00:00:00.199
so I have the honor to announce
00:00:02.560
Alexander who will talk about laziness
00:00:04.720
driven development and we'll open up the
00:00:06.879
very first H the the uncom for today
00:00:09.280
with his talk so give him a warm
00:00:11.799
Applause
00:00:18.560
please so hello everyone and as announed
00:00:21.560
today we're going to talk about laziness
00:00:23.480
driven development but uh at the
00:00:25.840
beginning I need to start my timer so I
00:00:28.400
know when how much time I spend
00:00:31.119
um and we do a round of introductions so
00:00:33.960
you know better uh who I am my name is
00:00:37.480
Alex uh I'm not really active on social
00:00:40.680
media you could always send me an emails
00:00:42.800
I'm an old school guy but I have a
00:00:44.840
profile on
00:00:46.039
GitHub and my personal homepage um I
00:00:49.960
have uh 20 plus something years of
00:00:52.359
professionally uh programming experience
00:00:55.520
which like professional means that I was
00:00:57.879
paid for this offer that I was uh Crea
00:01:01.800
currently I work at bab for the last 6
00:01:04.360
years uh and my current position is
00:01:06.799
principal uh software engineer which
00:01:09.200
should mean that I know something about
00:01:11.520
making
00:01:12.720
software uh and what is Babble bubble uh
00:01:15.720
is an ecosystem of different products
00:01:18.119
that help uh people people uh to learn
00:01:21.240
languages and behind the scenes it's
00:01:24.320
more than thousand of people uh working
00:01:27.439
in in two different offices the one in
00:01:29.119
Berlin one in in New York and as you
00:01:31.759
might guess uh bab is looking for uh
00:01:34.520
Engineers Ruby included so please check
00:01:37.560
this uh this this
00:01:40.040
URL so uh now back to the main topic and
00:01:44.799
it's a laziness driven
00:01:46.840
development um if we talk about laziness
00:01:50.399
I could imagine that we might have
00:01:52.240
different definitions of what it is so I
00:01:55.439
thought it would make sense to uh stick
00:01:57.680
to the one which is close to the
00:02:00.119
laziness D development itself so
00:02:03.000
definitions um the Cambridge Dictionary
00:02:05.880
says that the laziness it's basically
00:02:08.520
our quality that um kind of forces us
00:02:11.879
not to do any work at all right this is
00:02:14.959
very common understanding of what
00:02:16.519
laziness is and this is not this type of
00:02:19.120
laziness we're going to talk today there
00:02:21.599
is another definition of
00:02:23.599
laziness given by Pearl uh glossery and
00:02:27.519
it's a very long definition but uh in
00:02:30.080
summary uh it's the quality that
00:02:32.480
actually uh kind of forces you to work
00:02:35.680
more uh uh and Achieve great better
00:02:38.800
quality of the whatever you you're
00:02:40.640
producing in order not to do work later
00:02:43.599
on this that's the laziness we're going
00:02:45.879
to explore today and that's the laziness
00:02:48.080
I highly recommend to possess at some
00:02:51.440
point so
00:02:53.959
um rules right so rules of laist driven
00:02:57.959
development uh rules number one one is
00:03:01.080
avoid manual work uh you try to come up
00:03:05.920
with the with some with some tooling or
00:03:08.200
with some with some processes that help
00:03:10.760
you not to have uh manual work you also
00:03:13.080
learn how to identify what is so
00:03:15.440
sometimes doing some manual work is fine
00:03:17.040
I'm not saying that you have to avoid it
00:03:19.640
completely
00:03:21.519
uh the the one example is testing right
00:03:24.920
if you have to run test one time it's
00:03:27.239
okay but when it's already thousand
00:03:29.040
times the same
00:03:30.400
uh thing try to automate it right we all
00:03:32.959
love uh writing test for all our
00:03:36.400
applications that's one of the the Ruby
00:03:38.920
strongest uh things um but in order to
00:03:42.360
automate tests uh or testing of of our
00:03:45.400
software uh the software needs to be
00:03:47.799
designed in in a way that it's testable
00:03:50.959
right uh all possible edge cases
00:03:53.599
especially uh they should be covered
00:03:56.239
with tests so we don't have
00:03:58.280
bugs so that means when you design your
00:04:00.760
software and you don't want to run test
00:04:02.400
or you do manual testing with a button
00:04:05.000
clicking or something you try to design
00:04:07.680
uh the code in a way that software could
00:04:10.360
do testing for you uh your routines uh
00:04:15.120
right it's if you have to repeat the
00:04:17.440
same thing all the time let's say when
00:04:19.400
when you deploy something uh or when you
00:04:22.079
do a setup of the project that you just
00:04:24.720
W from GitHub uh try to see what are the
00:04:28.160
things uh that could be scripted and
00:04:31.880
automated uh and again similar to
00:04:34.600
automated testing right these routines
00:04:37.320
they should be designed in a way that
00:04:39.240
they actually scriptable right you could
00:04:41.039
come up with some uh conventions like
00:04:43.759
naming conventions or conventions where
00:04:46.120
you put files uh on on the disk so then
00:04:50.240
you could write scripts and then you run
00:04:51.960
one line uh and everyone is happy and if
00:04:55.000
you work in a project and you're not the
00:04:56.639
only one there and you have uh teammates
00:04:59.120
or you have New Jers from time to time
00:05:02.080
they will be definitely happy so they
00:05:04.680
don't have to run a thousand commands or
00:05:06.479
look in in some uh source code to just
00:05:09.039
run one command and everything works for
00:05:10.800
them and they they could be
00:05:13.919
lazy for the next couple minutes so
00:05:18.639
that's the summary of avoiding manual
00:05:20.919
work doing manual work sometimes it's
00:05:23.840
fine but if you see the pattern that you
00:05:26.080
have to do something multiple times uh
00:05:28.919
look for ways to automate
00:05:31.039
this avoid answering same questions so
00:05:35.720
it is okay to talk to people I'm not
00:05:37.720
saying that you don't have to answer
00:05:39.880
questions but try to uh look for
00:05:44.000
patterns when the same questions about
00:05:46.080
your software comes again and again uh
00:05:49.880
maybe there is a way uh to fix this uh
00:05:53.199
because you don't want to spend your
00:05:54.800
time answering the same question all the
00:05:56.479
time uh the one way or one of the best
00:05:59.680
best ways to avoid doing something is to
00:06:02.360
simplify the thing that to to the point
00:06:05.039
that it it is not possible to have
00:06:07.599
questions right everything is so simple
00:06:10.080
uh so people just don't have
00:06:11.880
questions it's always not always uh it's
00:06:14.840
not always possible uh that's why from
00:06:17.680
time to time we document our code right
00:06:19.919
we try to help people to find answers
00:06:23.759
themselves and the better you document
00:06:26.080
the less questions you
00:06:27.680
have um the next idea is to rely on
00:06:31.680
standard tool sets and the standard is
00:06:33.759
in quotes here in purpose um because
00:06:36.919
standard might mean uh different things
00:06:39.199
in different companies in different
00:06:40.639
teams in different projects but it means
00:06:43.599
something that people don't have to uh
00:06:45.919
don't have to learn right uh because
00:06:48.880
sometimes people don't want to learn so
00:06:50.199
they come to you and say oh I see that
00:06:52.400
you using this this thing how to use
00:06:54.520
it so uh and when it's multiple times uh
00:06:59.720
maybe you would like to switch to
00:07:01.080
another tool which most probably people
00:07:04.120
know how to use and they don't come to
00:07:06.800
you but still talking to people is
00:07:09.400
awesome you have to use
00:07:11.360
this so uh I avoid Reinventing the wheel
00:07:16.360
um this is one of the things that I
00:07:18.800
personally struggle as a software
00:07:21.160
engineer because I love making stuff uh
00:07:24.440
and sometimes when I see something uh
00:07:26.599
like a software tool I say I could build
00:07:28.840
it better right I could make it uh I
00:07:31.360
don't know 5 or 3% better um personally
00:07:35.759
I watch for this for this science this
00:07:37.400
is not not a good science in my case I'm
00:07:39.720
not I don't know about you but um to
00:07:43.599
avoid uh Reinventing the whe because
00:07:45.800
this is the price that you pay next it's
00:07:47.879
of course it's fun to build these things
00:07:49.879
but then you have to maintain them uh
00:07:52.039
and if if it's if the software that
00:07:53.919
you're working now it's just for you and
00:07:55.520
you're the only user do whatever you
00:07:57.560
want uh these rules and the laziness
00:08:00.400
even development development uh is
00:08:03.360
mostly about when you work in the team
00:08:05.039
when you work in the company uh or the
00:08:06.960
project invols more than one person
00:08:09.479
that's why to avoid Reinventing the
00:08:11.479
wheel uh try to use common uh common
00:08:14.280
tools right so people know how to use
00:08:16.800
them and they most probably already uh
00:08:20.080
bulletproof and then there's a history
00:08:22.199
and then all the possible buxs they
00:08:23.759
already discovered so you don't have to
00:08:26.120
do this
00:08:28.120
um Rel on well-known uh protocols and
00:08:32.080
patterns uh again try not to invent yet
00:08:35.760
another or protocol or or GTP or
00:08:38.479
something like this it's great to be uh
00:08:41.159
to be creative uh that's why we have all
00:08:44.240
these protocols but sometimes we have
00:08:46.000
too many of them I think uh so if you
00:08:48.800
really feel the passion that you would
00:08:51.080
like to build uh to to to create a
00:08:53.560
design yet another protocol look around
00:08:56.120
most probably it's already invented but
00:08:58.959
you could still contribute and make this
00:09:00.920
thing better and not uh invent another
00:09:05.120
one so um and again try to uh have as
00:09:09.800
less
00:09:11.040
customizations uh as as possible because
00:09:13.480
then you have to maintain them and uh
00:09:15.880
having something like half per uh
00:09:18.200
different than something else it means
00:09:20.160
that next time when there is an update
00:09:22.720
uh for this for the for the standard
00:09:24.480
tool you have to do extra work uh and as
00:09:28.240
a lazy person you don't want to this
00:09:32.160
right so um this one might be
00:09:36.440
controversial again as talking to people
00:09:38.839
and answering questions I'm not
00:09:40.839
suggesting that you have to uh
00:09:42.680
completely avoid third party
00:09:44.760
dependencies in your projects but you
00:09:47.360
always have to question whether it makes
00:09:49.680
sense or
00:09:50.839
not and there is a good side uh behind
00:09:53.959
behind this so I'll try to prefer uh
00:09:56.880
simple libraries over complicated
00:09:59.480
Frameworks because you need to you need
00:10:01.839
to spend time to learn them and and
00:10:04.959
understand them uh there's a time to
00:10:07.120
support them as well and you're not the
00:10:09.480
only one again uh every new Joiner on
00:10:11.839
your team or your company will have to
00:10:13.560
go through this through this process so
00:10:16.200
that's why uh try to uh if you make
00:10:19.640
decision to take some some third party
00:10:22.320
uh library or framework on board make
00:10:25.040
sure that it's actually uh worth it
00:10:27.600
maybe there is something that could do
00:10:30.000
the same could produce the same results
00:10:32.920
but it's inside of the pro language that
00:10:35.720
they use like Ruby Ruby has a great uh
00:10:38.720
core Library so you could build a lot of
00:10:40.760
stuff just using
00:10:43.079
this and uh again try try to learn and
00:10:47.600
understand what's already comes with the
00:10:49.399
language itself uh in many many cases uh
00:10:53.519
you could find something and I saw
00:10:55.120
examples when people just don't know
00:10:56.720
that something exist in in cor Library
00:10:58.760
one of the devices that I use uh and
00:11:01.000
then give to uh to to my colleagues that
00:11:03.920
every time when I don't know how to do
00:11:07.000
something or I found a new method in in
00:11:11.279
car Library I just look around like
00:11:14.120
within the same module I just read
00:11:15.639
documentation just to get an idea what
00:11:17.399
else is possible and sometimes you could
00:11:19.279
find diamonds there and if you have time
00:11:22.440
look in the source Cod it's also awesome
00:11:24.760
but time box
00:11:27.519
this so uh
00:11:30.360
um I don't want to compete with
00:11:36.160
eile that's why I have a comment on this
00:11:39.000
one try to avoid magic uh what it means
00:11:42.680
um it's there is no magic if you
00:11:45.959
understand things right uh that's why
00:11:48.920
rails has a lot of magic and the magic I
00:11:51.200
mean something that is a uh a method
00:11:54.040
that uh does a lot for you and it's
00:11:56.240
awesome right you don't have to do it
00:11:57.959
hides all the complexity
00:12:00.160
um but you have to kind of dissolve this
00:12:03.240
magic and look inside so you understand
00:12:06.040
so uh if sometime there's a fire on
00:12:10.440
production uh this magic could bite you
00:12:12.959
the magic means that you don't know what
00:12:14.399
it is right you don't know uh what's
00:12:16.920
behind so it's better to know and then
00:12:19.279
there's no magic this is how you avoid
00:12:20.880
it you learn you de inside um and you
00:12:25.000
dissolve the magic this this time so uh
00:12:28.199
in my opinion so the second best cot is
00:12:31.160
the one that you understand the one
00:12:33.399
which is not magical to you uh and if
00:12:36.360
you don't know the best code is the one
00:12:38.560
which does not
00:12:40.120
exist so the one which you don't have to
00:12:42.639
maintain right you don't have to learn
00:12:45.160
and you have to answer questions on this
00:12:47.320
that's like the the pure goal of
00:12:48.839
laziness driven
00:12:50.560
development but it's not possible yet so
00:12:55.480
summary and I spent just 12 minutes so
00:12:58.480
we'll have uh some extra bonus uh
00:13:01.880
laziness driven development it is not
00:13:04.360
about avoiding doing work it's not this
00:13:06.920
this cambrid definition of being
00:13:08.880
laziness and just sit on the couch and
00:13:10.639
watch Netflix it is not it's actually
00:13:12.680
about uh making better things uh like
00:13:16.320
with things with the with the greater
00:13:18.240
quality uh that you don't have uh spend
00:13:21.360
time on on the maintenance uh or you
00:13:23.680
could spend this time that that you uh
00:13:25.600
that you get this this way to do on
00:13:28.000
something more awesome some possibly
00:13:30.160
software spending time with your family
00:13:32.240
something this type of
00:13:34.160
laziness
00:13:35.720
so and uh as we have time I could uh
00:13:39.480
show you an example how this mindset and
00:13:42.480
the Las development and this rules also
00:13:46.519
um they are not super strict uh it's
00:13:48.880
basically a mindset that you apply so
00:13:50.959
every time when you have a problem you
00:13:52.639
apply you try to uh get this this
00:13:55.759
feeling I don't want to do this uh and
00:13:58.720
there might might be reason why you
00:14:00.000
don't want to do this this this laziness
00:14:02.480
speaks to you uh and uh use this use
00:14:06.199
this feeling of being lazy and and not
00:14:09.199
wanting to do something uh to identify
00:14:12.320
how you could do it better how you could
00:14:14.399
automate it or maybe how you could
00:14:16.279
change design of your software uh that
00:14:18.839
you don't have this problem anymore
00:14:20.360
that's that's the best case um so listen
00:14:23.480
to your laziness in a productive way of
00:14:25.880
course so uh task right and let's do an
00:14:30.279
example and talk about the task task is
00:14:33.279
very simple it's it's epimeral just to
00:14:35.959
give you an idea how this uh mindset
00:14:38.240
could be applied let's say you have a
00:14:40.440
task to create a tool uh could be common
00:14:42.759
line U interface like a CLI Tool uh that
00:14:46.560
gets data from from files uh in some uh
00:14:50.279
directory on a disk it passes these
00:14:53.480
files um does some processing and then
00:14:56.560
output results pretty uh common thing so
00:15:01.360
solution that I observe talking to to
00:15:04.680
other people and doing this this kind of
00:15:06.480
stuff myself these are the steps right
00:15:08.560
this is what needs to be done in general
00:15:10.560
right you iterate our files you read uh
00:15:13.120
each file you read data uh do processing
00:15:15.839
and then you output very simple uh if
00:15:20.000
you if you pay attention to the steps uh
00:15:23.040
the paring data and processing that's
00:15:25.360
the core of our task this is the core of
00:15:27.199
the tool that we need to build
00:15:29.639
uh without this like nothing nothing
00:15:32.120
makes sense but these other three things
00:15:36.360
they look to me so when I apply uh this
00:15:39.360
laziness driven development mindset I
00:15:42.000
see that these things they're quite
00:15:43.480
common right if the pro data processing
00:15:46.360
could be very unique to my uh to my tool
00:15:49.480
it's sometimes hard to find something
00:15:51.440
that uh does it for you but you have to
00:15:54.519
there that's important so you don't have
00:15:56.120
to build this tool that would be ideal
00:15:57.759
if there's something which does exactly
00:15:59.639
this just use it uh but let's say there
00:16:02.480
is there is uh nothing existing yet so
00:16:05.440
you have to build it let's um explore
00:16:09.079
these three things which are creating
00:16:10.920
now there something which sounds to me
00:16:13.120
like extra or external stuff
00:16:17.319
uh so we could of course and then people
00:16:20.959
do this that's the first thing that that
00:16:22.600
comes uh to our mind um let's have a
00:16:26.839
dysfunctionality the the uteran of files
00:16:29.240
reading them uh and writing output as
00:16:32.480
part of our program let's say it's a
00:16:34.600
script it's a very simple thing it's
00:16:36.199
just a ruby script uh then U pretty soon
00:16:41.519
you discover that some files might have
00:16:45.240
some some like different permissions you
00:16:47.600
could not read them or they don't exist
00:16:49.480
the directory does not
00:16:51.000
exist um or something something like
00:16:54.920
this uh and then you have to support
00:16:57.519
this in your script right you have to
00:16:59.240
add some reaction the script should be
00:17:01.120
user friendly right it should tell you
00:17:03.040
what needs to be done of course it could
00:17:05.000
just explode but this is not the nice
00:17:07.559
thing because people come with questions
00:17:09.400
and you want to avoid
00:17:12.160
this right uh or if you want to avoid
00:17:15.679
them then you you do this this exception
00:17:17.600
handling but then you have to think uh
00:17:20.360
how to explain people and and and so on
00:17:23.360
uh in your script so we could avoid this
00:17:26.839
uh we could kind of offload this task to
00:17:29.720
something which is standard right uh you
00:17:32.480
could read files and then directories
00:17:35.200
using standard so uh operating system
00:17:38.720
tools uh Ruby is great so it has uh
00:17:42.679
really nice interfaces three from stand
00:17:45.120
on input so the the OS level tools uh
00:17:49.559
like cat uh could read this file for you
00:17:53.480
uh fit it to standard input and then you
00:17:55.960
you just get it from there if there is a
00:17:59.080
problem and file is not readable because
00:18:01.480
uh not enough permissions to access this
00:18:03.919
file there will be of course exception
00:18:06.400
uh but because use standard tool most
00:18:08.440
probably people know already what this
00:18:11.440
uh error mean because they use uh this
00:18:14.080
this these tools so you don't have to
00:18:15.640
explain them you don't have to do this
00:18:17.400
this handling uh in your
00:18:19.799
tool um and the same when you do an
00:18:23.440
output uh of the the the results of of
00:18:26.799
data processing to standard output and
00:18:29.440
not writing this into files or some some
00:18:32.799
path on the dis from your script uh
00:18:36.000
because again um someone might provide a
00:18:38.799
pass but it might not be uh writable
00:18:41.440
right and again it's the same the same
00:18:43.159
story you have to explain uh you have to
00:18:45.480
do error handling just print it to
00:18:48.240
standard output the benefit number one
00:18:51.080
uh for yourself is that you see the
00:18:53.039
output you don't have to read the file
00:18:54.640
you run the script and you see it right
00:18:56.120
away on the screen uh if the the thing
00:18:58.480
does uh uh processing correctly but it
00:19:02.320
also gives a freedom uh for the users of
00:19:05.120
the script to do maybe some post
00:19:07.240
processing or put files where they want
00:19:09.760
to put them uh and it's very simple and
00:19:12.400
again it's already OS level it's kind of
00:19:14.880
out of scope of your application which
00:19:16.799
is great and if there's any problem uh
00:19:20.120
that the OS will tell uh uh about these
00:19:23.240
problems and people know about this the
00:19:25.880
this this errors already so they will
00:19:27.880
most probably handle
00:19:30.039
them uh as a grade developers we do
00:19:34.600
extra steps and the extras that you
00:19:36.880
could add uh to to the tool is to
00:19:40.720
provide uh in RM some copy and pastable
00:19:44.600
examples uh how to run your tool right
00:19:47.400
uh again using this this uh the OS uh
00:19:51.280
utilities uh you could say this is how
00:19:53.440
you read files from this directory and
00:19:55.080
then and fit uh to my tool or this is
00:19:57.159
how you could store them uh and if it's
00:19:59.559
on GitHub you have this nice little uh
00:20:01.679
copy uh button right so you just people
00:20:05.480
click this button and then run your
00:20:07.799
script uh you could add a built-in help
00:20:10.799
so people don't come with questions to
00:20:14.600
you uh and is very additional bonus
00:20:19.000
extra thing um you could write this um
00:20:23.440
this this this tool in the language
00:20:25.080
which compiles into binary um and
00:20:29.240
then for the users of your tool there is
00:20:31.880
no need to add uh Ruby is tricky
00:20:34.400
sometimes that you need Ruby itself and
00:20:36.440
run time um and if it if if this tool is
00:20:40.480
just uh if there is no necessity for
00:20:43.520
people uh for users of your tool to have
00:20:45.919
Ruby around uh something which compiles
00:20:48.159
to Binary is better and I've heard many
00:20:50.880
many from many people that Crystal is
00:20:53.400
really nice for rubby Ruby developers
00:20:56.480
Crystal language I mean
00:20:58.919
so uh profit right you don't need to uh
00:21:03.520
remember how to explain uh and and use
00:21:06.679
the tool you have read me for this uh
00:21:09.080
many this this file operations they
00:21:11.120
already on OS level they're not part of
00:21:13.320
your uh of your tool of your software
00:21:16.480
which means you don't have to test this
00:21:17.960
you don't have to Pro provide
00:21:19.120
documentation for
00:21:20.679
this you don't have to maintain it uh in
00:21:23.520
the future um so and then it also gives
00:21:27.480
you this greater f exibility right uh
00:21:30.039
you give the the users uh in a common
00:21:34.000
line interface of course example you
00:21:36.000
give them flexibility to decide uh where
00:21:39.720
the files are uh and then how they um
00:21:43.640
how they how they tool receive these
00:21:45.360
files and also how these files are
00:21:47.919
stored so it's kind of profit for for
00:21:50.679
everyone and not just for you who is
00:21:52.559
trying to be lazy and then do some other
00:21:55.960
software development so thank you very
00:21:58.720
much we have 9 minutes left so if I'm
00:22:02.480
allowed uh then I'm super happy to
00:22:05.000
answer any
00:22:18.559
questions just raise your hand
00:22:23.120
and hi um can you elaborate a little bit
00:22:27.039
on the question how to decide between
00:22:30.799
don't reinvent the wheel and don't use
00:22:33.000
third party dependencies so for example
00:22:35.440
if I'm trying to build a system and
00:22:37.240
there is a third party dependency that
00:22:40.400
does exactly this um for example yeah
00:22:44.279
like uh one possibility for you in your
00:22:47.400
example would be to use a library that
00:22:50.360
reads and pares uh files for me and that
00:22:53.320
outputs uh nice error messages and all
00:22:55.720
of that if that would exist um so yeah
00:23:00.039
what would you suggest how to decide
00:23:02.080
between the don't reinvent the wheel and
00:23:05.440
don't use external dependencies so if
00:23:08.200
the library so you don't have to
00:23:09.600
reinvent the will in this case if
00:23:13.080
um if the number of functions this this
00:23:15.760
third party provides is very big so it's
00:23:18.240
not just one uh one functionality right
00:23:21.320
it's not just uh I know it does not
00:23:25.679
let's say you want to read files you do
00:23:27.240
some processing you do some output uh
00:23:29.760
and you actually need just the the part
00:23:32.200
that reads the files uh most probably
00:23:34.559
you could implement it yourself if it's
00:23:36.159
not super complex and not something uh
00:23:38.559
exotic right then you uh implement it
00:23:41.039
yourself it's fine but if you actually
00:23:43.799
need all the five things that this third
00:23:45.880
party uh provides then then go for it uh
00:23:49.600
because then you kind of find with the
00:23:51.919
maintenance burden that it that it uh
00:23:54.360
comes automatically and uh necessity to
00:23:58.039
explain how to use uh this this third
00:24:00.440
party for your for your colleagues or
00:24:02.120
for new joiners it's okay avoiding
00:24:04.640
doesn't mean that you don't have to do
00:24:05.960
it at all it means that every time you
00:24:08.799
you ask yourself whether it's actually
00:24:11.039
necessary I'm fine uh to take this
00:24:13.840
responsibility and if the answer is yes
00:24:15.840
then just go for it the same with
00:24:18.240
Reinventing the wheel if there's no
00:24:21.080
nothing existing and you really need to
00:24:23.559
do this and you try to change design
00:24:26.360
that existing let's say uh
00:24:28.919
protocol uh still doesn't fit right you
00:24:32.640
have a some special case and you have to
00:24:35.399
you invent something like unique product
00:24:37.399
which didn't exist before of course
00:24:39.080
there's a chance that uh there's nothing
00:24:42.120
you have to you have to invent so this
00:24:44.360
is not about uh kind of killing
00:24:46.960
creativity completely it's about asking
00:24:50.200
uh I did all the things necessary I and
00:24:53.399
there is no way then yes just go for it
00:24:56.279
but avoid the magic uh if if you took
00:24:58.799
this third party please make sure you
00:25:00.760
understand how it
00:25:04.000
works I you said that you occasionally
00:25:08.200
see software and immediately think or
00:25:10.799
Maybe not immediately but think I could
00:25:12.880
do this better and I can completely
00:25:15.520
relate to this have you ever considered
00:25:18.320
improving on the thing that you think
00:25:20.600
you could do better instead of doing it
00:25:23.520
you know completely new that that's what
00:25:26.039
I try to do myself
00:25:28.960
personally uh and recommend to everyone
00:25:32.120
just to look around most probably the
00:25:34.480
things this thing is already invented
00:25:36.960
and most probably it does even better
00:25:40.000
better uh quality results that that you
00:25:42.720
imagine because uh this tool or this
00:25:45.159
protocol or this thing exist for long
00:25:48.520
time and then people solved all this
00:25:50.120
issue this it's highly recommended look
00:25:52.159
around use the time like the if you
00:25:55.520
would U employ lasz divent development
00:25:58.360
you get some extra time that you could
00:26:00.240
spend on learning and exploring what is
00:26:02.880
possible and then again use this next
00:26:06.600
time when you feel lazy and you don't
00:26:08.840
want to do this you kind of help
00:26:10.360
yourself all the
00:26:12.120
time does this answer the question yeah
00:26:14.440
thank
00:26:15.480
you more questions we have five
00:26:20.799
minutes thank you for your talk I wanted
00:26:23.559
to ask if you have any specific tricks
00:26:25.919
in mind uh to promote laziness in how
00:26:29.640
many how much mental effort do we need
00:26:32.360
to write and understand the
00:26:35.320
code I am not sure if I understand the
00:26:37.559
question how uh um how do you like write
00:26:41.799
the code to decrease the mental effort
00:26:43.960
needed uh to understand it for other
00:26:47.080
programmers and to ah okay so uh this
00:26:49.840
one is tricky I think only with practice
00:26:52.080
so every time when you uh write the code
00:26:55.279
try to imagine yourself uh in
00:26:58.799
half a year or try to imagine yourself
00:27:01.600
being a new Joiner of this company um
00:27:05.360
and then from this position look at the
00:27:07.399
code if it's unclear what it does or
00:27:10.120
especially why it does something uh this
00:27:13.200
way uh try to either change design if
00:27:16.480
it's not possible then uh add
00:27:19.159
documenation right there so people don't
00:27:21.000
have to hunt for this doc Commendation
00:27:23.000
this would help with this with this with
00:27:24.559
the mental burden
00:27:28.679
there was H yeah um yeah so what if you
00:27:32.279
choose like a path between uh inventing
00:27:35.799
the new wheel and using the third party
00:27:39.080
dependency you um you want to actually
00:27:42.640
contribute to a third party dependency
00:27:45.440
with a feature that wasn't there before
00:27:48.279
um but then you realize oh um maybe this
00:27:53.200
is avoiding future work for other people
00:27:55.720
but it's actually adding future work for
00:27:58.200
for me because now people are asking me
00:28:00.600
to maintain the future so where where is
00:28:04.919
the value here like where what would you
00:28:07.279
say on the Spectrum like are you um is
00:28:10.799
the laziness only in your own scope or
00:28:13.600
is the laziness like in a bigger picture
00:28:16.039
of the community you you should strive
00:28:18.000
for for bigger picture of course but if
00:28:21.480
there's a choice between inventing
00:28:22.960
something just for yourself and
00:28:24.200
contributing to something which exists I
00:28:26.480
would probably go for the the the second
00:28:28.880
choice because then you have people and
00:28:31.080
then you also share uh the maintenance
00:28:33.559
with them right if if this change was
00:28:35.399
accepted it means someone also decided
00:28:38.159
to share responsibility with you so
00:28:39.760
you're not the only one uh who has to
00:28:42.440
maintain it uh later
00:28:47.760
on seems like we have no questions
00:28:49.960
anymore thank you very much you're
00:28:51.919
awesome
