Panel Debate Led by

Dan North

Recorded at GOTO 2015

okay so we're going to start the panel a
couple of minutes early because then we
can just have more fun with it I've been
a bit opinionated with the panel so I've
looked through all the questions of
people posted and a sexy what I've done
is I've picked out questions and put
them into topics i did in preparation
which is really unlike me I know like
Hank like what and so what I figured is
what was a couple walked conversation
were sitting down having a beer I
thought that would work better if we had
a bit so this example we can pretend
we're sitting down but I thought we
couldn't fake beer that's not good so so
our first question is a bit meta and
it's more of an observation really this
is I don't know if it's the best place
to post it but do you think posting
questions to an app is agile I quite
like that i'm not sure i understand
serve i thought i'd share that with you
i'm gonna start with right it's also
hundreds or not 100 several questions so
we're sort of back and forth on the
panel but reasonably quickly and i said
something you really want to climb into
you in which case go nuts so people's
stuff there's and i love that you that
there's a tone for some of these
questions it kind of comes along with it
across in the question so I'm going to
just read them verbatim what do you mean
this is I presume it to Hank what do you
mean you make all your engineers speak
in public is this real life ? said in
horror aghast in horror I think I talked
to the guy outside excellent and I told
them hey if you look you will see that
there was this little diamond shape to
how we want our engineers to rate and
speaking and writing tutorials is
especially a skill that we require for
proficient and expert and then he said
but I don't want my engineers to speak
and be very great at that because then
they might leave so then we had a
discussion what I said hey you're
keeping them down my friend right is
that right they could be so much more
than they are today i talked to adrian
quickly and he said we need to get there
so many talented people out there we
need to stop wasting their talent right
so the big question is i think how do we
get to a place where we can nurture
people skills and have them all become
awesome yeah i think i forgot your name
my friend even though you gave me three
cigarettes already today so thank you
for that I hope my wife is not listening
to this don't smile kids yeah don't
smoke but but yeah that's really the
discussion or we do we really want to
spend energy and time nurturing the
skills of where our engineers fantastic
thank you anyone else want to speak to
that in terms of getting your people out
speaking in public we've works this is
compulsory and I think it's similar to
the attitudes wave by Rachel where's
kind of owning live there's another
aspect of this owning the customer
experience in a live application if
that's what you're responsible for
developing goes hand in hand and things
like documenting it talking about it
with people if you're not willing to do
that then we don't want you to be in our
team we I've got yeah I've got a night
here we don't force people but what we
do instead is say people well you know
you will probably in your career need to
speak to people you will need to stand
up in front of people this would be good
to practice and you can practice it in a
safe environment so having tech talks
back at the office where you're just
talking to your colleagues that's an
easy way to do it running a meeting
though I mean there are some people who
really get paralyzed just to stand in
front of a white board and talk to their
teammates but if you do it enough times
you realize it's not that bad and I
think just supporting people is
important yeah it's a skill you can
learn it so an example of that I had a
team I was coaching a very small teams
like four people and part of what they
were doing is the first-ever a July
anything and so I said okay so you guys
are going to demonstrate progress at
your showcase okay so they're going to
showcase I'll be business people along
when your show them what you did and
they're like color drains out their face
right for the first time the first time
they do this they're kind of all we did
this thing and there's a screen and the
business sponsors like wow you did that
in a week like normally I'd be seeing
diagrams of that in three months right
and you've done that in a way they're
going yes and then I've sent out the
second week so I can showcase they're
like well this week we did this and this
and blah and and they're feeling quite
pleased about progress because people
aren't you to see in progress and that
was going well and then the third week
this guy starts bringing his buddies
along right other other business guys
saying look at what these guys can do so
now there's like through by about what I
was there for eight weeks coaching them
and by the eighth showcase it was a show
right they would come on and they were
presenting and they had a thing and a
big screen and there were probably 30
people nothing to do with the project
two people actually on the project
everywhere else going blindly where can
we get some of this and and yeah you're
right a learnable skill and they were
enjoying it themselves they were having
fun they were energized there's a really
sad corollary to this story which is
then so where I left and they carried on
doing this thing and then the project
basal is that I was a great pilot was
successful we learned load of stuff
right you to go back into that team and
you go into there and you go over there
and carry on doing what you used to do
two of them promptly quit and went and
worked for much much more enlightening
companies one of them the last time I
saw was looking just a little bit
Haggard and sad and I lost track of the
other one yeah so the whole thing just
dissipated back but yeah it's definitely
learnable skill cool hurrah now oh ok so
there's a similar sort of thing this is
about changing people so hanger into
Jamal he was speaking of his questions
coming in about 40 questions just on his
page so there's lots of sort of change
and scaling things tips for engaging
developers who
working within the agile framework not
sure the agile frameworks a thing but
we'll go with it but seemingly do not
see the big picture or reason why we're
adopting this so so so tips can what
tips would you give for engaging
developers who are kind of there in this
sort of transitional thing or some big
transformation and they're not getting
it they don't see the big picture and
they're lost what how can how what tips
would you give for engaging with those
people I really don't know other than is
famous article I read from of course
again Netflix where which basically said
if some engineer is failing to get the
context then some manager has failed or
some engineer makes an error than some
manager has failed to give the correct
context right so I think IT managers
should be able to transmit the right
context and make people feel
enthusiastic about this HL adventure oh
this is an answer or what about the ones
that don't there's only the body count
right so okay can I open so well I think
so you forget the agile framework and
getting excited about the other
framework but maybe just getting people
to talk to users that are using
something they've made and like actually
seeing them use it and have and how bad
it potentially might be can often it I
think often software developers are
motivated by problems so if you show
them the bigger problem rather than
being isolated from that problem that
can engage people if they understand why
we worried and why we try to do
something different can it also maybe
terrify them or paralyze them will freak
them out or something how do you know
which you use to shower to protect on
say just get introduced them to people
who are actually used
the software because sometimes those
people are not the scary high up people
they're usually the lowly people who
just trying to I just wanted to do this
and I can't and meeting somebody who has
a genuine problem you think I could
probably fix that might help people to
kind of go oh it's not about this big
and real transformation plan it's more
probably about connecting with people
who work in our business all people are
working outside okay I've seen this
problem in a company I used to work for
which specializes in pretty XP style
agile development and there was a rule
that if you interacted with members of
the team who were doing this you had to
work in their way which a lot of the
developers in the team that I was
associated with found extremely
difficult and frustrating because they
they was one particular thing they'd
absolutely did not want anyone that was
pair programming and it wasn't just
ordinary pair programming but we all sit
in front of the screen it was replicated
screens you know each person has a mouse
but it controls one pointer all the
whole thing and they just found it to be
slow they didn't like necessarily
working at the same speed as another
person so in that situation we had a
problem because you had two different
teams of totally different cultures now
the objective was to get them to work
together not to necessarily adopt each
other's practices wholesale so the
solution in that case is not to make to
force people doing something they really
don't want to do but instead to go to
the question behind the question will
say what are we trying to achieve here
and what we were trying to achieve was
collaboration so what we lot of the
people ended up being happy just sitting
next to another programmer for a day or
two working with them side by side not
necessarily always doing identical
things but just actually going through
the code I mean walkthroughs and
discussions and that was kind of an ice
breaker and a few a few people model
after that we're happy to try out the
replicated screens thing some lighters
some didn't but the point was that we
actually thought off
so I think not being too stuck in the
mud about those things is good I want to
try to get a bit of a discussion and a
bit of contention going so I'm just
going to think about that as we talk
I've got a question specifically for
Frank but I think we can broaden out
this isn't a general theme of testing
but this I love this how did you test
that it would work with big crowds right
because because you can't just get a
hundred thousand test people together so
it says I guess when you are developing
you only had a small number of
wristbands or people to test with yeah
there's a fine yeah there's a fine line
between testing at some point it goes
into praying because you know obviously
you don't really know we had like we had
a limited number of wristbands at
someone we just put some wristbands on I
went into the city when it was crowded
to figure out because human bodies also
make a difference and in the end like
music festival is in wireless terms the
most noisy place on planet they're like
huge like a forty thousand watt amp is
going to cause some interference or
everything that's there and you can't
replicate it you can't do that in a lab
so at some point it is just scary so I
didn't mean how do you mitigate that you
just say right we're just going to go
with this and see what happens oh well
yeah at some point you just have to bite
the bullet maybe just get on and print a
hundred thousand won anything yeah well
I think much are you asking well so the
question was how do you what kind of
testing do you do so that you're a bit
more confident when you put it give it a
hundred thousand table and first of all
it's going to be okay well yeah we do as
much as we can but we can't have on the
thousand people so we can do all kinds
of tests on the wristband individually
on the software on the hard way and also
just testing if it stays together for
long enough but yeah there is going to
be a difference between a music festival
and your test setup and that's something
you have to accept and take a deep
fantastic and then a very quick one a
great talk to Frank does the wearable
watch actually show the time no ok that
was nice quick question so ok something
hopefully little bit contentious I need
a more than one word answer here do you
think containers are going to make Java
interesting because we've got an osgi
guy and we've got a couple of Java
people say it was that question and if
you're one of our our co-sponsor IBM you
might want to look away now the app
server has made itself redundant by
growing into incredible complexity it's
been replaced by things first by Tom car
which is kinda at least that mail by
spring boot which has got rid of the app
server completely just in the JVM it's
got nothing to do with containers what
containers are doing is they're
standardizing deployment which is one
layer down in the stack that I drew so I
think you know there's a bunch of trends
going on at the same time not quite in
sync of which you know i think you're
touching on a bunch of themes with your
question but i don't think the container
itself is going to make the app server
redundant i think at the Apsaras done a
really good job on its own i do think
that like observers like tomcat were
used as containers to separate different
apps right you can run several separate
like an Universal classf oh yeah true
but I think that works a lot better with
containers I think you can treat rust to
two different apps you don't really
trust in the same jvm exciting sorry
another point which is a technical point
is that with containers you'll see
people building more distributed
applications that was the point of the
cloud native stack story most java app
servers presume a very restricted notion
of a distributed app which essentially
is the old layer cake style I mean they
exist because people wanted
to replicate the business model
developed by jeff bazos and amazon
anyone else wants a offer anything Hank
like yeah whatever okay and so there's a
regulatory thing here and I'm interested
especially well on the on the hardware
side because there's a whole bunch of
hope things you have to jump through as
well in terms of getting something a
physical device that you can
mass-produce important people's risk
venue but the question here is have you
and it was aptly relating to hang so in
a banking environment but I think it's a
broader thing than that have you how
have you met some of the regulatory or
compliance challenges on your agile
journey yeah I'm just thinking very
quickly now if I am being very naughty
by telling this story got this friend
that works at a dutch bank and hey tell
me and the European Central Bank came to
visit the end here and they send a
couple of smart alecks with their
preconceived notions of how things
should be done so question number one is
where is your test department and open
it quite friendly we have no test
department so I saw his pencil goat a
book he has no chest Department yeah so
I said if you're interested I will tell
you how a continuous delivery pipeline
works and how we gather the evidence
that everything has been tested before
it can go to production and he met made
a mistake of saying oh yeah please tell
me so I explained to him in horrendous
detail for two hours how that worked and
one of the two had difficulty keeping
awake
I wasn't dead entertaining obviously but
their credit that we had a different way
of doing testing without having a test
department and i have not heard new
questions since which could be that they
are still studying on how to tackle this
i don't know is this an answer to the
question really yeah so i think i used
to work in the bank notes the issue with
financial regulation is one where people
have a transactional presupposition
about everything you do so they assume
that everything is a transaction meaning
like it all go back I've wake up that
you can't go back so they think that
they don't understand concepts like roll
back or you know if we commit this
change to the app and we made a boo we
can just take it away and do something
else it's really alien to them so my
recommendation for dealing with these
people if they have to come around to
your office is the austere what do you
think about Bitcoin and then when they
run around the room screaming I've
interested in right what do you do in
terms of like regulation and and all
that type of stuff I don't really know
there's a lot I mean there there there
are some things that need to be done but
generally they are yeah you have to get
f6a compliance nice benefit exactly that
kind of stuff but that's not like a huge
a hurdle it's more like a public
personality in process yeah yeah
basically that cool okay so it's much
more systemized don't say another thing
sorry don't apologize you're on a panel
that's what you here for the companies
that that we've been speaking out today
these sort of pioneers like Netflix and
so on our consumer facing companies with
big web presences which do things like
take details personal details of
customers online and hold them in trust
and they are perfectly able to comee
engl agile practices with strong
compliance around holding their
consumers personal details so they can
do it so can everybody else oh great I
think Munsen I found as well what's a
bunch of financial places now is there's
some regulatory compliance two very
different things so I'm regulatory is
tends to be we would like you to provide
this information or we would like you to
if we ask you this question you'll have
an answer for it so things like client
money is a popular one right which of
the money did in your pot is your money
in which of the pond money in your part
is someone else's money because you need
to know that and the way people engage
those things is very much on a scale
right so I've seen some Institute's that
will that will remain nameless that I
kind of they know that if they haven't
got an answer by X point in time there's
going to get fine and they say well that
fine is one possible option right and
that's a cost of business if we don't do
that and if we do do this thing how much
it's going to cost of be compliant will
take the fine this quarter and we'll
have it done by next quarter or
something it's a commercial decision a
lot of places I find they knit or a lot
of places misunderstand compliance rules
in a way that means that they
overcompensate so things like
sarbanes-oxley I've seen sort of very
very integrated teams completely
compliant the sarbanes-oxley because I'm
a dog who says two pairs of eyes need to
have seen something I need to have
approved something so that it can go
into production it doesn't say you need
to have a hermetically-sealed operations
and release group that's how everyone
implements it and then they go oh you
can't do agile in a in a regulated
environment Allah we can't do it in that
regulated environment because you know
you're not talking to the people this
discussion I really recognize it was
translated to us as the separation of
duties yes I think and we told them well
as the deployment is completely done by
pushing on this button so this system
basically with that which does the
deployment is separated from the guy who
fantastic we didn't get away with it
yeah but that's a different thing right
if everything is auditable it is we've
made sure of that there's another thing
that's the thing is that you can design
in auditability and the same way you can
design insecure ability so it'll wear
the rugged thing so security is a
characteristic of a system secure
ability is a characteristic of the
design auditability monitor ability all
of those illa tease our choices you make
when you're building something but
another thing about this I think just
what people have said in the last few
minutes has made me realize that if it's
a Sybok style philosophy where to
individual autonomous agents have to
sign off on something for it to go
through it's a lot like pair programming
code review you could argue that all of
the philosophies of regulation of sort
of baked in at a very deep and
fine-grained level at in agile
development so what is what why does it
why then is it challenging I think
perhaps it's because the regulatory
mindset is one of wanting to point the
finger at individual and say it was your
fault you are accountable and you you
can't do that when it's the process
that's replace the person to to deliver
quality and I think that's quite quite
mentally challenging but a good thing
still want to react to that and
especially in the light of the rugged
discussion that we had yesterday right
we have to assume that one of our
colleagues is or can be brought into a
position where he or she does something
that is not allowed and then we have to
catch it and that's really a thing today
and for a bank it absolutely is a thing
so of course we have our measures in
place for that and increasingly so and I
think this is going to be a thing for
many many teams who write software so
being able to at least detect that
somebody did something at that could
potentially be harmful and that that
event triggers somebody to have a look
at it immediately all these kinds of
measures will become more and more part
of continuous delivery and the
pipeline's that we're building was in a
learning experience for me really when
we built the central pipeline how many
of our risk people really were
interested in hey how do you build this
whereas for them it was also a
manageable when everybody did their own
continuous delivery and their own had
their own build surface but as soon as
we created the central thing it felt
like everybody was hitting on me to make
it perfectly perfect from the from day
one well that happens too and I thought
I mean certainly in your talk like you
can see that there's an art because as
different things at all as you as you
pick your battles which things you're
going to tackle now which things are
the other stuff needs to bed in yeah so
okay so there's more questions about
kind of scaling and change and that kind
of stuff so mmm that's yeah I'll just go
just gonna read this funny all of this
this again is is for Hank but I think
though it does apply to yo guys as well
all this IT focus are you still running
a bank
or has your well now and then and then
the the the the payoff part of the
question or has your business evolved
into something else I think a bank is ID
and I t is the bank this is an answer so
what's different is it up take it no
more and more are it sounds like big
phrases so I feel a little bit awkward
but and I can see that that a large part
of the population that is actually
driving and building software for our
customers is all righty folks right and
they are very much part of the
discussion of the what the next feature
will be because they know from a
technology perspective what's important
and they more and more interested in
what our customers say about a product
so having this very direct feedback loop
for example from the app stores where
customers either complain or have
suggestions and today even developers
are reacting to that or giving their
comments really really helps it's a form
of engagement in a form for them to
understand what is valuable in their app
or not i'm not sure if there's this big
difference anymore i think also when you
start to blow the edges of teams and
organizational boundaries the certainly
the successful technologists I've seen
in banks understand banking that's the
thing that they do so I've been no
worked in trading firms where you got
programmers who understand the domain of
trading as well as the traders and then
the traders are going I'm going to learn
me some Python so I can put some ideas
together and show you guys and you can
go and turn it into a product and you
get is fantastic cross-fertilization
going on so I think it's where we go yes
we're still running a bank we're still
having a banking licence that's a pretty
good indicator right and we intend to
have that one for a long long time
that's a big difference between traders
and bank customers so you know you can
think of a business as a collection of
AP is for interacting with it has
customers the customers there to buy
things from it or to put money into it
too
you know to xq some kind of action and
those interfaces that consumers are
using are like AP is and more and more
that they actually are api's so it's
natural if the customer experience is
digital to think of them as my teeth
problems that need to be solved but i
think what a traitor is in a very
different situation because they usually
inside the bank making decisions on
behalf of the bank as to what risk to
take there's a slightly different thing
I wouldn't come mingle the 20 stone i
think there's there's something there
about understanding all the different
stakeholders involved so you got the
upstream guys in a business the
downstream guys and operations you've
got the customers out there and you have
many many different types of customer
and you're as you say it's a bunch of
API as your engagement with each of
those is something that's going to be a
function of time I think as the bank
changes marvelous so okay and I suspect
so he's all for Hank but these are the
other guys please jump in there's
there's a reason it's very obviously how
do you get talent to join your company
okay pause especially in fields that are
considered stuffy dull or boring right
like making awesome watches to hand out
to people at festival node like banking
Rachel you can speak to this as one I
that's media darling that's cool having
one day a week to do whatever you like
it's kind of a cool way to attract and
but we do I think there are people who
are drawn to the idea of collaborative
working there are also people who don't
like that kind of thing and so if we
make it really really clear that the way
we work is collaborative and we kind of
get out then we have a meet we host a
meetup and we put post out and talks
about what we're doing it helps people
to understand oh that's what I would be
getting myself into if I did go there
and it's kind of part of the interview
process so you if you didn't like pair
program
and you would walk out in the pair
programming through as we have
occasionally had people do but it's if
you make it clear this is the kind of
world you're going into then people
couldn't decide whether they subscribe
to that and want to be part of it or not
I say yeah i agree i think a lot of this
branding as well as a another conference
in london i was at last year last year
maybe year before where bank of america
so big conservative you know american
bank type thing how to stand and it
wasn't like her you know come and work
at bank of america those all gleaming
max and they've written a python
challenge thing with a bunch of stages
and the last stage was time then it had
a leaderboard right and there's like a
guy's a geek conference this thing was
buzzing the whole time and there's other
folks who are other sponsors are going
to come and live here so he's a free
spongy thing right and they're going how
fast can you hack on python they're
going to let me just find out and it
sends a very different message yeah and
i think it people muddle the app you
build with how interesting it is to
something's being used at festivals
doesn't mean it's like exciting code
might be i think it probably is but do
you see what I mean they're serious you
don't then just because it's in banking
doesn't mean it's not exciting like
they've got some very interesting
challenges so it makes it hard so hard
problems or what people look for right I
completely agree with that yeah I can
say something about it because I'm quite
active in miyapur meetups around
Amsterdam and the ing turns up quite a
bit so they are really visible there and
that's for me if I think about AIG I
think other their deadly do cool stuff
I'm not looking for a job but still I
know that in the back of my mind so I
think that's really good to have
visibility there I mean that's it's
always a bit of a sometimes there are
too many recruiters at the meetup
nowadays but yeah I think it's a good
place to be i watch honestly makes
watches for festivals things your bank
is a cool place to be you should just
I'm a bit put off by the question really
because I don't recognize it and and so
thank you this gave me a little bit of
time to prepare my answer but here it is
what talented developers really want is
hard challenges and to try their hands
on new stuff if I look at the amount of
new stuff that we have rolled into the
bank in the last two years it's amazing
if i look at the platform that we're
giving two engineers to become really
great that's truly amazing so two of my
colleagues are in Silicon Valley next
week to give a keynote at Cassandra
event so so I think a bank right now
today and the situation editing is a
fantastic place to work as an engineer
there's so much so easy to influence the
people it's so easy to influence the
course of the the strategy if you have a
great plan it's really an awesome place
to work right now and banks have to be
that or they will be poked by adrian's
unicorns right so there's an image so I
thought so so all I want to say is be
aware of the unicorns with the big
pointy things on their heads back but
hey thanks or transforming at least ing
is transforming at an amazing speed and
and if if there's something that you
want to be a part of at least well come
and talk to me fantastic excellent so
lovely panelist one are likely you're
placing comments on a on what do you
think say we all assemble next year this
time and have a beer what will be
different what would you want to be
than one beer I like that but yeah so
what do you think is going to be
different in the world in a year
will you still be in these corporation
I'll be working in a much bigger company
dad knows what the future may bring I
think it's really hard to predict I
position to adapt to change and to
instead of trying to think like big you
know where will we be now kind of thing
what would be good for us to be able to
do so that we're more flexible so I tend
to think if you can bring more
flexibility to your approach then that
allows you to take the chance of
Alexis we will have taken over the world
everyone will be using it I mean really
important changes in technology are the
ones that impact people when they take a
long time to be recognized usually only
in retrospect I mean I was astonished
when Amazon Web Services released their
actual revenue figure some people
express surprise of them I mean it was
obvious to everybody who's paying
attention of these guys were making
money at quite high profit margins for
some time ago yet people insisted until
the last possible minute that it was
impossible to make money in the cloud
and all you were doing is adding a tiny
percentage to you know procuring a
machine it was put as pathetic so that
just goes to show you how hard it is for
to understand really significant
technological change you know there's
all these unit Adrienne's unicorn so
I'll use that phrase indefinitely now
Adrian the unicorn appearing all unicorn
herder it was still insisting that I
check problem number one has been sold
it can't be true I'll fantastic there we
um thank you if you look back over the
last segments what's the most
significant change in technology can
anyone even answer that question I can't
but I think that I'm I hope there'll be
more people here I it's a good
conference so thanks for coming and you
know I hope we're still in business that
would be good I think we're doing pretty
well maybe working with some of you guys
and I suppose the really big question in
our industry there's two questions one
at the macro level one of the smaller
level in the industry at large there's
concerns about whether the entries
overfunded I think if that proves to be
true and i'm not sure if it is that will
have a negative impact of all kinds of
things in the way that the tech crash
did in two thousand on businesses
outside tech and on a smaller scale
within the world of cloud native and
containers we're all waiting for dr to
start making any real money as soon as
they do it's going to help everybody
else in the industry so that's that's
one that i'm curious to see in the next
12 months thank you you triggered me
there so there's next year we are here
again and there's basically two topics
that i would love to be to be addressed
here so one is hey take a night what do
people actually need to learn to become
great engineers fast right I think
Adrian himself has some ideas and a lots
of experience on that and the other one
is in you triggered medeiros hey what is
this and there's this lady called
Carlota Perez and she's made a study of
how financial bubbles and Technology
bubbles go together and we that we've
seen that before so if we could get her
to also talk very good
very specific shout out there okay notes
are being taken i'm fairly short well
yeah no that's a tricky one I think it's
always less than we think that changes I
mean for us in the wearable space if we
look at the google glass at the Apple
watch not that much changes I mean
people still don't want to wear the
stuff so so I think it's less than we
think and there will be there will be
changes but slow really slow people are
slow so here's my prediction for not for
next year but for let's say 10 15 years
time I think right now but ok so we meet
back here 10 or 15 years time right I
years Ron alexis here population of
software developers is about 10 million
people today I think in it won't be long
before it's 100 million people I think
that's going to lead some pretty
fascinating changes I think that's cool
so it just remains for me to thank my
panelists thank you very much indeed
guys brilliant
you