Moving to weekly releases – presentation at Test and Verification Conference

At EuroSTAR 2014 I delivered one of the Keynotes and it was about our time when we were moving to weekly releases.

It focused around how our testing changed, how we made the change and how we relied, and still rely, on lots of data to make releasing as safe as possible for our customers.

Continue reading “Moving to weekly releases – presentation at Test and Verification Conference”

EuroSTAR – Star – Paul Gerrard

This is a short series leading up to Christmas where I feature a tester that was at EuroSTAR 2014, who I personally believe you, my readers, will benefit from knowing.

The last tester in this short series before Christmas is Paul Gerrard.

Paul Gerrard at EuroSTAR 2014
Paul Gerrard at EuroSTAR 2014

For those that don’t know – Paul Gerrard was conference chair at this years EuroSTAR conference – so we have Paul, and his excellent conference team, to thank for the content we enjoyed at EuroSTAR.

I’ve known Paul for years now, mostly through his excellent UKTMF event in London each quarter. Paul is very active in the testing community; he’s published many books, speaks at conferences, runs a consultancy business, organises events and is contributing greatly to our industry.

You can find Paul on Twitter – https://twitter.com/paul_gerrard

Paul’s website is here – http://gerrardconsulting.com/

Paul blogs here – http://gerrardconsulting.com/?q=blog

EuroSTAR – Star – Stephen Janaway

This is a short series leading up to Christmas where I feature a tester that was at EuroSTAR 2014, who I personally believe you, my readers, will benefit from knowing.

Next up is Stephen Janaway.

Stephen Janaway at EuroSTAR 2014
Stephen Janaway at EuroSTAR 2014

I’ve known Stephen for a few years now and we regular speak at the same conferences.

Stephen works for Net-A-Porter and is currently working as an agile test evangelist! Sounds pretty awesome.

I didn’t get a chance to see Stephen’s talk on mobile testing at this years EuroSTAR but Stephen is clearly an expert in the mobile testing world. He’s also a great person to be around at the social events – another tester who’s passionate about mobile, testing and sharing this love of testing with the wider community.

Stephen blogs here – http://stephenjanaway.co.uk/

You can find Stephen on Twitter – https://twitter.com/stephenjanaway

EuroSTAR – Star – Fiona Charles

This is a short series leading up to Christmas where I feature a tester that was at EuroSTAR 2014, who I personally believe you, my readers, will benefit from knowing.

Next up is Fiona Charles.

Fiona Charles at EuroSTAR 2014
Fiona Charles at EuroSTAR 2014

I see Fiona at pretty much every EuroSTAR and every year I thoroughly enjoy chatting to her about software development.

The reason I wanted to include Fiona in this series is because she’s refreshingly straight talking and direct about the challenges facing the testing industry. She’s also supremely knowledgeable – I suspect a week spent learning from Fiona would be invaluable.

Fiona is on Twitter here – https://twitter.com/FionaCCharles

Fiona blogs here – http://quality-intelligence.blogspot.co.uk/

EuroSTAR 2014 – Adam Knight

This is a short series leading up to Christmas where I feature a tester that was at EuroSTAR 2014, who I personally believe you, my readers, will benefit from knowing.

Next up is Adam Knight.

Adam Knight at EuroSTAR
Adam Knight at EuroSTAR

Me and Adam have often been at the same conferences and seem to share a similar career trajectory from testers to “anything that the company needs to get software shipped”. For me this resulted in engineering manager, for Adam it’s resulted in looking after testing, support, technical documentation and release management!

This year at EuroSTAR Adam spoke about on the challenges of testing a Big Data product in agile.

Adam’s outlook on testing is fabulous and he’s clearly a brand evangelist and someone who’s intent on evolving the testing to meet the needs of the company he works for, Rainstor.

Adam is also a good laugh to be around and very much someone I like to sit down with a beer and chat to about testing, life and work.

Adam is on Twitter – https://twitter.com/adampknight

Adam also has a terrific blog – http://www.a-sisyphean-task.com/

EuroSTAR – Star – Kristoffer Nordström

This is a short series leading up to Christmas where I feature a tester that was at EuroSTAR 2014, who I personally believe you, my readers, will benefit from knowing.

Today is Kristoffer Nordström.

Kristoffer Nordström at EuroSTAR 2014
Kristoffer Nordström at EuroSTAR 2014

I first met Kris about 4 years ago at a EuroSTAR conference – Amsterdam I think. We spent a night talking about testing, life and families and struck up a great friendship. Now getting together with Kris each year at a conference is a great treat.

Kris is the owner of Northern Test, his own consultancy. Kris is big in the Python for Testers world and offers some great talks and training on the subject. If you’re interested in learning to code Python then Kris is your man.

I attended Kris’ talk at EuroSTAR on Gamification – one of the best that I saw at the conference. Kris was talking about how a company was looking at ways to improve feedback from users – add gamification. Kris’ message was insightful, interesting and incredibly balanced. Kris warned against communicating the need for gamification with power and turning what can be a good game in to manipulation. Great talk.

Kris always has his Moomins with him. Here’s a post from earlier this year about the Moomins.

You can find Kris on Twitter – @kristoffer_nord

His website is – http://www.northerntest.se/

EuroSTAR 2014

It’s been a week since the EuroSTAR 2014 conference kicked off and I’m still trying to digest everything I learned.
R0001330_1680x1050
This years conference was one of the best I’ve been to for a number of reasons.
The talks were very forward looking in some respects – they seemed to show me the future of what testing will be (Internet Of Things, Lean and Kanban in testing, Mobile testing challenges, DevOps).
It was my first proper Keynote which pushed me even further out of my comfort zone.
The vibe was really relaxed. There were lots of community activities and lots of discussions happening around the event.
R0001360_1680x1050
  R0001330_1680x1050
The Test Lab seemed a lot busier than usual and was back to back with the community hub meaning it became a central place for discussions.
R0001359_1680x1050 R0001329_1680x1050
The social events were really good.
A lot of my good friends in testing were there this year – and this makes a huge difference (a lot of my good friends weren’t able to make it this year though sadly) 🙁
R0001399_1680x1050
R0001362_1680x1050
Last year and this year seemed to show more clearly the move I’ve personally seen in the industry – a shift away from the Test “phase” to a more continuous testing lifecycle. A lifecycle where the testers do less testing at the back-end of the project, but they have way more input in to the design, code and building of the software.
This highlighted itself quite clearly in the discussions I had with many of my peers – we find ourselves in a strange place where we do a wider range of activities than testing (management, leadership, recruiting, strategy, support, marketing, evangelism of our company’s brand) – basically whatever it takes to ship great software and make the business a success. Is it as simple as start-up versus established company? I don’t think it is. I’ll be exploring this idea next year on this blog.
R0001331_1680x1050 R0001354_1680x1050 R0001334_1680x1050
Next year EuroSTAR is in Maastricht, Netherlands and is chaired by Ruud Teunissen – the theme is “Walking the Testing Talk”. I’m very much hoping I can make it to next years conference. The call for papers is out now – and the most cost effective way to get to a conference – is to present at it 🙂
 R0001351_1680x1050 R0001338_1680x1050
Some more photos can be found over on my g+ page

SIGIST – It’s better, but is it enough?

The BCS SIGIST (Special Interest Group In Software Testing) was the first conference event I ever attended. It was about 6 years ago and I remember being amazed that people actually got together to talk about testing.

SIGIST was where I first saw Michael Bolton, James Whittaker, Dot Graham, James Lyndsay and Julian Harty. There were also plenty of vendors and lots of interesting discussions in the pub after.

In my opinion though the quality of the talks soon went downhill and I soon stopped going.

There were alternatives that I felt offered better value and a more consistent and relevant set of talks. Meetups were better attended and more sociable. The UKTMF had more interesting topics and more discussion. The London Tester Gathering had more beer. The SIGIST just felt old fashioned in a rapidly changing conference world.

Yesterday I went to the SIGIST event after a three year hiatus. And it was good.

It wasn’t amazing, but it was good.

The event is now held at The Barbican Centre which, for me personally, is a lot trickier to get to than their previous home at RCOG centre. It’s a personal thing, but it was an opinion shared by a couple of other attendees.

The food, drink, rooms etc were good although the rooms were not easy to find and a distance from the main networking area – cue people getting lost.

There was one vendor stand selling some sort of TMMI service so nothing to see there. 🙂

I reckon there were 50(ish) attendees – not a great turn out and in a big(ish) lecture hall it felt woefully under attended.

The annual AGM was also the opening part of the day. If you’ve never been to an AGM (at any organisation) then keep it that way. It was formal and boring, but essential for the BCS I guess….

An introduction from Stuart Reid kicked things off. A great keynote by Declan O’Riordan then followed. Declan talked about being an assertive tester.

Declan’s interest is in security testing and he talked about how a lack of assertiveness can lead to a lack of good security testing. He also talked about how lack of assertiveness can manifest itself, such as in people being Passive/Aggressive. Declan talked about some interesting behavioral facts whilst tying it to testing. Declan’s got a really easy presentation manner which added to the talk.

After a networking slot it was time for my morning workshop with Paul Gerrard. The main hall continued with track sessions.

Paul presented his ideas about a new model for testing. You can read more about the model here.

The session was good. Paul spoke about modeling our testing and presented his new model of testing. It’s worth reading and applying it to your own thinking to see if it works. Paul’s after feedback about it.

We use modeling as an aid for creating tests here at NVM but we don’t use it as a way to agree scope – something I’m sure we’ll start doing. We also don’t share our models across teams which is something I’d like us to try more of. So I took a lot away from Paul’s workshop.

After a decent lunch I went to the presentation tracks.

Up first was Russell Gallop who did a good presentation on Agile and CI for embedded software but it was clear he was at the wrong event. No one in the audience was working on embedded software and few of the audience were programmers.

It’s a shame when this happens as the speakers put a lot of time and effort in to presentations only to be presenting to the wrong audience.

Next up was David Orr who presented about “Testing Your Mind”.

I’ll be honest I was expecting something different but it was good. David spoke about his experience of working in three different changing contexts and how each one challenged him. I did expect it to be about the way testing can become frustrating and stressful sometimes and how our minds cope with it. But it was more about how the testing changed to cater for this changing process and environments. Not a bad thing but a mis-match in expectations.

Then came a discussion panel.

I must admit, although I have deep respect for all the people on the panel, I didn’t get much from this part of the session. The panelists are all great but they are all similar in the way they think about testing. The panelists were Tonnvane Wiswell, Ole Hansen, Mike Jarred and Declan O’Riordan.

They shared common values and ideas about the testing topics discussed. This resulted in four people all agreeing with each other and not being able to add much to what the previous panelist had said.

I want to see some Jeremy Kyle sort of panel with chairs thrown and security getting involved. Only joking, but you get the point. A panel discussion should be just that – a discussion – and there wasn’t much of that. Not the fault of the panelists. Maybe just needed a divisive topic or more diversity in panel members.

There was also no audience participation in the discussion. This assumes that the audience (for who’s value the panel is there for) had nothing valuable to add.

I like panel sessions where a seat is available for audience members to chip in with opinions and observations – a Goldfish bowl approach.

It also felt like the initial topic was one possible solution to a bigger problem, rather than the underlying real problem. Starting with a possible solution leads to discussions around this topic rather than the real problem.

“How do we get testers involved in earlier stages of projects?” was the opening gambit. A common question in the community. But what problem is that solving?

  • Why do people feel the need to involve testers earlier in the project?
  • What problem will that solution solve?
  • What problems will it create?
  • Will it work? Why might it not work?
  • Are there other people we should involve earlier?
  • What does early mean?
  • What about those who work in a collaborative agile environment already?
  • What benefits have been seen by early collaboration?
    Yada Yada.

Like I said, the panelists were great but the panel session just didn’t seem to work. Not enough discussion and not enough rigorous debate for my liking. 🙂

More coffee.

Then Paul Gerrard closed with a presentation about the Internet Of Things. This was good. It’s a topic I’m super excited about. It’s an infrastructure and network of devices that I think the testing community are ill prepared to test.

Paul talked about what the Internet Of Things is and what it means for society. He talked about the problems with it, the good it can bring and of course, the challenges we will face testing it.

How will we test it? Can we even test it? What happens when we miss something? What impact will it have on society? How can we recover from an epic fail? It’s a fascinating topic.

Good talk and lots to think about.

The SIGIST face stiff competition in the testing community and I do wonder how long the event will remain a viable option to run.

Meetups are free, better attended and more social.

Other conferences are much better value for money, more modern and more engaged with the community.

Online courses and other resources are more convenient for learning.

And of course, social media has replaced the need for many people to meet face to face anymore.

Can the SIGIST position itself somewhere in the market and regain it’s audience?

See you at……Nordic Testing Days

I’m heading to Nordic Testing Days (NTD) in a few weeks.

It’s not a conference that normally pops up on my radar. But this year is kind of different. I’ve got two of my team speaking at NTD.
nordic
Both Dan Billing and Raji Bhamidipati are presenting and I’m heading along to support them both (not that they need it) but also to learn from what looks like an interesting line up of topics.

It’s funny how conferences slip past you but on deeper inspection they look like they will be a great place to meet people and learn more about testing.

I’m very much looking forward to attending both Dan and Raji’s sessions and I have no doubt both of them will deliver insightful and informative presentations. No pressure 🙂

It’s still not too late to register for Nordic Testing Days….

 

Moving to Weekly Releases – Webinar

On 25th March I’ll be doing a webinar for EuroSTAR.

PIE

The webinar is entitled “Moving to weekly releases” and instead of repeating the presentation I delivered at EuroSTAR last year I thought I would open this topic up early for questions – that way I can tailor the webinar to try and answer some typical questions around this topic.

So here’s the basic outline I’ll cover:

  • We took our deployment lifecycle from 8 months to 1 week in about 1 year.
  • I stopped using any sort of test completion metric as a marker for done/not done or coverage (pictured)
  • We made testing the centre of everything.
  • We adopted lots of automation.
  • We started using metrics to help us test.
  • We put out lots of fires that most people run away from. (We went around a few also, and bought some fire proof suits).

It was a hard journey and many of us still wear the battle scars.

I’ll be walking through some of the reasons why this change was possible, some of the changes we made and some of the challenges we faced.

If you’d like me to focus on specific areas or have specific questions I can cover off then please do get in touch via Twitter @rob_lambert or email me : questions@thesocialtester.co.uk

Getting Hired – At Conferences

One of the things that I have observed from a number of testing conferences is that none of them have any sustained focus on hiring or getting hired *.

There have been one or two sessions about the topic of hiring but nothing sustained.

The occasional tracks that I have seen have been mostly focused around the hiring strategies of big corporates where bums on seats is sometimes more important than cultural team fit.

Most testers don’t know how to get hired – I wrote a book to help bridge that gap. Those that do know how to get hired are truly in the minority and appear, at least on the surface, to be overall better testers. Mostly this is not true – they are good, but they are often no better at testing than others, it’s just they are much better at getting hired. Getting hired is a skill.

Hiring and getting hired is a vast topic and one which is fraught with contextual challenges, but I believe that a dedicated set of talks from hiring managers from a wide variety of contexts, and maybe some sessions and tutorials on writing CVs, interviewing etc would go down well at most testing conferences. It’s great being good at testing but how do you then go on and get hired…

There are supporting topics such as social profiles, writing clear CVs, networking, self education and interpersonal communication that might also make interesting tracks. Or maybe they wouldn’t. Maybe people go to testing conferences to learn about testing and not the other stuff that comes with our working world…

What are your thoughts?

* The conferences that I have been to

EuroSTAR 2013 – Thanks

EuroSTAR

Well that’s it. Another EuroSTAR done. I’m writing this on the plane home after a tiring few days in the amazing city of Gothenburg, Sweden.

This years EuroSTAR was a really great event.

I saw a great deal of fun and excitement about the industry. I also saw a great deal of standardised thinking, some very shady conclusions from potentially unsound data and a whole lot of reasons to think we’ve still got a long way to go to see deep changes to the way we build and test software. But that’s the sign of a good conference – diverse opinions and much food for thought.

The social side of the event was excellent. There were more social events put on by the organisers (who did a sterling job as usual) and I think the location of the event meant a significant number of hotels were very close by; this kept most people together for the socialising. After all the session track topics are often the fuel for the discourse and deep learning that happens in the bar over a drink.

This was also the first year that I have done a presentation at EuroSTAR. I was deeply worried and very nervous but the presentation went well and the feedback was overwhelmingly positive.

As with most conferences I knew a lot of people from Twitter who I finally got to meet in person. I also met lots of new friends whilst solidifying existing relationships further.

I’m looking forward to next year. DUBLIN! If you do get a chance to go to EuroSTAR I think you’ll enjoy it.

The rise of the technical tester, and other thoughts from the UKTMF

The UKTMF this week was very good. Lots of interesting discussions to be had.

Being at the event triggered some thoughts about testing that I have been meaning to air for some time now.

Mastered Functional Testing

There were strong arguments by a few influential people that we, as testers, have mastered functional testing.

I don’t agree. I think we’ve got lots to learn.

I don’t believe we can ever master something as everything is forever changing; software changes, contexts changes, we change.

There’s also no single source of right (despite what some certification boards may tell you), which means that we’ll never truly know whether we have ever mastered functional testing.

I think we have a long way to go as testers before we can say we’ve got close to mastering functional testing.

Testing is only done by testers

This is a common misconception we have when talking about testing; that testing is done just by testers.

We are quick to put the tester at the centre of the development lifecycle when in reality we are just part of a wider team. The team will typically do testing of many different types.

The discussion at UKTMF centred around technical testers and the assumption was that technical testing was done by technical testers. Not so for many companies. We need to separate out the role of tester and the act of testing. Developers test, customer test, product owners test, testers test. Not all testing needs to be done by testers. Therefore not all technical testing needs to be done by technical testers.

All products must have uber-quality

Many statements and discussions centred around the product quality and how it must always be amazing. This is not true.

As much as we would like to think all companies need testers (is this us putting testers at the heart of the process?) and need to create high quality products it’s simply not true.

There are many companies producing software with no “testers” (although they do testing) and there are many products on the market (doing well) that aren’t (or weren’t) great to use.

Products, like humans, go through life stages and quality isn’t always important at all life stages of a product. Sometimes market share or a proof of concept is more important than a quality product.

Early adopters of Twitter will know the “Fail Whale”. Years ago about 3 in 6 tweets I posted would result in a Fail Whale – yet Twitter itself is now thriving, and I’m still using it. I’m happy to live with poor quality (for a period of time) if the problem the product is solving is big enough.

We often lose sight of the context the product is used in when talking about testers and testing. We are not the centre of the Universe and we are not always required on all products.

Testers are not generic

Another point I observed was that many people spoke about testers as though they are resources. Just “things” that we can move around and assign to whatever we want and swap as we please.

It was refreshing to hear a few people telling stories about the best teams being made up of a mixed bag of people from many different backgrounds, but they (we) were in the minority. A good team, in my experience, has people with very different backgrounds and approaches. We are not carbon copies of each other.

Testing (Or more to the point, social science) is not technical

Many people seem to be quick to replace the word”technical” with “being able to write code”.

It’s almost as though “programmer” is a synonym of “technical”.

So specialists in exploratory testing, requirements analysis, negotiation, marketing, technical writing (there’s a clue in the title) and design are not technical…..really?

It actually lead to a great comment from Richard Neeve who suggested that actually a technical tester (and tester in general) could actually be described as a scientific tester. I like the thinking behind this and I need to ponder it further.

The whole “technical tester” discussion is interesting as it often takes hours just to agree on a general definition of exactly what a technical tester is – something Stefan Zivanovic facilitated very well.

And finally I thought it prudent to mention the views I aired at the UKTMF on why “technical tester” is becoming increasingly common and prevalent in our industry.

I believe that the clued up testers are the ones who realise the future is bleak for those who simply point and click. A computer is replacing this type of testing.

The switched on testers are learning how to code, specialising in areas like accessibility, training and coaching, UX, performance, security, big data and privacy or they are expanding their remit in to scrum master, support, management, operations and customer facing roles. They are evolving themselves to remain relevant to the market place and to make best use of their skills and abilities. Some of these people are differentiating themselves by labelling themselves as technical. I believe this is why we are seeing a rise of the technical tester. And just as a side note, not all of those who call themselves a Technical Tester are all that technical (in either coding, or other technical skills). As with all job labels there are those who warrant it, and those who are over inflating themselves.

I think it’s important that testers diversify and improve their technical skills (and I don’t mean just coding skills) to remain relevant. After all, if you’re interviewing for a tester and they have excellent testing skills then you’ll likely hire them (assuming all other requirements and cultural fit is right). But what if a similar candidate came along who was also excellent at “testing” but was also skilled and talented in UX, ethnographic research, performance testing or security testing……which one would you want to hire now?

Remaining relevant in the marketplace isn’t about getting another certificate, it’s about evolving your own skills so you can add even more value than simply pointing and clicking. You need to become a technical tester (as defined above) – in fact, scrap that – you just need to become a good tester – because after all, good testers are indeed technical.

http://www.uktmf.com