Arriving at EuroSTAR
I arrived on Tuesday, the day the conference part of EuroSTAR started. It was a really easy journey taking the AirLink bus in and my hotel was round the corner from the stop.
I had just enough time to drop my bags off and then head to registration.
I had a brief look around the expo before heading to the intro and first Keynote. Below are my notes of the day. They may work for you or they may not. But I hope you find something useful! 🙂
Professor Andy Stanford-Clark, IBM Global Business Services, UK
This was a fun and informative talk. Andy made the subject really approachable and relatable with real life examples.
The internet of things is becoming more and more intrusive into our daily lifes as access to ready to go devices such as arduino’s and raspberry pi’s has increased.
Andy highlighted that based on data we can create actionable insights.
instrumented – the devices
interconnected – communications
The initial examples he provided were around home automation, where by monitoring your energy usage, you can then plot graphs and visualise the data to then encourage new behaviours such as turning lights off.
But the internet of things also means more engagement for businesses.
one to many
tools for testing automatically at scale
What will be important for testers in this world is simulating and emulating the environments these are performed in. As well as error reporting.
We need to ensure we build models of the things we are testing based on how it is being used in the world.
Alongside this is the weighing up of value vs security and privacy invasion. Is it worth it?
As testers we are facing a new level of technicality, but does not necessarily mean a high involvement in the small details but we should be aware of the technologies and implications of the internet of things to confidently test applications.
During the break I mixed it up and did a puzzle as part of the test lab. It was about focusing on the task at hand and getting tunnel vision. I had ignored certain parts of the puzzle and how it reacted and discarded them as not important. This was a nice lesson and the debrief chat afterwards was fun.
Amy Phillips, Songkick, UK
Amy gave us another side of her story at Songkick.
The emphasis was on startups being about uncertainty.
This could be extended to: are you a start up if you are developing while facing huge uncertainities? If yes then I have been at a start up development wise for the last 7 months! 🙂
Startups come from a problem to be solved. But it is the problem of an individual, to make a success from the solution the problem may have to be slightly altered and then in turn the solution.
What problems do other users want solved?
There may be a discovery phase and development phase – development can be moved between those two phases.
Notably these phases affect testing.
Discovery – low assumptions and design. it works to show the user and get their feedback. Need to leave time to go back and re-develop properly if the solution worked.
So to get this feedback you need real users to use your software to get real feedback. This means shipping often and quickly. Startups need to get as much info from their users in as short a time as possible to be able to satisfy investment, no matter if it is money from a mother company, investor etc.
Testers are often outnumbered in the start up world. And if you are a tester at a startup you are likely looking after several streams of development and ence unable to test everything. To tackle this issue Amy proposes education. Spend more time educating and pairing with testers to give them some responsibility back.
cannot test all of the features.
teaching good testing practices early on are essential to not build up too much technical debt.
Key qualities of a tester in a startup world:
focus on quality and not testing – not a lot of time
willing to compromise
spot opportunities – what is better value time spent on? testing or educating?
handle uncertainty well
be able to adapt and change – learn every day
align testing with the team’s goal
no test reports
Testing becomes hard as testing provides information on an undefined product that is looking to solve and undefined problem for an undefined person. What to do with the results?
The main thing to do is to learn together and communicate the learning – foster a good communicative environment to enable the startup to strive.
Alexandra Schladebeck, BREDEX GmbH, Germany
This was a fun talk and I am unsure how to relate my notes. Why not talk about it on the forum?
Alex gave us 8 recipes in total.
These made me think about what works for us and what are the downside and upsides.
I could definitely relate with the themes of the talk. I liked the fact that thinking about testing practices as recipes, let’s you think about adding different spice and hence a different spin on a process to try something new.
I felt this was one of the main messages, follow a process/guideline and adapt it and then make sure you review if things improved or not. Even at the risk of personalities clashing, make sure to hold retrospectives about what you tried.
My take away is to try and play with estimates and story points. I liked the proposed approach of an ABC system and make the story fit the story points.
a – no time to 1 day and a half – dev feels confident
b – day and a half to three and a half – more complex, may contain unknowns
c – too big, needs refining
I will propose this new variation on the story point recipe in the next retrospective. How are you adding spice to your processes?
I also want to add the question:
What can WE get done today? This is to emphasise that testing is also a team responsibility. To foster this Alex mentioned that she did not introduce more swimlanes on her JIRA board but tried labels instead. We work quite well with swim lanes but there is definitely more of a “over the fence” approach.
Other things I need to think about from this talk are introducing:
bug fixing days?
dealing with finding more than fixing
dont talk about bugs but talk about analysing the stories
Sometimes the team needs to understand the problem before they can accept the solution. Let them feel pain first but work as a team – let go of personal views in favour of the team.
Rob Lambert, New Voice Media, UK
The last talk was a keynote by Rob who I met earlier this year for the first time. He talked about how he changed the culture from yearly releases to weekly releases.
Previously test were slow boring and ineffective. This is a popular pattern that companies often rely on.
During crunch times anyone who can click a mouse becomes a tester – what does that say about testing?
Let’s do something different:
Pervasive testing – all the time by everybody.
Have a releasable candidate as often as possible so the business can decide to ship it.
What do your customers need? Do your customers’ needs move quickly? So need your releases then.
Customers provide the ultimate feedback loop – have we built the right thing?
Start change with a concrete vision:
everyone tests all the time
become one with data – understand it
remove testers from the centre of the universe and put testing at the centre
metrics – visualise your impact
pre production – real world kick from real users
test and monitor in production
every team supports their stories not the ops/maintenance team
real data is more powerful in guiding tests
Exploratory testing is only possible if checks are automated. The biggest challenge may be people and politics and not the technical details. Be aware and address this.
And that was all for the day from me! Why not share your experiences on the forum. I would love to read about them!
I hope the notes are useful. Below a pic of the view as the sun was setting and the drummers who led us downstairs to the party sponsored by HP.