Pythonneries - Making Of 8

Hello,

I believe that the sets example at the end
of the previous video where I was switching

shapes and images illustrates rather well my philosophy:
I like what doesn’t show. Nobody has ever

commented on Youtube on my use of
Powerpoint, and I believe that it’s

because few people see it as Powerpoint
and can imagine that in a lecture or seminar

it looks exactly the same – in fact, slides
from my classes or seminars end up on

Youtube and vice-versa.
When your goal is to make your presentations

lively, the key is to see movement
as an integral part of the story, and

not as a gimmick to try to
instill life where there is

none. The best proof of it may
be a counter-example.

I have attended long ago in California
an IT presentation by a speaker who

was young and enthusiastic, had a sense of humor,
and knew very well what he was talking about.

He made only one mistake, but a capital one:
he chose to have random transitions

between his slides, and it looked
a bit like this. Even he seemed surprised.

It should have been one of the good talks
at this conference, and all I can remember

is this major mistake of random transitions.
You could see nothing but them. In fact it

remained a trauma for me, so much
that I didn't use transitions for many

years.
My standpoint changed one day when

I wanted to talk about networks, and explain
that in a computer network, at least with

the TCP/IP protocol, everything is sent
as small packets, about two of them for a

single text page, and when you want to send
something big, it’s sliced at one end

and reassembled at the other end.
To represent the network, a cylinder

shape, which I rotated and colored
with a gradient to make it look more

like a “pipe”. Something big?
Let’s not fear clichés, I found

a fun elephant clipart, that will do.
My goal was to make the elephant appear

at the other end, and to have an
“exit” animation for the elephant on

the left and a simultaneous “entrance” animation
of the same type for the elephant on the right

that would carry the deassembling/reassembling idea.

And I was stranded without any idea. Fade?
I doesn’t work, the message doesn’t

become thin air.
Dissolve, it doesn’t work either,

it looks like pixie dust. Checkerboard
was a bit better, but it misses the

randomness that I was looking for,
as packets don’t necessarily arrive

in order.
I can’t remember why I tried transitions,

but I’m going to show you with
two Rembrandt self-portraits,

painted 40 years apart, what I discovered:
the “dissolve” transition wasn’t at all

the same as the “dissolve” animation.

And I said to myself: “but what would happen
if instead of having both my elephants on

the same slide I put them on two different
ones with the second slide dissolving

into the first one?”.
Well, I’m going to show you. Here is

what it gives, and it’s exactly the effect
I was looking for.

What surprised me when I tried it is that
the pipe isn’t affected.

In fact, the transition only shows on
what isn’t on both slides.

This isn’t surprising with no transition
or a fade; but I suddenly understood

that there were other transitions that
could be seen not as the switch from one

slide to the next, but as a particular type
of animation. Their characteristic is that

if two successive slides have
common elements when this transition

is set on the second slide,
nothing that is common will

be affected. In fact, if I had two
successive identical slides, I would

see nothing at all.

I didn’t know how to name these transitions,
and I suggest using an adjective

borrowed from mathematics to call them
idempotent transitions.

In maths, an idempotent element changes
nothing to the result of a given operation,

such as 0 for addition or 1 for multiplication.
An idempotent transition is a transition

that is invisible between two identical slides.

Idempotent transitions allow in a number
of cases using indifferently animation

or transition, and there are numerous
practical consequences.

For instance, suppose that for once we
have to mention three points.

That might be the presentation agenda,
or a summary of the most important

points. Obviously, core principle,
we aren’t going to dump the slide

and let everybody read it before we get
a chance to talk. The natural move

is using animations. We are going
to select the text, and in the animation

menu select “appear”. We might as
well choose “fade”. Except that

of course we don’t want to see everything
appear at once. We are therefore going

to “reorder”, select the animation

and in the text animation options say that
we don’t want to see everything appear

at once but by paragraph. This time, every
click will make a paragraph appear and

we’ll be able to comment it. There is
an even better option that puts the focus

on the current paragraph. In “effect options”
I’ll ask to dim after animation.

I can choose a color, so that

past text remains there but kind of
fades into the slide.

Here is what it gives, and there are two issues,
not without a solution, but which make matters

more complicated: first of all, by default
dimming applies to all paragraphs, and

I feel like a moron when everything is dimmed.
I don’t want to dim the last paragraph,

after it I want to directly switch to the
next slide. The second issue, is what

to do if we want for instance to make
a new image appear next to the

current paragraph?
All of this can be solved far more easily

with idempotent animations.
What shall I do?

I take my slide, without any animation,
and copy it as many times as I have

paragraphs. On the last slide, the color
of every paragraph but the last one is

changed to the dim color, and on previous slides
I remove what follows the current paragraph,

and set the previous ones to the dim color.
To make it better, one different image

each time. To every slide but the first one
I associate an idempotent transition. Very

easy to achieve, far easier than animation,
and when I play it the result is

flawless.

For all the available transitions,
forget about the headers, “subtle”,

“exciting” (personally I need more to
get excited) or “dynamic content”.

What matters, is idempotent or not -
will the transition merge into my

story and bring continuity, or will
it indicate a break.

I have tested everything that was available
in my Powerpoint version; in logos,

B indicates the current slide and A the preceding
one. I have checked with green everything

that is idempotent, and with red everything
that isn’t. Let’s keep aside “push”

that isn’t really idempotent but is
in a category of its own and that I’ll

talk about shortly. None of the last transitions,
which, for most of them, I never use,

is idempotent.

It’s obvious that when you look that way
at transitions, the “slide” concept

becomes rather meaningless, and I prefer
substituting to it a “sequence” concept,

an indefinite number of slides that flow
through idempotent transitions.

I have already talked about it in the
3rd video in this series, idempotence

allows maintaining continuity and therefore
staying with the story. For me, the famous

concept “one slide, one idea” means
nothing and I believe that what is interesting

is “one sequence, one idea”, knowing
that several illustrations of the

idea, or looks from different angles, can
occur in the sequence.

What about transitions that aren’t idempotent?
Many of them don’t look really necessary to me,

but some of them are useful for indicating
major breaks if several unrelated topics

are covered, or for a parenthesis. For smaller breaks,

an idempotent transition (and preferably no
transition) will do the job if

all elements on the slide change.

The typical major break is for me,
in a seminar that is over half a day

or more, the break after 60 to 90
minutes. If the seminar organizer

has done things well, that will be
coffee and cookies, otherwise it will be

around the coffee-machine ...
Usually I pair highly visible transitions

with reverse moves.

I mentioned previously that I would elaborate
on the “push” transition, which is rather

peculiar. Nancy Duarte is very fond of “push”,
and I’ll try to illustrate what she does

with it. Let’s start with a Swedish flag.
Why a Swedish flag?

Because of the “Nordic cross” that is common
to all Scandinavian flags and goes from

edge to edge, you are going to see very
soon what I liked in it, and because contrary

to Norway or Iceland there are only two colors,

so it’s simpler, and these two colors are
better for a Powerpoint presentation than

those in the Finnish or Danish flag. If
you are in Sweden and move up North,

far, far away, you arrive to the North pole.
Let’s head back southwards, then westwards,

and we meet Norwegian vikings. Sweden had
vikings as well, but contrary to the Danes

and Norwegians they were heading towards
Russia and Byzantium more than Great-Britain

or France.
What do you think you saw?

Something larger than the visual scope
of the presentation, on which we have

displaced a window.
What have you really seen? Four different

slides, two of them duplicated, with
“push” transitions in-between,

first from top, then from bottom

(which is the default) and finally from
left. There is a strong feeling of

continuity, reinforced by the graphical
continuity of the Scandinavian cross. The effect

wouldn’t have been as good with the Swiss flag,
for instance.

On the topic of flags, I’d like to underline
that, if instead of using a Swedish flag

as background I had used a French, Belgian
or Italian one, pushing up or down

would have been an idempotent transformation.
And pushing left or right would have

been an idempotent transformation
with a German, Dutch

or Polish flag for instance.
The “push” method is a very elegant

way to present much information as consistent

chunks while having something that
remains legible. To present program code,

a mathematical proof or a complex
industrial process, it’s an excellent

solution. By the way, I don’t know if you
have noticed it, but I have just been using

it to explain how I had done the Swedish
sequence, which was a 6-slide sequence.

Showing the 6 slides on one was too small,

with too much emptiness on the right. So I have
shown the first half of the sequence on one

slide, push up, and second half of the sequence.

One last point, just to remind you that everything
is linked: if you choose for your slides a fancy

background, “push” will look bad. And if
your background is a linear gradient,

pushing in the same direction as the gradient
will look bad; however, it will work fine if

you push perpendicularly to the
gradient. One more reason to favor

plain backgrounds, which allow everything.
The next time, I’ll discuss animations

proper.