1
00:00:00,000 --> 00:00:03,840
Even with all the time given up to me
now by lockdown

2
00:00:03,840 --> 00:00:07,200
there still isn't enough time. it takes
for ever... !

3
00:00:07,200 --> 00:00:12,080
And it's a good job i'm semi-retired from

4
00:00:12,080 --> 00:00:16,720
being a part ofthe School of Computer
Science at [University of] Nottingham because

5
00:00:16,720 --> 00:00:21,039
honestly, as a regular employee,  I could not justify

6
00:00:21,039 --> 00:00:27,680
the hours i've spent on my latest ultra-fascinating project.

7
00:00:27,680 --> 00:00:32,480
I refer you, my dear Watson, to the
strange case

8
00:00:32,480 --> 00:00:38,719
of Dennis Ritchie's missing thesis.
We all know who Dennis is. UNIX was

9
00:00:38,719 --> 00:00:42,160
invented by Ken and Dennis:  Ken Thompson 
and Dennis Ritchie.

10
00:00:42,160 --> 00:00:46,320
If you want to simplify things
Ken was sort of in charge of UNIX the

11
00:00:46,320 --> 00:00:50,800
operating system. [But] anything to do 
with creating C language

12
00:00:50,800 --> 00:00:55,039
facilities and C compilers
because, remember, UNIX was one of the

13
00:00:55,039 --> 00:00:59,120
very first operating systems
to be written in a specially invented

14
00:00:59,120 --> 00:01:03,600
high-level language called
C. And Dennis was an absolute whiz

15
00:01:03,600 --> 00:01:07,920
with the C compiler, And it needed both
of them to do this,

16
00:01:07,920 --> 00:01:13,200
So, again, many of you will know Dennis because

17
00:01:13,200 --> 00:01:17,200
one of the best-selling computer
programming textbooks of all time

18
00:01:17,200 --> 00:01:21,200
is "The C Programming Language" by Dennis Ritchie

19
00:01:21,200 --> 00:01:24,560
and our good Computerphile colleague and friend

20
00:01:24,560 --> 00:01:28,560
Brian Kernighan. There's some big gaps in
our knowledge about

21
00:01:28,560 --> 00:01:32,720
Dennis and I should remind you, I think,
sadly, he died in 2010 [Correction: October 2011]

22
00:01:32,720 --> 00:01:39,119
and he has been must much missed in
the 10 years or so since that.

23
00:01:39,119 --> 00:01:45,119
But recently the Computer History Museum (CHM)
in the USA

24
00:01:45,119 --> 00:01:51,200
came up with a story which just amazed
everybody. People who knew him said:

25
00:01:51,200 --> 00:01:55,040
"Yeah! Yeah! he went to Harvard 
and we think he did an

26
00:01:55,040 --> 00:01:59,600
Applied Mathematics degree.
So of course he was Dr

27
00:01:59,600 --> 00:02:03,759
Ritchie wasn't he? He did a PhD ? 
Well the answer is he

28
00:02:03,759 --> 00:02:09,599
studied for a PhD
but it never properly got submitted.

29
00:02:09,599 --> 00:02:14,400
The Examination Panel was appointed; they
never met.

30
00:02:14,400 --> 00:02:21,200
And the CHM story, basically, says "We don't totally

31
00:02:21,200 --> 00:02:25,840
feel happy about this story... " 
The cover story is that

32
00:02:25,840 --> 00:02:29,040
there was a standoff and that there was a need

33
00:02:29,040 --> 00:02:34,160
to submit bound library copies of his thesis. 

34
00:02:34,160 --> 00:02:40,800
Dennis objected to having to pay
to have this done. He thought he

35
00:02:40,800 --> 00:02:44,560
made enough sacrifices already
(presumably) in keeping himself alive

36
00:02:44,560 --> 00:02:48,800
while he did his PhD thesis.
It's worth saying his PhD was a very

37
00:02:48,800 --> 00:02:54,000
theoretical one. In coming out of the Maths 
department,

38
00:02:54,000 --> 00:02:58,480
perhaps not too surprising. But he
regarded his thesis as something to be

39
00:02:58,480 --> 00:03:01,120
got out of the way. But he was far more

40
00:03:01,120 --> 00:03:03,440
interested in playing Space Wars and writing

41
00:03:03,440 --> 00:03:06,720
real (prototype) C programs that were useful to 
help out Ken

42
00:03:06,720 --> 00:03:11,120
his new-found friend, you see. I know
you'll all say: "You [seem to] know everybody,

43
00:03:11,120 --> 00:03:15,360
Don't name-drop !" [But], yes, i have to confess 
I did meet Dennis

44
00:03:15,360 --> 00:03:19,920
Ritchie once. It was in 1986.
Brian had been over to see me in the

45
00:03:19,920 --> 00:03:23,120
early 80s. I paid a return visit to Bell Labs

46
00:03:23,120 --> 00:03:29,599
i had a very enjoyable lunch: Dennis
Ken, Brian and me, in which Ken regaled

47
00:03:29,599 --> 00:03:34,159
us with his story
about [how], if you were a qualified pilot

48
00:03:34,159 --> 00:03:38,319
you could get to fly a MIG in Russia, if you 
played your cards right. They [in Russia] were so

49
00:03:38,319 --> 00:03:40,560
short of foreign currency, foreign exchange, at

50
00:03:40,560 --> 00:03:44,080
that time. [So]  any gimmick that brought in
tons of dollars, you see, was

51
00:03:44,080 --> 00:03:48,159
absolutely fine.  So, yes, i did meet Dennis
just that once

52
00:03:48,159 --> 00:03:53,040
and ...  totally charming guy. It takes one 
to spot one [but] he was a computer scientist!

53
00:03:53,040 --> 00:03:57,040
And so was Ken. They were so similar to each other but

54
00:03:57,040 --> 00:04:01,760
so different, you know. 
They had this wicked way of answering

55
00:04:01,760 --> 00:04:06,480
exactly what you'd asked
not what you wish you'd asked. 

56
00:04:06,480 --> 00:04:10,879
Ken didn't get a PhD but he did enroll 
for a Masters [degree], Ii think, 

57
00:04:10,879 --> 00:04:13,439
at Berkeley and got that.  So he was known about.

58
00:04:13,439 --> 00:04:17,120
But everybody just assumed that Dennis
must have got his PhD from Harvard

59
00:04:17,120 --> 00:04:20,959
There were no hitches? Turns out there were!

60
00:04:20,959 --> 00:04:27,199
And Bell Labs were not pleased
to find that this person they'd

61
00:04:27,199 --> 00:04:30,479
appointed, who they knew was good. They chased

62
00:04:30,479 --> 00:04:34,160
after him as solidly as they chased after Ken.

63
00:04:34,160 --> 00:04:36,960
I think they were tipped off how good he was but

64
00:04:36,960 --> 00:04:40,560
to find suddenly ... I think it was somebody from
Bell Labs who wrote to Harvard and said:

65
00:04:40,560 --> 00:04:45,680
"What is going on ?"
When will Mr Ritchie get his PhD. 

66
00:04:45,680 --> 00:04:49,520
And they got back the frostiest reply you will
ever see,

67
00:04:49,520 --> 00:04:54,320
effectively saying: "We have no plans - nor
are we likely to have any plans -

68
00:04:54,320 --> 00:05:02,960
to give Mr Ritchie a PhD ... " 
He had upset somebody mightily

69
00:05:02,960 --> 00:05:06,720
and neither I, nor Brian nor his 
[Dennis's] brother Bill,

70
00:05:06,720 --> 00:05:10,000
still we don't know the true story. 
Maybe we never will.

71
00:05:10,000 --> 00:05:17,039
1967 was a long time ago, even for me.
Just think of that - I was only 23 !

72
00:05:17,039 --> 00:05:25,520
It's like back in the Jurassic, you know.

73
00:05:25,520 --> 00:05:32,720
And just to put things absolutely clear,
What would have been available to Dennis

74
00:05:32,720 --> 00:05:38,479
for his output medium for his thesis?
Really something like a clattery

75
00:05:38,479 --> 00:05:41,600
Model 37 Teletype were becoming available
[at that time].

76
00:05:41,600 --> 00:05:44,880
They had a fixed character set;  you couldn't

77
00:05:44,880 --> 00:05:47,840
sort of extend it in any way. But they were very 
good [for medium quality]. i mean, you could have a

78
00:05:47,840 --> 00:05:51,280
Teletype plugged in [to your computer] 
They were used in Stock Exchanges

79
00:05:51,280 --> 00:05:56,560
so they were a common currency [computer peripheral].
So, you could plug a Teletype into your computer.

80
00:05:56,560 --> 00:06:00,479
>> Sean: So, this thesis does exist then - it did exist?

81
00:06:00,479 --> 00:06:03,600
>> DFB: It did exist,  you see. His sister Lynn, who I think died

82
00:06:03,600 --> 00:06:06,720
two or three years ago, was very fond of her brother and

83
00:06:06,720 --> 00:06:11,360
looked after him carefully.
She had a copy of his finally submitted thesis.

84
00:06:11,360 --> 00:06:19,680
One of his fellow PhD students also from Harvard,

85
00:06:19,680 --> 00:06:23,840
Albert Meyer, quite a name in theoretical
computer science but of course,

86
00:06:23,840 --> 00:06:26,880
again now, he'll have retired. He had one and

87
00:06:26,880 --> 00:06:30,639
the other one, if you like, was Dennis's
own working copy.

88
00:06:30,639 --> 00:06:36,800
which eventually got submitted - or was
about to be submitted. Normally, at this

89
00:06:36,800 --> 00:06:41,120
period, what you would do was handwrite
all your great thoughts and hand it to a

90
00:06:41,120 --> 00:06:44,880
[technically trained] typist.
The idea of being able to do it on a

91
00:06:44,880 --> 00:06:50,319
computer and with computer assistance
was at its very very infancy.

92
00:06:50,319 --> 00:06:54,800
We have a copy here of what it was like.
It could have been submitted [but] it didn't get

93
00:06:54,800 --> 00:06:59,520
submitted. But Dennis decided, quite rightly, that

94
00:06:59,520 --> 00:07:03,919
a Teletype terminal wasn't powerful
enough because it didn't have enough

95
00:07:03,919 --> 00:07:07,680
characters. So if you're going to either hand type it,

96
00:07:07,680 --> 00:07:11,840
or typeset it from a peripheral stuck into

97
00:07:11,840 --> 00:07:16,960
a computer, what were your options? Well, your options

98
00:07:16,960 --> 00:07:21,199
were either an ultra-expensive, very early,

99
00:07:21,199 --> 00:07:24,960
photo-typesetter [with] 
font characters as photos on a 

100
00:07:24,960 --> 00:07:29,280
film strip.
So, your a - z were 26 different little

101
00:07:29,280 --> 00:07:32,639
bits of film,
probably wrapped around a drum that you

102
00:07:32,639 --> 00:07:37,280
could illuminate from behind.
Massively, massively, expensive. Even Bell

103
00:07:37,280 --> 00:07:40,319
Labs hadn't invested in one of those at
the time.

104
00:07:40,319 --> 00:07:45,440
Although a few years later they did so.
If you come back from those

105
00:07:45,440 --> 00:07:48,560
rarified heights you could type it by hand

106
00:07:48,560 --> 00:07:52,800
and if it was an old-fashioned
typewriter you had special characters on

107
00:07:52,800 --> 00:07:55,440
little things that look like nail-varnish sticks [Typits]

108
00:07:55,440 --> 00:07:58,639
that you put into the throat of the typewriter

109
00:07:58,639 --> 00:08:02,800
and you hit any key. And it was like a sort of

110
00:08:02,800 --> 00:08:06,160
relayed impact. That alpha character [say] was
actually

111
00:08:06,160 --> 00:08:09,440
typed [as]  'i' or something, any character would do
And it hit

112
00:08:09,440 --> 00:08:12,639
the back of this special gadget and you got,

113
00:08:12,639 --> 00:08:17,759
at the right place, the letter alpha.
But going beyond that -- because that was a

114
00:08:17,759 --> 00:08:23,039
pain --  IBM came up with the idea of a 
golf-ball typewriter.

115
00:08:23,039 --> 00:08:27,039
[IBM video of Selectric in action]
"This is the best thing that's happened
to typing since electricity".

116
00:08:27,039 --> 00:08:32,200
"The IBM Slectric typewriter." They are
miraculous things,

117
00:08:32,200 --> 00:08:36,080
unbelievable. What they did was that,
instead of having

118
00:08:36,080 --> 00:08:39,200
26 different dipsticks [Typits] for weird
characters

119
00:08:39,200 --> 00:08:42,959
they would put put them on, let's say, a
Greek characters golf-ball

120
00:08:42,959 --> 00:08:46,560
[So] for special purposes you take out your
regular golfball, you plug in a Greek

121
00:08:46,560 --> 00:08:49,200
characters one.
When you look at the way that they

122
00:08:49,200 --> 00:08:53,839
[IBM] embedded those metal characters 
onto the surface of a sphere,

123
00:08:53,839 --> 00:08:57,760
in such a way that if the sphere rocked

124
00:08:57,760 --> 00:09:03,279
ever so slightly, 
another row of characters became

125
00:09:03,279 --> 00:09:06,720
as it were 'normal' to the paper and could be hit.

126
00:09:06,720 --> 00:09:10,399
>> Sean: With a typewriter you can have little
arms with a metal

127
00:09:10,399 --> 00:09:14,160
kind of thing that gets .... 
>> DFB: at the end and
they all end up in the

128
00:09:14,160 --> 00:09:17,200
neck, or the throat, or whatever it's called and

129
00:09:17,200 --> 00:09:21,440
on a conventional typewriter you move
the sheet around [to place the characters]

130
00:09:21,440 --> 00:09:27,040
With a Selectric you only move the sheet
for line feeds forward and back.

131
00:09:27,040 --> 00:09:29,760
You didn't move the sheet for horizontal motion(s).

132
00:09:29,760 --> 00:09:33,120
This unbelievable cradle with a golf ball in it,

133
00:09:33,120 --> 00:09:38,000
going seemingly at the speed of sound
was leaping all over, left and right,

134
00:09:38,000 --> 00:09:41,760
putting in characters.
It is an electro-mechanical marvel of the

135
00:09:41,760 --> 00:09:46,800
first order.
And it did not come cheap! It cost as much as ...

136
00:09:46,800 --> 00:09:50,800
well the story at the time was -- those of
you on the far side of the Atlantic can

137
00:09:50,800 --> 00:09:54,800
translate this better than me --
it cost as much as a Buick. Now a Buick

138
00:09:54,800 --> 00:09:57,600
is not an utterly luxurious saloon, I don't think,

139
00:09:57,600 --> 00:10:01,279
But it's not cheap!  And it certainly wasn't cheap

140
00:10:01,279 --> 00:10:07,120
in those days, So it [a Selectric 2741 peripheral]
cost as much as a medium-sized saloon car,

141
00:10:07,120 --> 00:10:11,839
just for this output device! And Dennis
having been taken on by 

142
00:10:11,839 --> 00:10:18,800
Bell Labs as the 'bright new lad',
actually conned them into letting him

143
00:10:18,800 --> 00:10:22,720
have an online version of one of these devices

144
00:10:22,720 --> 00:10:27,279
it's called a 2741 and [it also had] a 
dedicated [telephone] line that

145
00:10:27,279 --> 00:10:30,640
had a monthly rental at eye-watering prices.

146
00:10:30,640 --> 00:10:34,320
just so that, as well as working for Bell
Labs in his new job,

147
00:10:34,320 --> 00:10:37,839
he could also have a golf-ball machine down 
there [in the basement of the Ritchie family house]

148
00:10:37,839 --> 00:10:41,200
where he did experiments on how to typeset 
his thesis properly.

149
00:10:41,200 --> 00:10:47,760
And brother Bill (We'll point to his Web site -- 
https://dmrthesis.net) aged nine [Correction: twelve] 
years old, at the time

150
00:10:47,760 --> 00:10:51,600
was just in awe of Dennis sitting in their basement

151
00:10:51,600 --> 00:10:56,160
(in the Ritchie family house).
Dennis rented the basement, just

152
00:10:56,160 --> 00:11:00,160
swapping golf balls and messing about
with this thing [which was] in

153
00:11:00,160 --> 00:11:03,279
one corner of the basement, from what I
understand.

154
00:11:03,279 --> 00:11:12,160
But, even so, how did he, in the end,
manage 180 pages of thesis

155
00:11:12,160 --> 00:11:16,240
with changes of characters all over the place?

156
00:11:16,240 --> 00:11:19,839
>> Sean: Because we're talking about [essentially]
typesetting equations here aren't we? 

157
00:11:19,839 --> 00:11:23,360
>> DFB: yes, I'd probably better turn our attention

158
00:11:23,360 --> 00:11:26,880
to this one I think [points at equation on hard-copy printout]
which is pretty horrifying. I mean when

159
00:11:26,880 --> 00:11:30,399
you look at it there are the obvious alphabetic

160
00:11:30,399 --> 00:11:34,880
characters there.
But at the moment, inked in, is

161
00:11:34,880 --> 00:11:41,120
a Greek 'rho', or is it a 'p' symbol
in Dennis's own handwriting. And just the

162
00:11:41,120 --> 00:11:44,399
sheer density of this stuff, one [equation] 
on top of another.

163
00:11:44,399 --> 00:11:48,480
A funny thing there which is a 'monus' operator.

164
00:11:48,480 --> 00:11:54,160
So, this is horrifying to contemplate

165
00:11:54,160 --> 00:11:57,360
is: "How many golf-ball changes would you need

166
00:11:57,360 --> 00:12:02,079
to typeset that page?"  Because what you've
got to do is do all the easy stuff - and

167
00:12:02,079 --> 00:12:05,120
there isn't much easy stuff here - set it 
[the golfball carriage] back, in a

168
00:12:05,120 --> 00:12:09,040
sense, to a 'home' position and put the paper 
back in again.

169
00:12:09,040 --> 00:12:13,440
In other words, if it's on tractor feed,
make sure you get back to exactly the

170
00:12:13,440 --> 00:12:17,680
right start position, 
then go through it [the document] waving the golf ball

171
00:12:17,680 --> 00:12:19,920
around on its carriage and then deciding:

172
00:12:19,920 --> 00:12:23,680
"Ah!, I'm needed there. Let's type one
character there." So, you might end up

173
00:12:23,680 --> 00:12:27,760
having to do multi-pass processing on
just about every page, depending on complexity.

174
00:12:27,760 --> 00:12:31,279
Now some of you may remember we did

175
00:12:31,279 --> 00:12:33,680
- actually you may not even recognize it

176
00:12:33,680 --> 00:12:37,680
as a comparable exercise. About 6 or 7 years 
ago we did [on 'Computerphile] a thing about the

177
00:12:37,680 --> 00:12:41,680
Vacation Memo. It was a 13-page [Bell Labs] memo

178
00:12:41,680 --> 00:12:46,079
of the fights they [Bell Labs] had had with 
Mergenthaler, and with the [202 typesetter] hardware,

179
00:12:46,079 --> 00:12:50,079
to get everything working. But they
succeeded in the end. We effectively wanted

180
00:12:50,079 --> 00:12:53,760
to do, for Dennis's thesis, what we did in

181
00:12:53,760 --> 00:12:58,240
rescuing the Vacation Memo.
The trouble is it's not only ten times

182
00:12:58,240 --> 00:13:02,000
longer, is this thesis,  it's also easily ten

183
00:13:02,000 --> 00:13:07,760
times harder to do,
because his material is so complex.

184
00:13:07,760 --> 00:13:11,040
[For] the stuff on the 202 we were rescued by
the fact that

185
00:13:11,040 --> 00:13:17,760
fonts carried on beyond the 202 [into laser-printers] 
[and] into the great new world of Adobe Type 1 [fonts],

186
00:13:17,760 --> 00:13:21,920
Truetype fonts, Open Type fonts for
those of you into these things.

187
00:13:21,920 --> 00:13:25,680
There are standards.
You can miss out all the [electro-mechanical] 
stuff in the

188
00:13:25,680 --> 00:13:29,680
middle take a flying leap into the world of fonts 
And things become

189
00:13:29,680 --> 00:13:36,560
easier fortunately.
How do we make our flying leap here?

190
00:13:36,560 --> 00:13:39,920
Here is a font, if you
care to call it that, which basically

191
00:13:39,920 --> 00:13:45,600
exists on a golf-ball on a typewrite.r
Not a traditional conveyance mechanism

192
00:13:45,600 --> 00:13:48,079
for what a typographer might call a 'real font'.

193
00:13:48,079 --> 00:13:52,639
Nevertheless, our good friend Chuck Bigelow,

194
00:13:52,639 --> 00:13:57,920
typographer and giant of the Lucida typeface family,

195
00:13:57,920 --> 00:14:00,959
when advising us said: "i'll have a look around and see

196
00:14:00,959 --> 00:14:07,120
if somebody has seen fit to turn the Selectric font

197
00:14:07,120 --> 00:14:10,560
into what we would call a 'proper font' "
And he found for us

198
00:14:10,560 --> 00:14:16,639
from the Bitstream foundry, a font which
I think is called Pica 10 and it's

199
00:14:16,639 --> 00:14:20,800
pretty darn close to what a standard
Selectric golf-ball has got.

200
00:14:20,800 --> 00:14:25,839
But unfortunately nobody did the exotic
symbols [e.g. Script, Greek, Math - all
at fixed pitch].

201
00:14:25,839 --> 00:14:29,600
And the thing about the main running
text is that because it came off a

202
00:14:29,600 --> 00:14:33,920
typewriter [device] it's what's called 'fixed pitch'. 
Maybe a

203
00:14:33,920 --> 00:14:38,320
lot of you don't realize this, but
if you have a fixed-pitch typewriter

204
00:14:38,320 --> 00:14:41,839
have you ever noticed that the problem 
is that the letter 'i' [for example]

205
00:14:41,839 --> 00:14:45,440
looks very lonely because there has to be a
lot of space on either side of it

206
00:14:45,440 --> 00:14:49,040
The fixed pitch of these characters is 602

207
00:14:49,040 --> 00:14:53,040
one thousandth of an inch  
[Correction: thousandths of an em]
- OK for ordinary alphabets.

208
00:14:53,040 --> 00:14:57,120
But when you get extended characters
like a [lower case] Greek alpha,

209
00:14:57,120 --> 00:15:00,160
which has got a floppy head and a floppy tail

210
00:15:00,160 --> 00:15:05,040
they don't look very good at fixed pitch.
[Ideally] you have to redesign them but they look

211
00:15:05,040 --> 00:15:07,920
all constrained and not their usual florid self.

212
00:15:07,920 --> 00:15:15,120
There have been no *fixed pitch* redrawings 
as far as we can find

213
00:15:15,120 --> 00:15:20,079
[for creating] a typewriter-compatible font 
of the Greek Symbol golf-ball [say].

214
00:15:20,079 --> 00:15:24,480
It's a minority interest. So I've
had to do my best in

215
00:15:24,480 --> 00:15:29,279
rescuing Dennis's thesis; work is
ongoing it'll take quite some time yet,

216
00:15:29,279 --> 00:15:35,839
to say: "Well we'll have to use
a variable pitched typewriter-like

217
00:15:35,839 --> 00:15:40,240
font for the exotic symbols." 
So, for example,

218
00:15:40,240 --> 00:15:47,839
the [lower case] alpha symbol in Greek
requires a variable pitch of 890

219
00:15:47,839 --> 00:15:53,519
thousdandths [of an em] not 602.
And if you start to say: "Come on Brailsford stop

220
00:15:53,519 --> 00:15:56,560
being lazy! Just take the variable
pitched font

221
00:15:56,560 --> 00:16:03,519
and using your innate [type-design] skills
squash it down into being a fixed

222
00:16:03,519 --> 00:16:07,440
pitch version [602/1000] of it.
It's hard!  If you don't believe me read

223
00:16:07,440 --> 00:16:11,920
my semi-rant back in the Vacation Memo
paper it took me something like 

224
00:16:11,920 --> 00:16:16,079
120 - 150 hours to create a font called PrintOut and

225
00:16:16,079 --> 00:16:20,639
the results look rubbish unless you have artistic

226
00:16:20,639 --> 00:16:26,160
[type design] innate capability. Which i don't" 
It means that my lines on the [DMR thesis] rebuild

227
00:16:26,160 --> 00:16:30,079
are sometimes a bit longer than Dennis's
because i'm having to use

228
00:16:30,079 --> 00:16:34,959
quite a few variable-pitch characters
that are inevitably - they look more elegant -

229
00:16:34,959 --> 00:16:37,600
but they're longer [i.e. wider] than they would be in

230
00:16:37,600 --> 00:16:42,399
a fixed-pitch version. So under my left finger
[points at original thesis page]  is the

231
00:16:42,399 --> 00:16:45,680
stuff that Dennis left - his display equations.

232
00:16:45,680 --> 00:16:51,199
Swapping over quickly here [and] using the bit
stream Pica 10 font

233
00:16:51,199 --> 00:16:57,839
is my attempted rebuild of the same
material. I'm still working on it but can

234
00:16:57,839 --> 00:17:01,839
you see the 'pi' symbol there - it looks better

235
00:17:01,839 --> 00:17:05,919
There's all sorts of things. For those of
you saying: "Well how are you producing it?"

236
00:17:05,919 --> 00:17:09,760
I've allowed myself
the anachronism of going five years

237
00:17:09,760 --> 00:17:14,640
later in Bell Labs history.
Anybody who's into ancient UNIX would

238
00:17:14,640 --> 00:17:18,640
say that [this material] was yelling out 
for 'eqn' and 'troff'

239
00:17:18,640 --> 00:17:22,400
so  why didn't Dennis use them? 
They hadn't [yet] been invented !

240
00:17:22,400 --> 00:17:27,679
His thesis was in [late] 1967. 
Eqn and troff weren't there until 1972

241
00:17:27,679 --> 00:17:31,039
And [this material] it's got superscript and 
superscript positions.

242
00:17:31,039 --> 00:17:34,320
And you must respect those because those
affect the meaning

243
00:17:34,320 --> 00:17:38,000
of the equations. He was five years ahead
of his time and Lord

244
00:17:38,000 --> 00:17:44,640
knows how he did it but in following him
- may we be forgiven - we're giving

245
00:17:44,640 --> 00:17:48,559
ourselves five years
benefit and comforting ourselves that

246
00:17:48,559 --> 00:17:51,840
had Dennis done his thesis five years later, he

247
00:17:51,840 --> 00:17:56,480
would [almost certainly] have used these tools.
So we have used 'eqn' to set these

248
00:17:56,480 --> 00:17:59,360
equations feeding into the device independent

249
00:17:59,360 --> 00:18:03,919
[ditroff] version of 'troff'
because stylistically and date-wise

250
00:18:03,919 --> 00:18:06,960
they are earlier than things like TEX 
and LATEX, by quite a bit.

251
00:18:06,960 --> 00:18:10,720
So, this seemed to us to be the best

252
00:18:10,720 --> 00:18:14,240
reproduction compromise.