years he had worked

around big IBM’s and Honeywells, and while telecommuting, he was a

computer and management consultant with the navy. He even soldered

together his H-89—a kit. And, anyway, he was communicating with another

smaller computer, a Wang word processor.

Does that mean you should give up on do-it-yourself telecommuting

because of the technical complexities?

No. Not at all. Before hiring a consultant, you might consider two

possible sources of free advice: (1) your corporate data-processing

department, if you have one, and (2) users groups, whose phone numbers

you can get through dealers.

Before approaching anyone, though, learn the basics of the technology. A

=modem= converts the _0_s from your computer into one pitch and the _1_s

into another, and that whine, you’ll recall, goes out over the phone

lines. The “mo” in “modem” stands for “modulator”; the “dem,” for

“demodulator.” And a “dem” springs into action at the other

end—demodulating the whine back into the _1_s and _0_s.

Modems come in two varieties, =direct connect= and =acoustically

coupled=.

The direct-connect modem hooks between the phone line and your phone, or

between your phone and the handset. Its whine transmits better over

static-ridden lines than do acoustic couplers, but it’s normally a

little more expensive, and hard to install if your phone lacks =modular

plugs=, the tiny plastic ones that unsnap.

Using an acoustic coupler, you cradle your handset in rubber “earmuffs.”

A small speaker whines into the phone transmitter; a little microphone

picks up the sound from the handset’s receiver. Acoustic couplers don’t

need phones with modular plugs, obviously—a boon to traveling

executives—but they might not work if the handset is oddly shaped, as it

is with some of the new-style phones. Also, loud noises in the room can

confuse them.

A direct-connect modem, for that reason and others, would probably be

better for your regular office. Also, consider a =full-duplex= modem in

most cases.

No, it isn’t a modem for two-family houses. “Full duplex” is just jargon

for computers jabbering back and forth, both ways, at once. Like two

people on the phone. A =half-duplex= modem would allow just one-way

communications without switching; it’s somewhat comparable to a radio

with an unwieldy send-receive control. There are times, however, when

half duplex would be best. Normal phone lines, =voice grade= ones in

telephone company language, can’t handle computers jabbering at too high

a speed, and half duplex may work better than full.

Like most modems for personal computers in the early 1980s, Fuller’s

gizmo transmitted information at 300 baud—equivalent to 300 bits a

second, 30 characters per second, or 360 words per minute at 5 letters a

word. When telecommuting, however, you may be better off with a modem

going 1,200 baud. After all, it can zip material over the phone lines at

four times the speed, and that means you’re tying up your modem less on

long documents. You might not be tying up _yourself_ if you run an

operating system like Concurrent CP/M, which lets your computer handle

electronic mail while you’re using it for other purposes; but a

1,200-baud modem is still a good investment if you can afford it. You

may well be able to. Soon 1,200-baud modems with auto-answer features

will commonly sell for well under $300.

Sent through a modem, the =data bits= are those conveying the

information itself. And the =start bits= and =stop bits= keep track of

the beginnings and ends of your micros’ characters. The most common

setting is a =word length= of 8 data bits and 1 stop bit.

=Parity bits= may help the computers check for errors caused by static

and other electronic noise. Parity is said to be “even,” “odd,” or

“none.” This =error-checking= method has its drawbacks. It may fail if

the mistakes cancel each other out—and it’s slow since it checks every

character. Some =communications programs= (the ones allowing computers

to talk) lack this feature. The most common parity setting is

“none”—when you tell your software not to do parity checks.

A few other wrinkles in such programs are:

FILE TRANSFER

That’s the ability to send whole files of text or data—including

programs, too.

CAPTURE ABILITY

With =capture ability= you can keep a record of each end of a keyboard

conversation with both people typing away. You tell your computer to

open up a =buffer= in its RAM—the temporary memory. The buffer captures

the conversation. And if it fills up, some software will automatically

“write” to your floppy disk, then reopen the buffer. Some programs won’t

write to the disk at the ends of your conversations unless you command

them to, meaning that the absentminded will see their bits and bytes

vanish.

PROTOCOL FLEXIBILITY

Beyond changing bit numbers, etc., you may want software that can handle

different kinds of protocols—sets of rules telling how computers

transmit different sets of information.

One possible protocol for you to use—at least in conversations with

other micros—would be the =XMODEM Protocol= or =Ward Christensen

Protocol=. The receiving computer tallies up the number of bits and

bytes transmitted from =blocks=—tiny parts of material—and compares this

=check sum= to the number that the originating machine says it sent. And

if there’s a mismatch? The receiving computer asks for the block again.

You might also use no protocol at all. It’s as if you’re sending

material simply by tapping away on the keyboard without any elaborate

rules for the other person’s machine. You two should still, however,

agree on basics like baud rate, data bits, stop bits, and parity.

Protocols use =handshaking= as part of their bags of tricks.

Handshaking, computers swap bits and bytes to verify that they’re on

speaking terms. A common form is =XON/XOFF=.

Some mainframes use XON/XOFF to talk to micros and vice versa. The

receiving computer gives an XOFF signal, telling the transmitting one to

pause while the receiving machine “writes” to its disk. XON means,

“Okay, my buffer’s almost empty. Send more!” Without this =flow

control=, you might overwhelm the buffer and lose information.

AUTOMATIC DIALING AND OTHER TRIMMINGS

Some modems will let the communications software dial for you or at

least let you feed the number in from your keyboard. Some also have

=automatic log-on=. That lets you check onto a network or bulletin-board

system without typing out your ID or password. Many =intelligent

modems=, by the way, even without special software, can remember and

dial phone numbers automatically and spit out log-on sequences and

passwords.

VERSATILITY

Want to speak to a number of computers? Then buy software allowing you

to set:

◼ Baud rate (speed of transmission).

◼ Data bits.

◼ Stop bits.

◼ Parity.

◼ Terminal emulation. It means making the micro imitate popular brands

of terminals.

◼ Half or full duplex. Full is most common in micro communications.

◼ The type of handshaking protocol.

Ideally, the software will also:

◼ Accommodate a variety of modem types.

◼ Designate the host—which computer will manage the =echo back= (which

makes the typed conversation appear on both screens).

◼ Let you choose different protocols.

All those features, of course, might still not let you communicate

easily with your corporate computer. It may not use the =ASCII= code,

for instance, which most micros do.

“Since you ASCII’d, anyway,” as _PC Magazine_ once punned, “those five

letters are an acronym for the American Standard Code for Information

Interchange.” With ASCII, an “A” is one combination of _1_s and _0_s,

“B” is another, “,” is another, and so on.

When communicating with a mainframe capable of ASCII, it may not matter

what brand of micro you’re running. So don’t reflexively think Big Blue

for telecommuting.

The proud IBM mainframes, however, like aloof mandarins of old China,

jabber away in a dialect of their own, =EBCDIC=—it’s similar to ASCII

but different enough to cause serious problems. Now that doesn’t mean

your home computer won’t work with an IBM mainframe. But it may require

special programming that will translate from one set of codes to

another.

Of course your corporate mainframe may also need a string of =control

characters= to open or close an electronic file, and here your

word-processing software may matter more than your communications

program.

The control characters normally are just letters tapped out in

combination with the control key to give commands to your computer. They

generally don’t print out on paper; they may or may not show on the

screen. In this case, controls characters pass the commands on to

another machine over the wires. And so you may have to =embed= the

control characters in your electronic files. If your word processor

won’t let you embed control characters, you may not be able to

communicate with some kinds of mainframes. WordStar lets you do this.

An issue more basic than control characters is whether you’re using

=synchronous= or =asynchronous communications=. The former is a staple

among the big mainframes; the latter, among micros.

In synchronous communications, the two computers or modems employ timing

signals to separate characters sent in one stream.

Synchronous has a big advantage. You can send at 4,800 baud over regular

phone lines; you can’t with the asynchronous method, or at least not

with late 1984 technology.

But “synch” is more expensive—the needed modems typically cost well over

$1,000. Asynchronous, moreover, doesn’t require timing signals. It’s the

simplest form of communications between computers; and more and more

data-processing departments have added asynchronous ports for micros to

dial up.

John Fuller himself was communicating not with a giant computer but a

dedicated word processor in his office, which, like most, talked ASCII

asynchronously. He solved other problems on his own. You may need a

consultant to do what Fuller did. But take heart. It could well be worth

it, what with the time and money your telecommuting may spare you and

your employer.

“I’m saving gas, dressing in comfortable grubbies when I write,” Fuller

said in an article about his experiences, “and with the direct

connection I can _prove_ that I’m working. I turn out my reports faster.

My computer is really paying off. Now,” said the self-deprecating

Fuller, “if you’ll excuse me, I have to repel invaders from space.”

BACKUP XII ❑ MODEM7: An Almost-Free and Fairly Easy Way to Talk to Other

Computers

MODEM7 may be free, but it _works_. With it you can talk to computers

thousands of miles away _if_ they’re running MODEM7 or compatible

software. Normally, your only real expense is for the modem and maybe

cables; you might spend as little as $100.

Using MODEM7, you can send letters, reports, or programs already on your

computer disk. Or you can receive them.

You can also:

1. Communicate teletype-fashion with the other person. You can keep

electronic records of what you both type and later print them out

using WordStar or another word processor.

2. Call up electronic bulletin-board systems (BBSs) or plug into The

Source and many other information utilities. In this case, too, you

can store material on your disk and later print it out.

3. Get copies of other programs that altruistic computer buffs have

placed in the public domain; you do this after you log on the BBS.

MODEM7 is a good example of free software that fills a niche; experts

say it’s better than some $150 products. Most free programs are losers

for the typical business. They’re either (1) games, useless except for

training, (2) unreliable lemons with bugs in them, the software kind,

(3) programs that might work but don’t do anything useful, or (4)

fiendishly difficult challenges because the instructions are confusing

and no one handy knows how to use them. But wonderful exceptions exist.

Never thumb your nose at something just because it’s free. In MODEM7’s

case you’ll find more happy, knowledgeable users than you will in the

cases of many commercial programs.

MODEM7-style protocols have become an industry standard. Even big

businesses and government agencies, including the IRS, have used MODEM7

for some purposes. It’s influenced the designs of commercial micro

programs, which employ similar techniques to facilitate linkups between

different brands of communications software. MODEM7 will also get along

beautifully with the popular PC-TALK III “freeware” program for IBM

micros and some clones. Freeware is low-cost software distributed in

many cases by the users themselves; the author may request a “donation”.

Granted, MODEM7 isn’t the ultimate. The instructions included on the

disk are confusing to the novice—a problem I’ll try to overcome here—and

ready-to-run versions aren’t available for every computer system.

Some commercial communications software, moreover, is easier to run and

can help you talk to more machines. And it may also make better use of

the “smart” modems and their software so that your computer can dial up

others automatically at 11:00 p.m., when the phone rates go down. You

can’t do this with your cheapie modem and the plain vanilla MODEM7

described here.

But for the average computer user, MODEM7 is a good communications

program to start out with. After reading the instructions that follow,

you needn’t be a micro whiz to use it.

Also, you _may_ find that MODEM7 is as convenient as some commercial

software for simple tasks, such as quickly checking your electronic

mailboxes on The Source or CompuServe.

Don’t forget: you shouldn’t pay a cent for the basic MODEM7 program

itself except for the cost of a copied disk or another token fee. It’s

public-domain software. So you can legally make copies for all your

friends and business contacts who have modems and the same machine that

you do.

But how to track down your own copy? Through CP/M user groups, you can

locate versions of MODEM7 set up for different computers and disk

drives, or you might find an organization of IBM owners, Kaypro

fanatics, or others. The better computer stores can steer you to the

right people. Incidentally, if you distrust your store, you might not

want to say exactly why you’re trying to catch up with a user group.

Bear in mind that the computer store isn’t going to make a profit off a

free program. Some organizations, however, are selling MODEM7 at next to

nothing through the mail. For instance, the Public Domain Software

Copying Company says it has versions for the IBM PC and clones, all the

Kaypro machines, the Morrow micros and Osborne, among others. The cost

as of late 1984 was $10. The address is 33 Gold St. cl3, New York, N.Y.

10038; the telephone number, (212) 732-2565. The company also offers

separate programs that will adjust stop bits and other settings,

eliminating the need for programming. These programs aren’t universal.

Like MODEM7 itself, they might not work for your particular machine. The

Public Domain Software Copying Company also says it offers a low-cost

program for the Apple II that lets it communicate with computers using

MODEM7.

The Silicon Jungle · 第 33 / 36 段