Programming
without programmers and what prevents its
development
A
mechanic, a chemist, and a programmer are driving in the car.
Suddenly
the car stalled.
Mechanic:
"Probably there
is something
with the motor."
Chemist:
"No, it's
a
bad gasoline."
Programmer:
"Raise my salary." © A
joke
Quarantine
and lockdown have forced businesses to go online: grocery delivery,
entertainment, shopping, meetings, games, chatting with friends,
and even sex - everything now happens on the Internet. Offline
businesses have intensified with vaccinations in the US and Europe.
Their development needs remain
relevant.
Accordingly, the already
considerable demand
for programmers
increased
greatly.
If
earlier they were recruited in third world countries
in
order to save money,
nowadays
the limits of this source have been reached. According
to djinni.co, the ratio between applicants
and vacancies has grown exponentially since November 2020, and not in
favor of recruiters. Business is
in agony:
there are simply no free hands and heads on the labor market.
You
have to lure
people
away
from their current employer,
offering
higher wage
(the Ukrainian Diya City was just made to enslave developers) If
the lure attempt was successful, uneasy one uneasy feeling will
change the other, because
tomorrow someone may
offer
even more to
the newly acquired worker...
What
can
be done in order to get out of
this unpleasant cycle of suffering?
Teaching programming to beginners is a task
with low time efficiency,
and there are no guarantees that a neophyte will not run away
to some other company that offered
him
a counteroffer with a higher monthly salary.
Is
it possible, at least in typical cases, to do without these
snickering, capricious,
and at the same time expensive bastards,
known as programmers?
Is
there something that can be adopted with the capability
to
create simple
programs without or with a minimum of code? Considering
that,
no-code platforms are gaining more and more popularity now. What do
they look like now, what will they look like in the future, what
hinders their development and why can't we refuse the services of
developers? I tried to answer the
questions below.
What
is no-code
When
robots become intelligent, their money will be called kilowatts.
No-code
nowadays
In
fact, no-code is now nothing more than a constructor. There are
components, they are linked using settings. The logic of the
application is created using code (such platforms are called
low-code) or drawn in the form of graphical diagrams (no-code). This
topic is not new.
Let’s
remember at least Delphi: by
applying
a
few components
to
the form, you can create a rather complex application.
You
can create a website, chatbot, document management system, ERP,
and much more in
the same way. Constructors for games look somewhat different, but
they still consist of an editor and logic. Examples of the
constructors
are Unity3D, Adobe Flex, and Unreal Engine. There are also accounting
systems: 1C, SAP,
and so on.
I see no reason why one cannot come up with a
system
for any other type of human activity. Question: does
the current state
of no-cone systems threatens
the profession of a programmer? The answer is no. Moreover, ABAP
(Input Language for SAP) programmers are considered to be among the
highest-paid.
And
no-code
in
the future
Astrolabe
... measures
by
itself, provided
you have
something to measure ©
Ilf
and Petrov. Twelve Chairs
As
I indicated above, at the moment, no-code platforms are not a threat
to the workplaces of IT specialists. But, let's say, over time,
artificial intelligence will develop.
It
will be able, for example, to create programs based on a text
description and
a drawing by hand. Will this be the end of high-paying IT
jobs
and the job seeker's labor market? Let's figure it out. First, how is
drawing fundamentally different from dragging and dropping controls
onto a form? The
only difference is
that for the former
you need to be able to draw with a mouse or with
a stylus. Secondly, the
free form of
textual description
will be limited because
it should describe the algorithm of the program's operation. Yes,
there are already experimental tools that allow you to convert
natural language into code, but even in
a
natural language,
you need to clearly describe what the program will do. It is quite
possible that knowledge of modern programming languages will
become a thing of the past, as has already happened with Assembler,
but the need to describe algorithms will not go anywhere in the
foreseeable future. It is the capabilities of a person, not
artificial intelligence, that stand in the way of progress.
People
simply do not know how to algorithmize. And as you
may
think,
the algorithm is far from just clever problems at the Olympiads. See
for yourself by
asking
your family members to write an algorithm for making coffee by
writing it down in words.
Granted,
that not everyone will succeed on the first try. On the other hand,
Lisp has been able for over half a century to create DSLs that are
virtually indistinguishable from a natural language. It should be
noted that progress in the
development of the tools
by
no means leads to a decrease in the number of programmers.
It
is in fact the opposite:
every year there are more computerization problems than enough
people
who know how to solve them. And there is
no
one
left
who is
willing to code
in
exchange for
"reasonable money". And the requirements
to developers
are only growing: in the early 90s, knowledge of a programming
language was enough to work, then the
knowledge of relational
databases was
needed, then programming web applications, and so on. Mobile
platforms, frontend, artificial intelligence
were added to this list
- everything without which modern IT is unthinkable. But what if you
really, really don't want to write code? Is it possible?
It
turns out, it is possible, because there are...
Flowchart,
flowcharts
everywhere
If
the logic cannot be written, it must be drawn, and, as a rule,
flowcharts are used for this. According to Wikipedia, a
flowchart
is nothing more than a common type of diagram
(graphical model)
describing algorithms
or
processes in which individual steps are depicted as blocks of various
shapes, connected by lines indicating the direction of the sequence.
In other words, a flowchart
is a set
of shapes
connected by lines with arrows. There may be something written inside
the shapes and on the arrows. It
seems to be nice,
but the devil is in the details.
They limit the application of the flowcharts
to
boring laboratory works
during your studying.
Reasons
behind the limited use of flowcharts
Let's
take a closer look at the disadvantages of flowcharts. We will use an
online editor as an illustration.
Lack
of mandatory structure
Flowcharts
can be drawn in any way: even from right to left, even from top to
bottom, even obliquely. No connection with the timeline, nor with the
flow of events. No wonder they are called "spaghetti". It
is almost impossible to understand these Inca writings.
The
standard allows the intersection of lines without semantic content,
even placing figures on top of others: you can draw an indecent word
with a flowchart and it will be quite valid. As an illustration, the
figures
above
are
flowcharts.
The figures and connecting lines represent the letters "F",
"U",
and “C”.
The fourth
letter
is left as an exercise for the reader,
in accordance with the measure of his licentiousness.
Lots
of empty space
Let's
see what a block diagram of a simple "for" loop looks like
(picture below):
At
the same time, in the C language, this block diagram looks like this:
Is
the block diagram clearer? No! Smaller? No! Why it
should be used then?
Redundant
complexity
See
how many shapes are offered for use in the popular online editor
www.lucidchart.com
For
each of them, there are rules of use described in hundreds of
textbooks, some of which also contradict each other. It is impossible
to understand all this without reading a few books. On the other
hand, after reading several books on the programming language, you
can already start writing code. At the same time, learning only the
flowcharts will not make you able to create programs.
As
a result, professional programmers usually do not use block diagrams.
Is there anything you can do about it? Yes. The disadvantages can be
overcome by introducing standards, restricting creative freedom in
such a way that only compact, clear and understandable schemes will
emerge from under the mouse or stylus. And such standards were
developed back in the USSR. By the way, nothing better has been
invented anywhere for the past thirty years, although there were some
attempts. One of these standards is R-charts
notation.
Introduction
to R-charts
The
R-chart
notation was developed in the USSR by Igor Vyacheslavovich Velbitsky,
Doctor of Physics and Mathematics, at the Kyiv
Institute of Cybernetics of the National Academy of Sciences of
Ukraine back in 1975, even before the birth of the author of this
article. Yes, with its help programs for missiles
targeted
at
the evil bourgeoisie
were created.
But it did not work out, since
the
USSR sank into oblivion, leaving us a legacy of the R-technology (and
much more). Firstly,
this language is graphic, and secondly, very simple, but this is only
at first glance.
It
has:
• points,
that contain no text;
• lines
that connect the points, which can be
horizontal or
vertical. You
can place
text on the
top and bottom
of these lines.
This
is enough to create programs of any complexity.
From
the theoretical
point of view, R-charts
are a directed graph
that is
loaded along arcs,
which means that
labels can only be on them, not on the tops.
This notation
is dictated by considerations of compactness: there is much more
space on the connecting lines than inside the points.
The
simplest program
Here
is the simplest program describing the behavior of a hungry person,
whose
actions depends on the available sum of money:
As
you can see, the condition is written at the top and the action at
the bottom.
Program
with transition
Suppose
someone went to the store to buy bread. If there
is no bread,
it doesn’t matter – there
should be some in the
next one. The
R-Chart describing this process is shown below.
This
is analogous to jumping to a line of code. Or at to
the
fork in this case. Some of the arrows do not have inscriptions above,
below, or at all,
which
is normal. In this case, it means an unconditional action, a jump on
a condition, or an unconditional jump.
Looped
program
Is
it possible in the previous case to make
a chart without drawing an arrow following
to the beginning? Yes, it
is possible by using
a loop:
See
the number
sign?
It means that this arc represents a cycle. But there can be more than
one action in a cycle. In
this case,
the R-chart will look like this:
At
the top of the arc marked with a cycle, we write a condition.
The
actions written on the arcs below it are repeated.
That's
all the theory
that should be mentioned.
Below I will develop the idea of R-charts in various
applications to real-life problems.
Advantages
over block diagrams
How
R-charts
are
different from dozens of UML block diagram
implementations?
Let's figure it out.
Clear
rules of construction and simplicity
R-charts
consist only of vertical and horizontal lines. This makes it
easier to read.
Compactness
R-diagrams
are loaded only along arcs, where there is a lot of space for
inscriptions, they can be used for programming from a small screen of
mobile phones.
As
an example, let's see how the for loop from the example at
the beginning of this article
would look like:
And
that
is it.
It
certainly takes less place than C code representation.
Using
R-charts
for No-Code Platforms
R-charts
are a good tool for declarative programming. It is applicable
wherever it is necessary to separate the algorithm itself from the
implementation of technical details, that is, what all no-code
platforms have been created for. Let's see how this technology can be
applied to no-code.
Chatbots
There
are literally hundreds of platforms for the
creation
of chatbots.
Typically they are
combination of
flowchart designer with
settings.
Below is a screenshot from the Landbot service. It shows that even a
small chat created for testing does not fit on the screen:
Let's
see how the same
chat
can be made
with
an
R-chart:
As
you can see, the circuit is much more compact. At the top,
we write the type of the element
and
its specific parameters at the bottom. For example, if you need to
display a message at the top of the line, we write "message",
and
its
text at the bottom. In the case of a choice of options, write
“Select” at the top and under this line the options for
selection.
Everything
is very simple.
Gaming
Platforms
Is
it possible to create a simple game without code? Yes! The
flowlab
platform
can fulfill this task.
This is a tool
that can be used
for creating 2D platformers. In it and in other similar platforms
there
are
game objects
that
have
the following characteristics:
• Object
type: floor, wall, robot, bush, character, and so on.
• Physical
characteristics: coordinates, speed, mass.
• Behavior
that depends on the physical characteristics of a particular object
at a given time and the game scenario.
The
toolkit for creating games, therefore, consists of two editors: a
graphical one for sprites and an editor for flowcharts. With the help
of the latter, the behavior of the characters is created. However, in
existing systems, the implementation of flowcharts
is extremely imperfect and ...
And
this is ... our favorite block diagrams with their shortcomings,
which can
be avoided
by the introduction of R-charts.
Let's
see how you can program the behavior of a spaceship in a scroller.
The spaceship
flies across the screen from top to bottom (in fact, it is stationary
relative to the computer screen - the game world is moving). The
spaceship shoots
if you press the spacebar. If the
spaceship
hits an obstacle, it explodes. Left
and right movements
across
the screen
are
made
by pressing the "A" or "D" keys.
Actually,
that's all. Those of you who have ever written games using, say,
Adobe Flash or Unity, know how much code it takes to create the
aforementioned logic. Spoiler for those who didn't -
a
lot
of code!
Robot
software
The
future has already become a reality, robots have already come
into our lives.
They
appeared in industries
long before the birth of the author. For example, you can create a
drawing
of
detail
in AutoCAD, then transfer it to a CNC machine, and the
detail
will be created without human intervention. But let's look at an
example closer to a mere mortal. Suppose we have a coffee maker and
it brews
the
coffee at a certain time on
weekdays.
On
holidays the
set time is different.
If
the day is marked as
a
vacation
on the calendar, the
coffee
maker will be inactive.
Let's see how this program will look in «R».
As
you can see, everything is so simple that it causes some feeling of
neglect.
But
this feeling
is
justified because
such
a
program can be written by any housewife.
So
what's next?
The
author
in
his spare time develops a constructor that allows using the R
language for a variety of applications. Development languages
are
TypeScript 2 + React.
Little
has been done so far
due to the
full-time
job and
non-work activities.
The
current
schedule leaves out
the
possibility to spend more time on the project. Therefore
enthusiasts
are invited for
cooperation.
This is not a startup yet –
rather an
open-source
development, but I plan to participate and win
the
grant. The
application for a grant is in an active phase.
p.s.
Follow the link for my interview
about
no-code platforms to Taras Luxurious, Vice President of the IT Guild.
Сообщения
