1
00:00:00,000 --> 00:00:10,956


2
00:00:10,956 --> 00:00:14,570
This is the Computation
Center at MIT,

3
00:00:14,570 --> 00:00:16,270
which serves more
than 40 New England

4
00:00:16,270 --> 00:00:18,250
colleges and universities.

5
00:00:18,250 --> 00:00:22,150
The heart of the center is
this IBM 7090 high speed

6
00:00:22,150 --> 00:00:24,490
general-purpose
digital computer.

7
00:00:24,490 --> 00:00:27,400
In addition to the routine
work which takes place here,

8
00:00:27,400 --> 00:00:29,800
experimentation is
underway to develop

9
00:00:29,800 --> 00:00:33,430
a new technique of computer
usage called time-sharing.

10
00:00:33,430 --> 00:00:36,610
At consoles like this one,
located in laboratories

11
00:00:36,610 --> 00:00:39,490
and offices throughout New
England, hundreds of people

12
00:00:39,490 --> 00:00:42,250
will one day be able to
use tomorrow's computer

13
00:00:42,250 --> 00:00:43,840
simultaneously.

14
00:00:43,840 --> 00:00:45,340
This man has been
using the computer

15
00:00:45,340 --> 00:00:48,520
to solve a relatively
simple mathematical problem

16
00:00:48,520 --> 00:00:50,800
and while he thinks about
what he wants to do next,

17
00:00:50,800 --> 00:00:53,830
the computer is busy working
on someone else's problem.

18
00:00:53,830 --> 00:00:55,690
In this way scientists
hope to break

19
00:00:55,690 --> 00:00:58,420
one of the major bottlenecks
in the use of computers.

20
00:00:58,420 --> 00:01:01,378


21
00:01:01,378 --> 00:01:04,829
[MUSIC PLAYS]

22
00:01:04,829 --> 00:01:28,020


23
00:01:28,020 --> 00:01:31,020
Hello I'm John Fitch,
MIT science reporter.

24
00:01:31,020 --> 00:01:32,910
I guess most of us
are aware of the fact

25
00:01:32,910 --> 00:01:35,670
that the computer has come to
play an increasingly important

26
00:01:35,670 --> 00:01:37,830
role in our everyday lives.

27
00:01:37,830 --> 00:01:41,700
It processes the checks that we
write, it handles our payrolls,

28
00:01:41,700 --> 00:01:43,620
and it even keeps track
of the availability

29
00:01:43,620 --> 00:01:45,390
of airline reservations.

30
00:01:45,390 --> 00:01:48,660
But in addition to solving our
problems which revolve around

31
00:01:48,660 --> 00:01:52,020
the relationship between
the man and his machine.

32
00:01:52,020 --> 00:01:53,820
I'd like to have you
meet Dr. Fernando J.

33
00:01:53,820 --> 00:01:57,120
Corbato who is the associate
director of this center.

34
00:01:57,120 --> 00:02:00,150
Dr. Corbato, maybe you can tell
us about some of these problems

35
00:02:00,150 --> 00:02:02,850
to put this whole business
in perspective for us.

36
00:02:02,850 --> 00:02:05,700
Well, to do that, we have to
go back about a decade, about

37
00:02:05,700 --> 00:02:07,020
the early 1950s.

38
00:02:07,020 --> 00:02:10,410
And the real problem was to
try and put together computers

39
00:02:10,410 --> 00:02:11,880
of this size and speed.

40
00:02:11,880 --> 00:02:13,620
Just to get computers to work.

41
00:02:13,620 --> 00:02:14,180
That's right.

42
00:02:14,180 --> 00:02:16,050
They were pretty
unreliable in the beginning

43
00:02:16,050 --> 00:02:18,920
and they had to be, or were
afraid they'd be unreliable.

44
00:02:18,920 --> 00:02:20,970
They managed to lick
all those problems.

45
00:02:20,970 --> 00:02:24,690
By about the middle 1950s, they
were being used very heavily

46
00:02:24,690 --> 00:02:25,710
already.

47
00:02:25,710 --> 00:02:27,330
What would have been
the next problem?

48
00:02:27,330 --> 00:02:29,610
Well, the next problem
was that of communicating

49
00:02:29,610 --> 00:02:30,702
with the computer.

50
00:02:30,702 --> 00:02:33,470
The computer basically
only understands numbers.

51
00:02:33,470 --> 00:02:36,560
I mean, you put numbers into it
and these represent something.

52
00:02:36,560 --> 00:02:40,940
Well, the next major
improvement was in the language

53
00:02:40,940 --> 00:02:43,465
that one talked to
the computer with.

54
00:02:43,465 --> 00:02:45,590
You could teach the computer
to understand English?

55
00:02:45,590 --> 00:02:47,090
Well, no, but we
can actually put

56
00:02:47,090 --> 00:02:51,260
a particular synthetic language
which is largely technical

57
00:02:51,260 --> 00:02:56,960
and which is, to some
extent, algebra, too.

58
00:02:56,960 --> 00:03:01,010
And this jargon, which is
understandable to a person,

59
00:03:01,010 --> 00:03:02,930
is much easier to work
with than numbers.

60
00:03:02,930 --> 00:03:03,890
I see.

61
00:03:03,890 --> 00:03:07,490
So once you're able to talk to
the computer, then what next?

62
00:03:07,490 --> 00:03:10,520
Well, the next major problem
is that of the ability

63
00:03:10,520 --> 00:03:12,380
to interact with the computer.

64
00:03:12,380 --> 00:03:15,800
The man-machine
interaction is very poor.

65
00:03:15,800 --> 00:03:20,030
Now this largely stems
from a problem of the fact

66
00:03:20,030 --> 00:03:22,760
that we must put only one
problem in the computer

67
00:03:22,760 --> 00:03:24,950
at a time, fundamentally.

68
00:03:24,950 --> 00:03:27,920
Now, in the process of trying to
put one problem in the computer

69
00:03:27,920 --> 00:03:30,710
at a time, in the early
days we wasted a lot

70
00:03:30,710 --> 00:03:34,190
of time putting in each problem
and taking him out and bringing

71
00:03:34,190 --> 00:03:35,270
in the next man.

72
00:03:35,270 --> 00:03:37,250
And these computers
cost a lot of money.

73
00:03:37,250 --> 00:03:38,190
Yes, indeed they do.

74
00:03:38,190 --> 00:03:41,917
They're between $300
and $600 an hour.

75
00:03:41,917 --> 00:03:43,250
So you don't want to waste time.

76
00:03:43,250 --> 00:03:44,880
No, we don't want to waste time.

77
00:03:44,880 --> 00:03:47,510
So we, as most
other centers, have

78
00:03:47,510 --> 00:03:50,390
resorted to a procedure
of batch processing.

79
00:03:50,390 --> 00:03:51,140
What is that?

80
00:03:51,140 --> 00:03:54,230
Well, we combine a
great many people

81
00:03:54,230 --> 00:03:57,510
who want to submit a job to
the computer on a single input

82
00:03:57,510 --> 00:03:58,625
magnetic tape.

83
00:03:58,625 --> 00:04:00,740
Now, these magnetic
tapes are made

84
00:04:00,740 --> 00:04:03,240
using computers such
as this one here.

85
00:04:03,240 --> 00:04:05,300
And they're
prerecorded and they're

86
00:04:05,300 --> 00:04:08,350
taken to the main computer
and used as an input tape

87
00:04:08,350 --> 00:04:09,350
to the main computer.

88
00:04:09,350 --> 00:04:10,250
Oh, I see.

89
00:04:10,250 --> 00:04:12,320
So you solve everybody's
problem in turn.

90
00:04:12,320 --> 00:04:13,310
Well, in a batch.

91
00:04:13,310 --> 00:04:15,560
A group of people,
like catching a bus.

92
00:04:15,560 --> 00:04:17,982
And the computer calculates
this batch through

93
00:04:17,982 --> 00:04:20,149
and produces an output tape
and then the output tape

94
00:04:20,149 --> 00:04:21,524
and brought over
to the computers

95
00:04:21,524 --> 00:04:23,180
and the results are printed out.

96
00:04:23,180 --> 00:04:25,340
Well, as you might
imagine, this takes

97
00:04:25,340 --> 00:04:29,200
a few hours to get anything
through the computer at all.

98
00:04:29,200 --> 00:04:30,830
And that raises more problems.

99
00:04:30,830 --> 00:04:33,230
That raises a lot of
problems and people

100
00:04:33,230 --> 00:04:36,380
are very frustrated with
this type of interaction.

101
00:04:36,380 --> 00:04:37,640
So what do you plan to do?

102
00:04:37,640 --> 00:04:40,340
Well, what we are
trying to develop

103
00:04:40,340 --> 00:04:43,180
is a new scheme of
using the computer

104
00:04:43,180 --> 00:04:46,400
that we call time-sharing
which consists

105
00:04:46,400 --> 00:04:48,290
of attaching a large
number of consoles

106
00:04:48,290 --> 00:04:51,310
to the central computer.

107
00:04:51,310 --> 00:04:52,540
It's a little noisy out here.

108
00:04:52,540 --> 00:04:54,760
I think it would be a lot
easier if I could explain

109
00:04:54,760 --> 00:04:58,360
to you in my office, how
to use one of these, where

110
00:04:58,360 --> 00:04:59,830
I have one remotely attached.

111
00:04:59,830 --> 00:05:04,490


112
00:05:04,490 --> 00:05:06,020
The idea, I gather,
then is to allow

113
00:05:06,020 --> 00:05:08,330
a whole lot of different
people to use the computer

114
00:05:08,330 --> 00:05:09,590
all at the same time.

115
00:05:09,590 --> 00:05:10,250
That's right.

116
00:05:10,250 --> 00:05:11,750
And this is one of
the consoles they

117
00:05:11,750 --> 00:05:13,280
might be using in the future.

118
00:05:13,280 --> 00:05:15,597
It looks very much like
just a regular typewriter,

119
00:05:15,597 --> 00:05:16,430
electric typewriter.

120
00:05:16,430 --> 00:05:17,222
Well, that's right.

121
00:05:17,222 --> 00:05:20,030
That's what it really is, it's
a small amount of electronics

122
00:05:20,030 --> 00:05:22,550
in the bottom and it's
attached to the computer.

123
00:05:22,550 --> 00:05:23,990
That's the real distinction.

124
00:05:23,990 --> 00:05:25,940
So you might have one
here in this office,

125
00:05:25,940 --> 00:05:27,570
and I might have one
in say, my office,

126
00:05:27,570 --> 00:05:29,195
and it might be one
in some University,

127
00:05:29,195 --> 00:05:30,153
or something like that?

128
00:05:30,153 --> 00:05:31,067
That's our hope.

129
00:05:31,067 --> 00:05:32,150
We think that can be done.

130
00:05:32,150 --> 00:05:34,100
How many might you
have all together?

131
00:05:34,100 --> 00:05:36,590
Well, at the moment
we're just working with,

132
00:05:36,590 --> 00:05:39,090
at the very moment,
5, and in few weeks

133
00:05:39,090 --> 00:05:40,940
we should be up to about 20.

134
00:05:40,940 --> 00:05:42,543
And ultimately, we
hope to see 100s

135
00:05:42,543 --> 00:05:44,210
but we'll probably
have to have a better

136
00:05:44,210 --> 00:05:45,607
computer before that is.

137
00:05:45,607 --> 00:05:47,690
But I still don't understand
now, how all of these

138
00:05:47,690 --> 00:05:50,270
can be using the computer
at the same time?

139
00:05:50,270 --> 00:05:52,910
Well, to really explain that
I have to backtrack a bit

140
00:05:52,910 --> 00:05:55,310
and talk about a little bit
about how a computer really

141
00:05:55,310 --> 00:05:55,920
works.

142
00:05:55,920 --> 00:05:56,420
OK?

143
00:05:56,420 --> 00:05:56,920
Sure.

144
00:05:56,920 --> 00:05:57,740
So.

145
00:05:57,740 --> 00:06:00,430
Let me show you.

146
00:06:00,430 --> 00:06:01,568
The blackboard.

147
00:06:01,568 --> 00:06:04,200


148
00:06:04,200 --> 00:06:06,360
First of all, let
me explain just

149
00:06:06,360 --> 00:06:09,360
for a moment what a program
is and how a computer really

150
00:06:09,360 --> 00:06:11,490
works.

151
00:06:11,490 --> 00:06:16,860
The computer's organized
around a large set

152
00:06:16,860 --> 00:06:21,150
of pigeonholes which is really
called a memory, usually.

153
00:06:21,150 --> 00:06:23,460
In each memory,
the memory consists

154
00:06:23,460 --> 00:06:27,030
of these various
cells or pigeonholes,

155
00:06:27,030 --> 00:06:29,130
which I can put a number into.

156
00:06:29,130 --> 00:06:35,720
Now the computer, say,
has a particular numbers

157
00:06:35,720 --> 00:06:36,540
in each cell.

158
00:06:36,540 --> 00:06:40,640
And it's very hard to
read this way, of course,

159
00:06:40,640 --> 00:06:43,950
but some of the
numbers represent

160
00:06:43,950 --> 00:06:46,190
instructions and
some of the numbers

161
00:06:46,190 --> 00:06:51,380
represent data and the person
who prepares these instructions

162
00:06:51,380 --> 00:06:55,490
and data is the one who keeps
it straight what they are.

163
00:06:55,490 --> 00:06:58,240
I see, those numbers
might not mean numbers,

164
00:06:58,240 --> 00:07:01,580
just as they might mean to add
something in one pigeonhole

165
00:07:01,580 --> 00:07:03,650
to something else and
something other pigeonhole.

166
00:07:03,650 --> 00:07:04,430
That's right.

167
00:07:04,430 --> 00:07:06,638
Some of them are instructions,
some of them are data.

168
00:07:06,638 --> 00:07:09,560
Now, the real
computer, in the sense

169
00:07:09,560 --> 00:07:15,140
that the, oops, device
that does the computation

170
00:07:15,140 --> 00:07:18,590
is a central processing unit
which has the calculating part,

171
00:07:18,590 --> 00:07:22,065
and this is usually called CPU.

172
00:07:22,065 --> 00:07:23,690
That's where the work
is actually done.

173
00:07:23,690 --> 00:07:24,232
That's right.

174
00:07:24,232 --> 00:07:26,330
This is the device that
really does the work.

175
00:07:26,330 --> 00:07:29,240
And it is able to operate on
one instruction at a time.

176
00:07:29,240 --> 00:07:31,700
And it does, the arrangement
is that it does the first one

177
00:07:31,700 --> 00:07:34,400
instruction, then the
next instruction, whatever

178
00:07:34,400 --> 00:07:37,135
that one might be, and
keep doing instructions

179
00:07:37,135 --> 00:07:38,510
and then some of
the instructions

180
00:07:38,510 --> 00:07:42,500
cause the program to
go back in a loop.

181
00:07:42,500 --> 00:07:45,410
Now this collection of
instructions, as I just said,

182
00:07:45,410 --> 00:07:49,100
is called the program and they
do a particular job for it

183
00:07:49,100 --> 00:07:51,620
and somebody has to
prepare them in advance.

184
00:07:51,620 --> 00:07:53,540
How do you get it
into this memory?

185
00:07:53,540 --> 00:07:56,480
Well normally one is read it in
from punch cards or something

186
00:07:56,480 --> 00:07:57,170
of this sort.

187
00:07:57,170 --> 00:07:59,840


188
00:07:59,840 --> 00:08:02,240
And this comes
back to the problem

189
00:08:02,240 --> 00:08:03,890
of how do you use a computer.

190
00:08:03,890 --> 00:08:10,640
And in the past, what
we used to do is,

191
00:08:10,640 --> 00:08:13,220
we have a computer,
like 7090 for example,

192
00:08:13,220 --> 00:08:14,990
and we used to read
them in by cards

193
00:08:14,990 --> 00:08:17,240
but then we got a little
more worried about efficiency

194
00:08:17,240 --> 00:08:19,365
and so we now read them in
from magnetic tape which

195
00:08:19,365 --> 00:08:21,470
has been prerecorded.

196
00:08:21,470 --> 00:08:25,160
Now this calculates
inside the computer

197
00:08:25,160 --> 00:08:27,080
and it's just one problem
at a time, remember,

198
00:08:27,080 --> 00:08:28,370
that's the problem.

199
00:08:28,370 --> 00:08:31,790
And after several minutes
the output results

200
00:08:31,790 --> 00:08:34,190
are accumulated on
magnetic tape and they're

201
00:08:34,190 --> 00:08:37,730
taken over to another
computer and printed later.

202
00:08:37,730 --> 00:08:40,190
Now this is the older
batch processing technique.

203
00:08:40,190 --> 00:08:41,240
One problem at a time.

204
00:08:41,240 --> 00:08:44,840
One problem at a time
and it's pretty clumsy

205
00:08:44,840 --> 00:08:47,690
because of the weak
interaction between the man

206
00:08:47,690 --> 00:08:48,560
and the machine.

207
00:08:48,560 --> 00:08:51,480
Now what we're trying to
do is superimpose on this.

208
00:08:51,480 --> 00:08:56,480
Instead of parallel
consoles which are not

209
00:08:56,480 --> 00:08:58,310
all near the computer, in fact.

210
00:08:58,310 --> 00:09:00,290
In fact, most of
them are remote.

211
00:09:00,290 --> 00:09:04,310
And let the users use
these with a reaction

212
00:09:04,310 --> 00:09:07,430
time of a few seconds
rather than a few hours.

213
00:09:07,430 --> 00:09:10,100
It takes a few hours to get
through this path because

214
00:09:10,100 --> 00:09:13,258
of the batching and queueing,
and it takes, we're hoping

215
00:09:13,258 --> 00:09:15,800
and we're actually able to do
it at the moment a few seconds.

216
00:09:15,800 --> 00:09:17,925
We'd like to get it down
to a fraction of a second,

217
00:09:17,925 --> 00:09:20,960
eventually.

218
00:09:20,960 --> 00:09:23,930
Most of these are typewriter
consoles much like we saw,

219
00:09:23,930 --> 00:09:28,490
and we actually are
getting to 21 consoles.

220
00:09:28,490 --> 00:09:31,220
Eventually, we'd like
to see graphical display

221
00:09:31,220 --> 00:09:34,040
on this kind of a console, there
are technical problems still,

222
00:09:34,040 --> 00:09:34,880
but.

223
00:09:34,880 --> 00:09:40,730
As an experiment, to try out
some of these more elaborate

224
00:09:40,730 --> 00:09:42,470
advanced ideas, we
actually are trying

225
00:09:42,470 --> 00:09:45,560
to tie two small
computers to this one.

226
00:09:45,560 --> 00:09:49,280
We have one, a PDP-1 computer
of the DEC Corporation

227
00:09:49,280 --> 00:09:53,660
and we have also a 1620
computer of the IBM Corporation.

228
00:09:53,660 --> 00:09:56,750
These are more experiments
which are a little too advanced

229
00:09:56,750 --> 00:09:58,430
to explain today.

230
00:09:58,430 --> 00:10:00,540
Oh, I can see now
what you'd like to do.

231
00:10:00,540 --> 00:10:02,090
But I don't see how you do it.

232
00:10:02,090 --> 00:10:02,600
OK.

233
00:10:02,600 --> 00:10:04,558
Well, I have to go back
then a little bit more.

234
00:10:04,558 --> 00:10:07,820
Let me come back here
to the same drawing

235
00:10:07,820 --> 00:10:12,590
of the memory and the
processor, which can only

236
00:10:12,590 --> 00:10:14,390
be at one place at a time.

237
00:10:14,390 --> 00:10:17,900
And let me redraw this a bit.

238
00:10:17,900 --> 00:10:24,300
What we have is
one master program,

239
00:10:24,300 --> 00:10:27,990
which I'll call a supervisor.

240
00:10:27,990 --> 00:10:30,830
And this is the program that
is going to run everything.

241
00:10:30,830 --> 00:10:33,950
And normally this computer, when
it's in the time-sharing mode,

242
00:10:33,950 --> 00:10:38,330
is operating a program
in the supervisor.

243
00:10:38,330 --> 00:10:41,840
We also have a program
for each of the typewriter

244
00:10:41,840 --> 00:10:48,540
users, so u1, u2,
u3, u4, and so forth.

245
00:10:48,540 --> 00:10:49,950
They might not be the same size.

246
00:10:49,950 --> 00:10:52,200
No, they're not all the same
size, that's quite right.

247
00:10:52,200 --> 00:10:55,230
That's what I meant
to try and convey.

248
00:10:55,230 --> 00:11:00,810
Now the consoles have to type
in to the computer and they

249
00:11:00,810 --> 00:11:02,950
type into the
supervisor and back out.

250
00:11:02,950 --> 00:11:05,010
So the messages are
handled by the supervisor

251
00:11:05,010 --> 00:11:06,990
for each of the programs.

252
00:11:06,990 --> 00:11:11,580
Now the supervisor is going
to relay these messages

253
00:11:11,580 --> 00:11:15,600
to the particular programs
as they are, and keep track

254
00:11:15,600 --> 00:11:19,980
of which user's talking to
which typewriter, of course,

255
00:11:19,980 --> 00:11:21,880
and so forth.

256
00:11:21,880 --> 00:11:24,990
And this would be
straightforward enough,

257
00:11:24,990 --> 00:11:27,330
except that quite often
we'll get more programs

258
00:11:27,330 --> 00:11:28,530
than we have room for.

259
00:11:28,530 --> 00:11:29,590
Yes.

260
00:11:29,590 --> 00:11:30,550
So we have a problem.

261
00:11:30,550 --> 00:11:35,010
And what we have here
is another device

262
00:11:35,010 --> 00:11:37,630
called a disk,
magnetic disk memory,

263
00:11:37,630 --> 00:11:40,170
which I can illustrate this way.

264
00:11:40,170 --> 00:11:41,820
Of course, schematically.

265
00:11:41,820 --> 00:11:44,280
The big whirling disk
on which you can put--

266
00:11:44,280 --> 00:11:44,890
That's what I meant to suggest.

267
00:11:44,890 --> 00:11:45,810
--a whole lot of tape?

268
00:11:45,810 --> 00:11:46,440
Right.

269
00:11:46,440 --> 00:11:49,050
Well, it's a magnetic material
much like magnetic tape

270
00:11:49,050 --> 00:11:52,830
on the disk and it uses
essentially a comb of reading

271
00:11:52,830 --> 00:11:55,560
heads to read material off the
tracks like a phonograph pretty

272
00:11:55,560 --> 00:11:56,060
much.

273
00:11:56,060 --> 00:11:56,820
I see.

274
00:11:56,820 --> 00:11:58,110
It has a very large capacity.

275
00:11:58,110 --> 00:12:01,050
Here we're talking
about 9 million words

276
00:12:01,050 --> 00:12:02,250
of computer memory.

277
00:12:02,250 --> 00:12:04,110
A word is basically
the information

278
00:12:04,110 --> 00:12:06,600
of 6 letters or 10 numbers.

279
00:12:06,600 --> 00:12:09,210
This is the kind of thing
we're talking about.

280
00:12:09,210 --> 00:12:15,330
The basic memory of the
computer is 65,000 words.

281
00:12:15,330 --> 00:12:19,320
So this gives some idea of the
scale that we're talking about.

282
00:12:19,320 --> 00:12:20,940
And in particular,
the supervisor

283
00:12:20,940 --> 00:12:25,650
can cause programs to read back
and forth from the disk memory.

284
00:12:25,650 --> 00:12:27,150
So this is the
magnetic disk memory.

285
00:12:27,150 --> 00:12:30,660


286
00:12:30,660 --> 00:12:35,550
And we are keeping all the old
programs of the different users

287
00:12:35,550 --> 00:12:37,050
on the magnetic disk.

288
00:12:37,050 --> 00:12:39,650


289
00:12:39,650 --> 00:12:43,070
As well as-- whew the
chalk, it's a little soft.

290
00:12:43,070 --> 00:12:45,680


291
00:12:45,680 --> 00:12:47,990
That's where the supervisor
tells the magnetic disk

292
00:12:47,990 --> 00:12:49,890
to put something into
the active memory

293
00:12:49,890 --> 00:12:51,057
when I hear this regular--

294
00:12:51,057 --> 00:12:53,390
So it's playing a juggling
game, to some extent, keeping

295
00:12:53,390 --> 00:12:55,482
all of the right programs
in at the right time.

296
00:12:55,482 --> 00:12:57,690
But I still haven't really
explained the idea to you.

297
00:12:57,690 --> 00:12:59,960
No because as far as
I know, that computer

298
00:12:59,960 --> 00:13:01,550
can still only do
one thing at a time.

299
00:13:01,550 --> 00:13:02,120
That's right.

300
00:13:02,120 --> 00:13:03,360
And it does just that.

301
00:13:03,360 --> 00:13:06,140
It runs one program at a time.

302
00:13:06,140 --> 00:13:10,250
And, in particular, it
proceeds to start one program,

303
00:13:10,250 --> 00:13:13,100
operate it for a short
burst, just a burst of time.

304
00:13:13,100 --> 00:13:15,380
By a burst I mean
maybe, and at the moment

305
00:13:15,380 --> 00:13:17,532
we're dealing in
about 2 seconds.

306
00:13:17,532 --> 00:13:19,490
That may not be enough
to do the whole program.

307
00:13:19,490 --> 00:13:20,250
No, it may not.

308
00:13:20,250 --> 00:13:22,880
And the way it, now
if it isn't enough,

309
00:13:22,880 --> 00:13:24,330
there's a special
device we have,

310
00:13:24,330 --> 00:13:26,230
which is essentially
an alarm clock.

311
00:13:26,230 --> 00:13:28,010
I use the term alarm
clock, it isn't

312
00:13:28,010 --> 00:13:29,980
like the conventional one.

313
00:13:29,980 --> 00:13:31,730
What it does is as
soon as the program has

314
00:13:31,730 --> 00:13:33,470
run for a certain
period of time,

315
00:13:33,470 --> 00:13:38,600
the computer is brought
back to the supervisor,

316
00:13:38,600 --> 00:13:41,840
automatically, without the
user's program saying anything

317
00:13:41,840 --> 00:13:45,590
and the supervisor now can
reassign the computer back

318
00:13:45,590 --> 00:13:47,060
to another user.

319
00:13:47,060 --> 00:13:52,040
Now this simple round-robin
is the basic principle

320
00:13:52,040 --> 00:13:55,070
of the computer
and time-sharing.

321
00:13:55,070 --> 00:13:56,540
This computer's
fast enough to get

322
00:13:56,540 --> 00:13:59,180
through all the different
users and still maintain

323
00:13:59,180 --> 00:14:02,210
a reasonably continuous service
at the different consoles.

324
00:14:02,210 --> 00:14:03,050
I see.

325
00:14:03,050 --> 00:14:05,360
It just seems as though
the computer is handling

326
00:14:05,360 --> 00:14:06,890
all these programs
simultaneously

327
00:14:06,890 --> 00:14:09,910
but actually it's just going
round and round very fast,

328
00:14:09,910 --> 00:14:10,740
one to the next.

329
00:14:10,740 --> 00:14:12,240
Yes, you can draw
several analogies.

330
00:14:12,240 --> 00:14:16,880
One nice one is that of, say
if the supervisor program

331
00:14:16,880 --> 00:14:20,060
could be compared to, say, a
chess master in a tournament.

332
00:14:20,060 --> 00:14:24,080
And the chess master is playing
a large number of people

333
00:14:24,080 --> 00:14:25,200
all at once.

334
00:14:25,200 --> 00:14:28,230
Now the chess master is
usually very fast and clever,

335
00:14:28,230 --> 00:14:31,010
and he can essentially analyze
the situation rapidly and make

336
00:14:31,010 --> 00:14:34,100
his move, and his
opponents are normally

337
00:14:34,100 --> 00:14:35,990
require considerable
time to think it over

338
00:14:35,990 --> 00:14:38,100
and figure out the implications.

339
00:14:38,100 --> 00:14:40,940
And this is the way it is here,
the opponents are essentially

340
00:14:40,940 --> 00:14:43,880
the people at the consoles
who need quite a bit of time

341
00:14:43,880 --> 00:14:47,410
to ponder over what the
computer's just done for them,

342
00:14:47,410 --> 00:14:49,560
for it to not be
under any pressure

343
00:14:49,560 --> 00:14:52,070
and the computer can be racing
back and forth, servicing

344
00:14:52,070 --> 00:14:54,410
each one as he needs attention.

345
00:14:54,410 --> 00:14:58,345
Does it give each one of
them exactly 2 seconds

346
00:14:58,345 --> 00:15:00,470
I mean, I was just thinking
if somebody isn't busy,

347
00:15:00,470 --> 00:15:01,760
somebody may have gone
out to lunch, what

348
00:15:01,760 --> 00:15:03,050
happens to that man's console?

349
00:15:03,050 --> 00:15:05,240
Well, there are two
things there First,

350
00:15:05,240 --> 00:15:08,000
we don't bother to serve anyone
who doesn't ask for anything.

351
00:15:08,000 --> 00:15:09,470
You skip right over him.

352
00:15:09,470 --> 00:15:10,848
Right, skip right over him.

353
00:15:10,848 --> 00:15:12,890
But there's an addition,
we've refined it already

354
00:15:12,890 --> 00:15:13,432
a little bit.

355
00:15:13,432 --> 00:15:15,440
The simple scheme will
work, but we can actually

356
00:15:15,440 --> 00:15:16,430
refine it a little bit.

357
00:15:16,430 --> 00:15:18,560
And what we do is
we say the man who

358
00:15:18,560 --> 00:15:22,400
hasn't had any computation
done for him at all gets

359
00:15:22,400 --> 00:15:23,810
the first priority.

360
00:15:23,810 --> 00:15:25,940
And so he gets
served right away,

361
00:15:25,940 --> 00:15:27,998
in a short burst
of, say, 2 seconds.

362
00:15:27,998 --> 00:15:30,290
But if he doesn't get, if he
isn't satisfied with that,

363
00:15:30,290 --> 00:15:32,990
and he needs more, then we
put him in a second priority

364
00:15:32,990 --> 00:15:35,820
and we give him lower, we
give him lower priority,

365
00:15:35,820 --> 00:15:38,360
but when he does run,
he gets a longer burst.

366
00:15:38,360 --> 00:15:39,290
Oh, I see.

367
00:15:39,290 --> 00:15:41,480
So this sort of
smooths things out

368
00:15:41,480 --> 00:15:43,280
and gives the people
that need the smallest

369
00:15:43,280 --> 00:15:45,140
amount but high
priority attention

370
00:15:45,140 --> 00:15:46,620
first crack at the computer.

371
00:15:46,620 --> 00:15:48,050
Incidentally, does
this save time

372
00:15:48,050 --> 00:15:51,530
by not having to re-enter
a whole program every time?

373
00:15:51,530 --> 00:15:53,030
Let's say you are
really going to be

374
00:15:53,030 --> 00:15:54,590
using the same set
of instructions

375
00:15:54,590 --> 00:15:56,245
for a number of problems.

376
00:15:56,245 --> 00:15:57,620
Under the batch
system you'd have

377
00:15:57,620 --> 00:16:00,438
to put that setup's instruction
in every time, wouldn't you?

378
00:16:00,438 --> 00:16:00,980
That's right.

379
00:16:00,980 --> 00:16:03,180
Every time it came your
turn or your problem.

380
00:16:03,180 --> 00:16:04,730
That's right and
this is of course,

381
00:16:04,730 --> 00:16:07,010
it would be a very
difficult task

382
00:16:07,010 --> 00:16:08,850
that we had to do it
through the console.

383
00:16:08,850 --> 00:16:10,490
So what we really
are doing is we're

384
00:16:10,490 --> 00:16:12,020
keeping all the old programs.

385
00:16:12,020 --> 00:16:14,210
One of the things that
are in the disk memory

386
00:16:14,210 --> 00:16:16,490
is old programs of each user.

387
00:16:16,490 --> 00:16:19,130
And they're all files so
they don't get mixed up.

388
00:16:19,130 --> 00:16:21,110
That's right and not
getting them mixed

389
00:16:21,110 --> 00:16:23,643
up is really the
crux of the problem.

390
00:16:23,643 --> 00:16:26,060
The disk memories have been
available now for about a year

391
00:16:26,060 --> 00:16:27,477
or so but not many
people know how

392
00:16:27,477 --> 00:16:30,470
to use them because you have
to get awfully organized.

393
00:16:30,470 --> 00:16:32,900
And the real difficulty
is keeping each person

394
00:16:32,900 --> 00:16:34,400
separated from the other.

395
00:16:34,400 --> 00:16:36,740
And we have had to solve this.

396
00:16:36,740 --> 00:16:39,230
In order to solve it
we have one section

397
00:16:39,230 --> 00:16:41,840
of the supervisor,
which is nothing

398
00:16:41,840 --> 00:16:44,450
but that responsibility,
and it's a major job.

399
00:16:44,450 --> 00:16:46,200
Just to keep things
from getting mixed up?

400
00:16:46,200 --> 00:16:46,700
Yeah.

401
00:16:46,700 --> 00:16:49,310
Well, could we say,
take an example

402
00:16:49,310 --> 00:16:51,060
and actually try this out?

403
00:16:51,060 --> 00:16:53,360
Why don't we call
ourselves number 1

404
00:16:53,360 --> 00:16:55,830
and actually go
through a process here?

405
00:16:55,830 --> 00:16:56,330
OK.

406
00:16:56,330 --> 00:17:01,800
Well, I can illustrate
this best by going--

407
00:17:01,800 --> 00:17:03,980
this person would
actually type something

408
00:17:03,980 --> 00:17:05,869
into the console
to the supervisor.

409
00:17:05,869 --> 00:17:10,069
The supervisor would
then start to take action

410
00:17:10,069 --> 00:17:11,150
on the basis of that.

411
00:17:11,150 --> 00:17:13,910
Now the way the person
really initiates

412
00:17:13,910 --> 00:17:15,440
anything in this
system is he types

413
00:17:15,440 --> 00:17:17,720
commands to the supervisor.

414
00:17:17,720 --> 00:17:21,410
There are about, in the order
of a few dozen commands.

415
00:17:21,410 --> 00:17:26,630
We, and these are mostly
fairly basic instructions like,

416
00:17:26,630 --> 00:17:28,880
I want to log in to
the system or something

417
00:17:28,880 --> 00:17:31,670
or I want to start
typing in a program

418
00:17:31,670 --> 00:17:35,720
or I want to run a program or
I want to put something away

419
00:17:35,720 --> 00:17:36,410
on the disk.

420
00:17:36,410 --> 00:17:39,655
These are rather basic
manipulating commands.

421
00:17:39,655 --> 00:17:41,030
So in general the
commander types

422
00:17:41,030 --> 00:17:44,960
to the supervisor and
they cause something

423
00:17:44,960 --> 00:17:46,960
to happen on the programs
that are on the disk.

424
00:17:46,960 --> 00:17:52,190
Now let me, actually, let
me explain just a moment

425
00:17:52,190 --> 00:17:54,108
what really happens.

426
00:17:54,108 --> 00:17:56,150
What really happens is
when a man types a command

427
00:17:56,150 --> 00:17:58,280
into the supervisor,
the supervisor

428
00:17:58,280 --> 00:18:00,620
goes and gets a
program out of the disk

429
00:18:00,620 --> 00:18:04,250
which it starts running
as say, u1's program.

430
00:18:04,250 --> 00:18:05,840
It brings it in
for him and starts

431
00:18:05,840 --> 00:18:08,540
to run it as his program.

432
00:18:08,540 --> 00:18:12,947
And, of course,
at that point he's

433
00:18:12,947 --> 00:18:15,530
got a program working for him
it may do something standardized

434
00:18:15,530 --> 00:18:18,330
in a standard way for him and
then drop out of the picture.

435
00:18:18,330 --> 00:18:20,360
Another command and a
particular one of them

436
00:18:20,360 --> 00:18:22,700
is he may want to start an
old program so it's bringing

437
00:18:22,700 --> 00:18:23,990
an old program for him.

438
00:18:23,990 --> 00:18:26,450
Now that's what we saw at
the beginning of the program

439
00:18:26,450 --> 00:18:28,695
and I'd like to
illustrate that first.

440
00:18:28,695 --> 00:18:31,320
I'd like to see that now that we
understand it a little better.

441
00:18:31,320 --> 00:18:32,778
I'm going to type
the start command

442
00:18:32,778 --> 00:18:34,400
and that's the first one.

443
00:18:34,400 --> 00:18:37,670
And that will bring
our old program

444
00:18:37,670 --> 00:18:41,790
which is already on file into a
position where we can actually

445
00:18:41,790 --> 00:18:42,290
use it.

446
00:18:42,290 --> 00:18:42,920
That's right.

447
00:18:42,920 --> 00:18:44,630
Now as soon as I
type start, I've

448
00:18:44,630 --> 00:18:46,040
actually started my program.

449
00:18:46,040 --> 00:18:47,610
So in order to see
this clearly, I

450
00:18:47,610 --> 00:18:49,860
think we'll have to bring
the camera in for you close.

451
00:18:49,860 --> 00:18:49,940
Oh, yes.

452
00:18:49,940 --> 00:18:51,023
We can do that, all right.

453
00:18:51,023 --> 00:18:54,530
Get a look at what's
being typed out.

454
00:18:54,530 --> 00:18:56,990
I'm moderately familiar
with this keyboard.

455
00:18:56,990 --> 00:18:58,657
Of course, one gets
more and more faster

456
00:18:58,657 --> 00:19:00,990
when you get used to it and
this is one of the problems.

457
00:19:00,990 --> 00:19:03,170
We have to study the
human engineering, too,

458
00:19:03,170 --> 00:19:04,940
of people learning
to use equipment.

459
00:19:04,940 --> 00:19:06,440
Too often the
equipment in the past

460
00:19:06,440 --> 00:19:08,690
has sort of been designed
for other machines,

461
00:19:08,690 --> 00:19:09,890
really not for people.

462
00:19:09,890 --> 00:19:13,250
But I start off by pushing
one button and then I'm off,

463
00:19:13,250 --> 00:19:15,680
I think I'm off to the races.

464
00:19:15,680 --> 00:19:16,700
Let's see how this goes.

465
00:19:16,700 --> 00:19:19,250


466
00:19:19,250 --> 00:19:21,960
Wow.

467
00:19:21,960 --> 00:19:29,550
Now, the computer
is pausing there,

468
00:19:29,550 --> 00:19:32,400
the little pauses are
actual hesitations

469
00:19:32,400 --> 00:19:34,740
when it's switching from
one user to another.

470
00:19:34,740 --> 00:19:36,210
Oh that "wait,"
for instance, meant

471
00:19:36,210 --> 00:19:38,252
that it wasn't ready for
you just at that moment.

472
00:19:38,252 --> 00:19:40,470
Well the "wait" meant that
the supervisor had gotten

473
00:19:40,470 --> 00:19:43,110
my command, "start"
and it actually

474
00:19:43,110 --> 00:19:44,880
started pulling in that program.

475
00:19:44,880 --> 00:19:46,920
If I wanted to type
there, I would really

476
00:19:46,920 --> 00:19:49,650
be trying to put input
into that command program.

477
00:19:49,650 --> 00:19:52,200
Now that particular, well
it's actually an old program

478
00:19:52,200 --> 00:19:55,170
that I wrote, that particular
program that I illustrated

479
00:19:55,170 --> 00:19:59,280
on the board doesn't
need any input

480
00:19:59,280 --> 00:20:01,480
and so it would have
been ignored by that,

481
00:20:01,480 --> 00:20:02,610
if I tried to type.

482
00:20:02,610 --> 00:20:04,340
Instead it wanted to talk to me.

483
00:20:04,340 --> 00:20:06,120
In particular, it says, "Hello.

484
00:20:06,120 --> 00:20:09,660
I do square roots and
compute hypotenuses.

485
00:20:09,660 --> 00:20:11,562
Please type either
root or triangle."

486
00:20:11,562 --> 00:20:13,020
Now would this be
a typical program

487
00:20:13,020 --> 00:20:15,420
or this is a
particularly simplified

488
00:20:15,420 --> 00:20:16,870
one for our purposes?

489
00:20:16,870 --> 00:20:19,620
This is an extremely simple
one and not at all typical

490
00:20:19,620 --> 00:20:21,660
of what we'd use.

491
00:20:21,660 --> 00:20:23,770
That gives us a chance
to understand it.

492
00:20:23,770 --> 00:20:25,470
But I have to pick
an example, which

493
00:20:25,470 --> 00:20:29,820
is good for pretty much
everyone, in a language

494
00:20:29,820 --> 00:20:31,560
that everyone understands.

495
00:20:31,560 --> 00:20:33,300
But it's illustrative
of the technique

496
00:20:33,300 --> 00:20:34,500
of man-machine interaction.

497
00:20:34,500 --> 00:20:38,160
I think it's very clear that
this is the technique that

498
00:20:38,160 --> 00:20:41,580
would work in, say, an airline
reservation or a small business

499
00:20:41,580 --> 00:20:45,053
asking about a payroll problem
or something of this sort.

500
00:20:45,053 --> 00:20:46,720
So, why don't we do
a square root, then?

501
00:20:46,720 --> 00:20:47,370
OK.

502
00:20:47,370 --> 00:20:48,180
So we type "root."

503
00:20:48,180 --> 00:20:52,635


504
00:20:52,635 --> 00:20:55,605
[PRINTER CLATTERING]

505
00:20:55,605 --> 00:20:59,493


506
00:20:59,493 --> 00:21:01,910
"Please type a number up to 6
digits, with decimal point."

507
00:21:01,910 --> 00:21:03,410
Want to pick one?

508
00:21:03,410 --> 00:21:05,410
I'm trying to think of
one I know the answer to.

509
00:21:05,410 --> 00:21:06,460
How about 49?

510
00:21:06,460 --> 00:21:07,240
49.

511
00:21:07,240 --> 00:21:08,680
OK.

512
00:21:08,680 --> 00:21:11,230
I don't think, the
computer, of course,

513
00:21:11,230 --> 00:21:13,050
doesn't have to have
the simple ones.

514
00:21:13,050 --> 00:21:14,968
I have no way of
checking it, if it was

515
00:21:14,968 --> 00:21:16,260
any more complicated than that.

516
00:21:16,260 --> 00:21:19,560


517
00:21:19,560 --> 00:21:24,360
Now, one thing
that's important here

518
00:21:24,360 --> 00:21:27,210
is that we could spend
considerable time arguing

519
00:21:27,210 --> 00:21:28,960
about what the right
number to have picked

520
00:21:28,960 --> 00:21:32,368
was and it really
wouldn't have mattered

521
00:21:32,368 --> 00:21:34,410
because the computers is
working for other people

522
00:21:34,410 --> 00:21:38,520
and we're not losing any of this
rather valuable computer time.

523
00:21:38,520 --> 00:21:41,040
I guess the computer's
right, it said 7.

524
00:21:41,040 --> 00:21:42,120
Now what does it say?

525
00:21:42,120 --> 00:21:43,890
"Do you wish to continue?"

526
00:21:43,890 --> 00:21:44,390
Sure.

527
00:21:44,390 --> 00:21:45,290
Oh, let's type that.

528
00:21:45,290 --> 00:21:48,218


529
00:21:48,218 --> 00:21:51,634
[PRINTER CLATTERING]

530
00:21:51,634 --> 00:21:52,610
Huh.

531
00:21:52,610 --> 00:21:54,650
Please type yes or no.

532
00:21:54,650 --> 00:21:58,400
Well I think the point here
is that we didn't really

533
00:21:58,400 --> 00:22:01,910
have to know exactly what
to type, see the program was

534
00:22:01,910 --> 00:22:05,270
prepared to check
and see if we made

535
00:22:05,270 --> 00:22:07,370
one of what it
expected us to assume

536
00:22:07,370 --> 00:22:09,410
was the right answer, yes or no.

537
00:22:09,410 --> 00:22:11,870
But when we didn't do it,
it's able to come back and ask

538
00:22:11,870 --> 00:22:12,917
us a question about it.

539
00:22:12,917 --> 00:22:15,500
That's interesting, because this
would be one of the problems,

540
00:22:15,500 --> 00:22:17,330
I imagine, in the batch system.

541
00:22:17,330 --> 00:22:18,980
That it might have
taken you 24 hours

542
00:22:18,980 --> 00:22:21,155
to find out that you had
given the wrong instruction

543
00:22:21,155 --> 00:22:22,530
or the wrong answer
on your part.

544
00:22:22,530 --> 00:22:23,940
That's all too true.

545
00:22:23,940 --> 00:22:25,290
That's completely the case.

546
00:22:25,290 --> 00:22:28,880
And it's one of the infuriating
features of the older

547
00:22:28,880 --> 00:22:30,590
way of doing things
and what slows down

548
00:22:30,590 --> 00:22:32,990
most problems with preparation.

549
00:22:32,990 --> 00:22:34,100
That's good.

550
00:22:34,100 --> 00:22:39,135
So let's say what we
should have said, I guess.

551
00:22:39,135 --> 00:22:42,045
[PRINTER CLATTERING]

552
00:22:42,045 --> 00:22:43,500


553
00:22:43,500 --> 00:22:45,910
It says, "Please type
either root or triangle."

554
00:22:45,910 --> 00:22:47,170
So we're back there again.

555
00:22:47,170 --> 00:22:49,420
All right, well, why don't
we do a triangle this time.

556
00:22:49,420 --> 00:22:49,920
OK.

557
00:22:49,920 --> 00:22:57,570


558
00:22:57,570 --> 00:23:00,080
Well this is going to give us
the hypotenuse of a triangle?

559
00:23:00,080 --> 00:23:01,622
I remember earlier
it said something.

560
00:23:01,622 --> 00:23:03,263
Yes, that's right.

561
00:23:03,263 --> 00:23:06,574
[PRINTER CLATTERING]

562
00:23:06,574 --> 00:23:10,370


563
00:23:10,370 --> 00:23:13,560
It says "Give me two sides and
I'll give you the hypotenuse.

564
00:23:13,560 --> 00:23:15,300
Type up the 6 digits,
with decimal point.

565
00:23:15,300 --> 00:23:17,250
The first side is."

566
00:23:17,250 --> 00:23:19,140
Oh, let's pick
something a little more

567
00:23:19,140 --> 00:23:21,240
difficult than ones we know.

568
00:23:21,240 --> 00:23:22,600
All right, you pick one out.

569
00:23:22,600 --> 00:23:23,190
OK.

570
00:23:23,190 --> 00:23:23,700
Let's try.

571
00:23:23,700 --> 00:23:27,480


572
00:23:27,480 --> 00:23:30,882
Typing 1.0

573
00:23:30,882 --> 00:23:32,500
"And the other side."

574
00:23:32,500 --> 00:23:34,870
Say, the other side,
let me make 12.

575
00:23:34,870 --> 00:23:36,680
All right

576
00:23:36,680 --> 00:23:38,930
I'm not sure, I think the
answer is going to be 13.

577
00:23:38,930 --> 00:23:40,180
But I don't remember for sure.

578
00:23:40,180 --> 00:23:45,110


579
00:23:45,110 --> 00:23:48,790
No, it's not 13.

580
00:23:48,790 --> 00:23:51,070
12.04.

581
00:23:51,070 --> 00:23:53,407
Well, we better tell
them no, we're finished.

582
00:23:53,407 --> 00:23:54,490
"Do you wish to continue?"

583
00:23:54,490 --> 00:23:54,990
No.

584
00:23:54,990 --> 00:23:55,780
OK.

585
00:23:55,780 --> 00:23:57,280
We won't make any
mistake this time.

586
00:23:57,280 --> 00:24:00,280


587
00:24:00,280 --> 00:24:01,950
Tell me one thing--

588
00:24:01,950 --> 00:24:04,525
Now this is the
program at this point.

589
00:24:04,525 --> 00:24:07,840
It is sort of cleaning
up the detail.

590
00:24:07,840 --> 00:24:10,720
Now does this, it seems to
me that this would actually

591
00:24:10,720 --> 00:24:13,240
save a lot of time and
that you might wind up

592
00:24:13,240 --> 00:24:16,620
with the same problem had
before the computer looping.

593
00:24:16,620 --> 00:24:21,060
It's a complicated question
about saving or using time.

594
00:24:21,060 --> 00:24:23,520
We actually save
quite a bit of time.

595
00:24:23,520 --> 00:24:27,300
In the sense that the person
is able to be more selective

596
00:24:27,300 --> 00:24:29,610
and be more shrewd in
the use of the computer.

597
00:24:29,610 --> 00:24:32,620
He doesn't set it on brute
force approaches to problems

598
00:24:32,620 --> 00:24:35,250
but actually can probe
around and examine

599
00:24:35,250 --> 00:24:36,870
some of the smaller
answers first

600
00:24:36,870 --> 00:24:38,670
before he decides to go ahead.

601
00:24:38,670 --> 00:24:41,640
So this saves time but then
there are other factors which

602
00:24:41,640 --> 00:24:44,550
make it look the other way.

603
00:24:44,550 --> 00:24:47,040
And one, well one
of them is the fact

604
00:24:47,040 --> 00:24:49,560
that we do have a certain
amount of overhead switching

605
00:24:49,560 --> 00:24:51,480
between one user and another.

606
00:24:51,480 --> 00:24:55,380
Now this is controllable
but, and we can keep it,

607
00:24:55,380 --> 00:25:00,900
say, down to 10%, but it's
one aspect of the system that

608
00:25:00,900 --> 00:25:02,838
uses more time.

609
00:25:02,838 --> 00:25:04,630
Probably the most
frightening part, though,

610
00:25:04,630 --> 00:25:08,910
is the fact that we really made
the computer extremely easy

611
00:25:08,910 --> 00:25:13,650
to use here and so it's very
clear that in the long run,

612
00:25:13,650 --> 00:25:16,920
we're going to increase
the need for computer time

613
00:25:16,920 --> 00:25:18,840
by a large amount.

614
00:25:18,840 --> 00:25:21,350
Well, we already have
increasing needs for a computer.

615
00:25:21,350 --> 00:25:23,850
They're being used more and
more and all sorts of businesses

616
00:25:23,850 --> 00:25:25,440
and science and
events and so on.

617
00:25:25,440 --> 00:25:27,720
What would happen if
we don't institute

618
00:25:27,720 --> 00:25:29,520
such systems as time-sharing?

619
00:25:29,520 --> 00:25:31,290
Well to see that
better, I really

620
00:25:31,290 --> 00:25:33,665
ought to show you our present
operation with a high speed

621
00:25:33,665 --> 00:25:35,040
printer.

622
00:25:35,040 --> 00:25:38,470
[PRINTER CLATTERING LOUDLY]

623
00:25:38,470 --> 00:25:52,150


624
00:25:52,150 --> 00:25:54,650
Oh, that is really printing out
a lot of material, isn't it?

625
00:25:54,650 --> 00:25:55,790
What is all that?

626
00:25:55,790 --> 00:25:59,240
This is output from the
computer and the people using it

627
00:25:59,240 --> 00:26:00,800
in the conventional way.

628
00:26:00,800 --> 00:26:03,620
Notice a tremendous amount
of material has to come out.

629
00:26:03,620 --> 00:26:05,900
That might all be one
problem, one person's problem?

630
00:26:05,900 --> 00:26:07,130
Yes, it's just one person.

631
00:26:07,130 --> 00:26:08,630
The reason he has
to ask for so much

632
00:26:08,630 --> 00:26:11,120
is that he really
couldn't anticipate

633
00:26:11,120 --> 00:26:13,800
having what he
needed to know, so he

634
00:26:13,800 --> 00:26:15,422
asked for more than enough.

635
00:26:15,422 --> 00:26:17,130
So what's likely to
happen in the future?

636
00:26:17,130 --> 00:26:21,020
Well just imagine having 10
times as much computer here,

637
00:26:21,020 --> 00:26:22,820
we'd have 10 times
as much paper.

638
00:26:22,820 --> 00:26:25,460
This would leave us with
freight cars of paper

639
00:26:25,460 --> 00:26:27,045
coming up to the
Computation Center.

640
00:26:27,045 --> 00:26:29,420
Well, by time-sharing, for
instance, in the demonstration

641
00:26:29,420 --> 00:26:31,980
that we did, how much time
did we use up on the computer?

642
00:26:31,980 --> 00:26:34,000
Well, the output here
is when we logged out

643
00:26:34,000 --> 00:26:36,110
it said it only used 18 seconds.

644
00:26:36,110 --> 00:26:39,170
Well, thank you very
much, Dr. Corbato.

645
00:26:39,170 --> 00:26:42,560
So, next week we're going
to visit the Bathyscaphe

646
00:26:42,560 --> 00:26:45,260
Trieste, which is
in Boston to help

647
00:26:45,260 --> 00:26:48,650
locate the ill-fated
submarine, Thresher.

648
00:26:48,650 --> 00:26:51,080
We're going to meet some
of the Naval officers

649
00:26:51,080 --> 00:26:54,290
who operate the Trieste,
and learn about some

650
00:26:54,290 --> 00:26:56,150
of the scientific
experiments that they've

651
00:26:56,150 --> 00:27:00,230
conducted in the ocean at
depths as much as 7 miles.

652
00:27:00,230 --> 00:27:02,780
This is John Fitch, your
MIT science reporter.

653
00:27:02,780 --> 00:27:05,440
I hope that you'll
be with us then.

654
00:27:05,440 --> 00:27:37,000