#UKSTAR2019 review
This is my quick review of 2019’s UKSTAR.
The conference and the venue
I dare say that this was the best conference I’ve attended. Not only the organisers were amazing, the venue was top notch and the catering was delicious!
Diversity
I want to give a huge shoutout to UKSTAR for doing an amazing job in bringing diversity to the testing community.
I’ve never seen a conference with such a high level of speakers and where most of them were women.
The environment during the whole conference was really good, the networking was easy and fun.
I’ve talked with people from all around the world, was a truly enriching experience.
It is a great feeling to be part of such a great community.
Even though many companies say they are pushing for diversity, we aren’t doing nearly enough. For example at Metropolis, we say we want to bring more diversity and we’re 95% white and 90% men 😭 .
I felt the struggle when hiring, even though I was trying to bring only women to my team, I couldn’t find enough suitable candidates, ended up hiring 1 of each.
Whova app === easy communication
Another great detail from the organisers was the use of the Whova app. Even before the conference started, the attendees were already networking and arranging gatherings. During the conference all the communication was done through the app which made everything much easier.
Small tip for next year: put the flags by default on all the cards by asking the spoken languages on the sign up form, it makes networking a lot easier when you see someone with a flag you either never talked to and want to get to know or find people from your nationality.
Talks
First up was the queen of the testing conferences, Angie Jones.
I’ve given her plenty of praise on my review from 2018’s Test automation day in Rotterdam. She gave the same talk on testing AI/ML with small improvements.
Her never ending efforts to contribute to the community go far beyond from giving talks and writing blog posts, now she’s also starting an online university to teach everything test automation related, for free! Is there anything this woman can’t do???
I never liked the theme too much but forced myself to attend in order to learn something new.
Dan covered a wide range of testing and attack techniques you can use to make sure your application has the basics of security covered.
I was a track chair volunteer for UKSTAR and my first talk doing that (helping out the speaker by introducing them and handling the Q&A at the end of the talk) was with Joep.
He came out as a very direct and succinct person when we were discussing what he wanted me to say when introducing him before his talk, so I was not surprised his talk was exactly like that.
Very direct to the point, short and sweet, it is on my top 3 talks for this conference.
Usually when you think about creating a test automation framework you think about all the different tools you need to put together.
Joep taught us how to get value from a very simple combination of pytest+requests just by looking at the logs and reports from this tools. No need for fancy html reporting or extra logging in your code, just dive in on the default logs from the tool and find issues with some debugging sessions.
Usually, when you go to a conference and someone talks about AI/ML or Visual testing, is just another sell pitch for yet another revolutionary tool that is going to solve all our problems.
Not only the talk was engaging, fun and creative, Viv talked about various different tools, open source and paid, and showed us that you can do visual testing with whatever budget you have.
Yes, Applitools is still the best one, but if you can’t afford it, for sure there’s something out there for you.
Chatbots and their usage of AI/ML is something we talk about and read about almost daily nowadays.
For me as a tester my first concern is “How do I test it?”.
Well, Hristo tried to tell us how. I say try because it is a very tricky and elaborate topic to discuss in such a short time, especially when you go so deep as Hristo went.
I’m sure we’ll be seeing more of him and this topic in the future.
Fiona Charles. She’s been in the software industry for more than 40 years, more than any of us ever will, most probably.
She’s very experienced in giving talks as well, and you could tell. Similarly to Angie Jones, the way she talks captures the listeners attention and the thought provoking content was perfect to end the first day of conference.
Nowadays we only think about getting the highest tech stuff out there, not thinking about the consequences.
Fiona showed us the lack of diversity in these technologies and their testing, the dangers of having computers overruling human interaction and the way historical data influences (BADLY!) the behaviour of machine learning algorithms.
A lot of the things mentioned will stay in our minds for years to come…
The 2nd day started off with Bas Dijkstra’s talk on “Why do we automate?”.
Something that both him and Angie have been preaching a lot lately which I totally agree: you should pick the right tool for the job and not the opposite. If Selenium is not the best thing to use in your case, don’t use it just because everyone is doing it.
Nowadays, whenever you jump into a team as a QA, a big part of the job is automating tests. Be it creating frameworks from scratch or adding tests to an existing framework.
We never ask ourselves, why do we automate? Do we really need it? Does it really save me time? Why am I using tool X when tool Y would be better applied here?
Bas made us think about this.
Stop automating just for the sake of automating. Find the problem you want to solve with automation (if there is one) and then pick the right tool to solve your problem, not the other way around.
I chaired this talk for Claire and, similarly to Joep’s, what a talk it was!
This topic is very close to my heart since I’ve struggled with it a lot when developers and managers think of us testers as a 2nd class citizen. We need to be heard and seen without being the black sheep of the team, we ARE part of the team, we are NOT the enemy.
Another topic discussed was the fact that nowadays, all QA job descriptions mention 100 tools that we must know about.
In a recent talk with a group of recruiters, we discussed exactly this.
They asked if we really had to knew about all the tools and my answer was no, we need to have an idea or know how they work but we don’t need to have experience in all of them.
If you find that a job description is too much for you, apply anyway and talk to the recruiter to figure out exactly what the team needs, most of the time they just put the tools because everyone else does but theres no real need for it…
I guess I just found another person to follow on social media, I hope I’ll learn a lot from her.
Another great talk, this time by Corina Pip from Romania.
Corina showed us how to hunt down a bug when you can’t immediately pin point where they come from.
Be it by looking at logs, checking metrics, checking the load balancers, the user’s location, the user’s device, the time that the bug tends to appear… All of this is important to track down those “random” bugs.
There’s a lot to digest from this talk, only people that have felt the issues in their skin understand how important the things mentioned are to us testers.
If you have the opportunity to see Corina live or a recording of this talk, please do it now, you’ll learn a lot.
2nd time I attend a talk from Amy. Last one was about implement CI/CD in your company, this time was a more general topic.
This is something that eventually you’ll go through in your career in software development. Be it applying the change or watching it / feeling it being applied.
As Amy mentioned, it is very hard to keep people engaged when going through a change in the way things get done. When there’s a process in place, people tend to stick to it. In order to change that mind frame, you need to give people some kind of benefit or a gift, similar to when you teach a dog new tricks and give him a treat so he keeps doing it — positive reinforcement.
After the organisers send off, we had one more talk before heading home.
I know he’s going to hate me for saying this (SORRY BAS!) but I was a bit disappointed by this talk.
His rhythm was off, too slow, not direct to the point and losing too much time with questions from the audience. Maybe the setting was not the best for this kind of talk, or I’m just too experienced for the level, but I didn’t like it so much. If this was a small workshop and people could discuss the questions the result would probably be better, but in a big room with so many people…
Of course the topics discussed were interesting: when you build a new test framework you need to pay attention to several things instead of just the tests. You need to have clean code, abstract the tests from the environment they’re running, make sure the tests are atomic and totally independent, use API calls when you can instead of doing everything through the UI, etc…
So, was it all worth it?
Absolutely, yes!
It was a great conference all around.
A bit on the expensive side (700 euros) but worth it in the end.
For sure I’ll come back next year.
If you’re looking for a new conference to go I totally recommend this one.
Thank you to Metropolis:lab for giving me the chance to attend this conference, thank you to the organisers for the excellent program and conference and thank you to all the speakers for the added knowledge to my brain!
Hope to see you again in a few months, maybe as a speaker!
