Skip to main content


Showing posts from November, 2022


  Over on the Association for Software Testing Slack our book club facilitator, Zenzi Ali , is guiding us through Elisabeth Hendrickson's Explore It! Last week she posed this question: "When joining a new team/job/project, how do you determine the core capabilities of the software you’ll be testing?" I surprised myself a little with my answer. Not because of what I described, but because I hadn't mapped it out this way before. It was only when I stepped back to try to summarise that I noticed the pattern in my intuitive behaviour. So here's what I wrote, edited so that it's less of a stream of consciousness. --00-- Rob Meaney has a great talk, A Tale of Testability , where he describes how his CODS criteria (controlability, observability, decomposability, simplicity) were applied to a product he worked on to make it more testable.  Now, I'm not suggesting that someone walks into a new team and shouts about re-engineering for testability before they'

CAST 2022 Recap

  Tl;dr: it was brilliant being at CAST 2022 , the conference of the Association for Software Testing , in San Diego last week. Even if I didn't get out on the water like Tina, it was brilliant to be at an in-person conference again, with the time and space and atmosphere and context to really talk to people about the presentations, about themselves, about their work, and about all the software testing things. It was also brilliant to be at a conference with the theme hands-on , set up to actually be hands-on. We wanted attendees to leave with a set of tools that they'd at least held in their own hands and felt the weight and balance of. Day one kicked off with three extended tutorials from experienced practitioners. Dawn Haynes gave us A Survey of Test Design Ideas which covered test design, heuristics, sampling, checklists, and other aids to generating possible areas to test. For those in a rush to get going with automation ag

BCS: Testing, AI, and Diversity

I spoke at the BCS Software Testing Specialist Group's Testing, AI, and Diversity Conference yesterday. That's me at the top, holding an imaginary marrow. There'll be videos of all of the talks online shortly but in the meantime, here's my sketchnotes. Sam De Silva , An overview of the draft EU AI Regulation. Adam Leon Smith , Software Testing Standards - why do we need them and when are they useful? Alan Giles , Giving ‘The User’ A Face - Accessibility Testing Using Personas. Andy Shaw, Mental Health, Testing and Me. Deborah Reid , Accessibility 101. Laveena Ramchandani , Testing Data Science Models.   Jonathon Wright , Shift Right into the Metaverse with Digital Twin Testing.  Image: Jonathon Wright on Twitter

What a Performance

  At CAST 2022 Eric Proegler delivered a keynote speech packed with information, audience participation, and humour.   Anyone Can Performance Test?!? he said, and then showed a roomful of sceptical testers that he wasn't lying by having us hammer the WOPR web site while he monitored the back end. In a second round we gathered stats from our browsers's dev tools and shared the findings into a Google sheet. From that we were able to start seeing some potential patterns in the speed at which pages were loading on the client side for different kinds of devices and networks. We also got a hint that those accessing the server at different times might be experiencing different performance. Eric then introduced us to WebPageTest where, for free, you can request a device, a connection type, a geographical location, and a URL, push a button, and get back data on page load times, accessibility scores, and other metrics for your chosen combination.  He continued to monitor WOPR on serv

Learning to Test

In Tariq King 's extremely interactive workshop at CAST 2022 we got a quick introduction to Artificial Intelligence and Machine Learning as we played with some toy online systems and trained our own model (using a Teachable Machine ) for recognising people or objects in the room. We didn't cover testing these kinds of systems specifically, but applied testing skills when trying to understand which features were being used by the model of workshop participants that we built. What's more discriminating: hair, skin colour, proximity to the camera, background, lighting, clothing, something else? Tariq also pointed to GANs ( Generative Adversarial Networks ) as tools that should be of interest to testers since they seek to model and then challenge the system they are in contest with. Finally we discussed how AI could be of use to testers right now. Tariq gave an example of tooling trained using computer vision and reinforcement learning to navigate login pages generally rather

Listener at Work

  That's me sketchnoting at CAST 2022 , taken by Pradeep Soundararajan . He tweeted the image and got a reply that made me smile:  My grin was wide for a few reasons: First, there is no way I'd flatter myself with that title. Second, this is not really work; being at CAST is an absolute pleasure. Third, I actually redrew those notes shortly afterwards because I'd only filled half the page when the talk ended. I'll forgive myself the last point. I'd thought it was another 90-minute session when in fact it was only 45. Doh! But that highlights an interesting thing about sketchnoting: how to determine what to put on the page, and where, and when.  Several people at the conference asked me about that and I said the same kinds of things I wrote in Beginning Sketchnoting a few years ago. My basic approach hasn't changed very much at all. I have a small repertoire of images and I draw them shabbily with the biros I have to hand. What has changed since 2018 is my c

The How of the Why

At CAST 2022 Amber Vanderburg told us that when you want to make a change, particularly something innovative, start with the Why. Why? So that everyone involved can understand the motivation and direction of travel. I've learned that myself over the years too, from Simon Sinek , and it has the added benefit that conversations about the merits of the Why can be had before any implementation starts. Of course, being honest about what the Why is is important too, if you care to maintain trust over time. Amber suggested a few things that help to navigate the change once it's underway. The How of the Why, if you like: proactive conversation: some kind of formal structure, some way for all voices to be heard, perhaps with a time box to stop things dragging on. awareness of fundamental attribution error: where someone's apparent motive (e.g. antagonism) may not be their actual motive (e.g. sharing potential alternatives). strength and weakness alignment: being clear with one ano

You Wish!

We got a slice of Ben Simo 's backstory in his CAST 2022 workshop, Testing Without Requirements, last week.  It was a trip through working environments in which requirements had varying levels of detail, through Ben's evolution in thinking around the kinds of constraints that imposed on his work, and into a model of the testing landscape which recognises the dynamic nature of what is known and unknown during testing.  A couple of takeaways: Stated requirements are not all of the requirements ... and we can test that. Firm requirements are not always available ... but we can test to solidify them. The requirements that matter meet a human need ... and we can test to learn what they are. I guess I could summarise the presentation as: Can we capture all and only the user wishes? You wish! In the interactive section of the workshop we used the FEW HICCUPPS mnemonic to help us to identify unstated re

Your Recipe for Suck Less

In Cindy Lawless 's keynote at CAST 2022 we spent a few minutes testing a simple application as a group and then collaborated on a test report. The report was the kind of thing a team lead might be expected to provide to management before a software release in a traditional software shop but, with Cindy's guidance, we avoided dumb artefact counts, meaningless charts, and level-of-effort analysis rather than risk analysis. Cindy's report recipe is straightforward: summary, strategy, coverage, risks, bugs and other concerns.  She gave a nice summary of how this can be a simple and clean Slack message rather than the cumbersome slide deck that is often requested, and copy-pasted, although we did build a deck in the session. I think there are a handful of key points to take away: Management won't read it (all). Make the important stuff is concise, clear, and prominent.  The important stuff is what's important to them . What could affect business value? Don't let th

Access all Areas

I attended Cordellia Yokum 's Usability for everyone: Are you excluding some users from accessing your website? workshop at CAST 2022 .  As I can't do justice to a packed and interactive five-hour session in a short post like this, I'm going to simply drop bullet lists and links from the notes I took here. Accessibility Not just about disability Auditory, cognitive, neurological, physical, visual US: one in four have some disability Small screens, elderly users, slow internet, poor lighting conditions, etc SEO benefits from being able to access content Accessibility is not a project. Like Quality, improve it incrementally and continually Assistive technologies Screen magnifiers (e.g. web browser zoom) Text readers (e.g. blind and partially-sighted, ADHD, dyslexia sufferers) Speech recognition software  Head pointers, motion tracking (often for paraplegic users) Single switch entry (e.g. sip and puff mouth operation; keyboards with "quick buttons") POUR Perceivable

CAST Lean Coffee

Lean Coffee? Over breakfast? With bacon? At CAST 2022 ? Yes please! Here's my aggregated notes from that early morning munch'n'chat with a bunch of dedicated testers and also some bakers. How have you shifted left? One of the ways I've used is to ask developers to write test notes with each story. It worked well once I'd persuaded one and he advocated for it to the others.  If there are no notes on the story, I'll initiate a conversation about risk. We have an Amigos session early in development. We identify things to test with the story. Someone on the team picks that task up ... ... and it's usually me or the developer that did the coding. They are supposed to write notes, but the quality is pretty variable. I am shifting left by doing it.  I will initiate conversations with people who matter at any time. I show value so that people are interested in working with me next time. At my place I ask to be invited to conversations. This has been successful, exc

Satisfying the Requirements

The Association for Software Testing is crowd-sourcing a book,  Navigating the World as a Context-Driven Tester , which aims to provide  responses to common questions and statements about testing from a  context-driven perspective . It's being edited by  Lee Hawkins  who is  posing questions on  Twitter ,   LinkedIn ,   Slack , and the AST  mailing list  and then collating the replies, focusing on practice over theory. I've decided to  contribute  by answering briefly, and without a lot of editing or crafting, by imagining that I'm speaking to someone in software development who's acting in good faith, cares about their work and mine, but doesn't have much visibility of what testing can be. Perhaps you'd like to join me?   --00--   "Testing is just to make sure the requirements are met" Yes, that's a common perspective, but ... But I disagree with it given a commonly-held view of the requirements . Let's gloss that view as: some list of self-an