Posts Tagged ‘design’

I recently attended the regional Scrum Gathering for 2016 in Cape Town. This is a post in a series on the talks that I attended. All posts are based on the sketch-notes that I took in the sessions. 

A lot of this talk was a repeat of things I’ve heard before:

  • Efficient feedback increases effectiveness
  • We need to build things to learn things through measuring
  • We need to start with understanding user needs
  • When we define a problem, we’re usually guessing, so we need to validate our guesswork (through feedback loops) as quickly as we can

A wicked problem is a problem that one can only fully understand or define once one has found the solution to the problem. Software in a wicked problem: when we find the solution (note: find not build), then we can usually answer what the problem actually was.

One of the new ideas for me from this talk was the #ConstructionMetaphorMustFall idea. Traditional engineering models treat coding as part of the build process, however Jacques argued that code should actually be treated as a design document and that writing code actually forms part of the creative design process. The code equivalent of bricks and mortar is actually only the conversion of code to bits together with the (hopefully automated) checking process i.e. when we release. In this model, things like Continuous Integration and Continuous Deployment are actually design enablers: they allow us to test our theories and verify our design through feedback.

Shifting your mindset to view writing code as part of the experimental design process rather than execution of a final plan/design would probably lead to a paradigm shift in other areas too. I’ve been saying for a while that, as software tooling improves and more “routine” activities become automated, writing software is becoming less of a scientific engineering field and more of a creative field.

What are your thoughts on shifting the “coding” activity out of the build phase and into the design phase? What does this mean for how we should be thinking about code and feedback?

Recommended reading: SPRINT

I recently attended the regional Scrum Gathering for 2016 in Cape Town. This is a post in a series on the talks that I attended. All posts are based on the sketch-notes that I took in the sessions. 

As we’re also looking to move a very complicated, fragile system over to a new technology stack, I found this talk by @NigelBasel quite interesting. One of the most interesting parts was his application of Conway’s Law to the problem: the software had evolved and modeled when there were only a handful of developers working on the system; and now it needed to change to model the communication structures required between a number of teams working on the same codebase. He also showed us the output from a really cool tool (I think) called Gource which he’d used to model changes in their source code repository over time. Made me wish I could see the same animation for ours! I’m sure it would be fascinating!

Nigel gave a suggestion of two steps one could take when faced with this legacy code-wool problem:

  1. Stop digging (yourself into the hole)
  2. Get out of the hole (carefully)

They’re still progressing on their journey, but these are the steps they’ve taken so far to try and bring their code base back under control:

  1. They identified and fixed any coincidental cohesion – they moved parts of code which logically belonged together and just happened to be where they were because they were written into things like libraries and services.
  2. They shifted their thinking of layers to services and considered things like whether certain services like authentication be provided by 3rd party tools and removed from their code base. If a service seemed to make sense, they created it and then migrated functions as they were needed, thereby “starving” the old code base.
  3. The considered their code base in terms of business features and the data required and tried to group these together
  4. They write all their new code using their new strategy (as far as possible)

One issue Nigel admitted that they haven’t really got to grips with yet was version control. He emphasised that their journey is not yet done. I’m hoping we will hear more about their adventures and learnings once they have traveled their path. Did you find any of these points helpful? Do you have experience changing old code to reflect new organisational communication patterns?

This is a post in a series on the talks that I attended at my first ever Agile Africa Conference in Johannesburg. All posts are based on the sketch-notes that I took in the sessions. agileafrica

Despite its short length, this was an interesting talk for me. There were two new ideas that I’d not been exposed to before and quite liked. The first was the term “Most Loved Product” (MLPLOL-48). A MLPLOL-48 (rather than MVP) is a slice of a Product that can be used to gather feedback about the user experience i.e. does my Product makes my customer smile? The important part about a MLPLOL-48 is to not forget the smiley!

CaptureSecond, and largely related, was the concept of what a MLPLOL-48 vertical slice needed to consist of. Whereas vertical slices traditionally relate to the stack on which the Product was built, the UX slice needs to include the functional, reliable, usable and emotional design layers to be a full vertical slice.

The presenters also touched on user feedback and how it was important to consider all aspects of the user experience: what they hear, see, say, do, feel and think. They did caution, though, that “not every colourless liquid in a glass is water”, so although user feedback can tell us what they are hearing, seeing, saying, doing, feeling and thinking; we need to probe a little deeper to determine WHY they are having that experience.

Have you ever heard of a MLPLOL-48? What are your thoughts on the UX vertical slice concept?

I recently attended the regional Scrum Gathering for 2015 in Johannesburg. This is a post in a series on the talks that I attended. All posts are based on the sketch-notes that I took in the sessions.checking

Initially, this talk was called “Automated Testing”. Apparently testing can not be automated because testing implies trying to break things when you don’t know what you’re looking for. The preferred term for when code can automatically check itself is “Automated Checking”.

Joshua had an interesting perspective in terms of a test approach. It’s based on the premise that the most important thing, at the end of the day, is the value delivered to the user and this is where all testing should start. His proposed approach is an outside-in approach:

  1. Identify tests/checks that confirm system behaviour and that the output produced is what the user would expect given the inputs.
  2. Once you have identified these checks (typically at the user interface level) then find the highest level where you can write automated checks without increasing the cost to a point where the cost of automation exceeds its benefits.

He did acknowledge that this kind of outside-in approach will not help one find where things are breaking. However, there are certain other benefits to consider when it comes to high level checks:

Benefits Table of High Level Checks

For those who are strong advocates of Test Driven Design, that did get a mention. Joshua acknowledged that when writing complicated pieces of code where there is a large amount of uncertainty, unit tests could be helpful in finding one’s way. He called these Scaffold Tests and drew the analogy to the scaffolding used when constructing tall buildings. Similarly to scaffolding, remove these tests once they have served their purpose, otherwise they become an unnecessary maintenance overhead.

What is your take on an outside-in rather than inside-out approach to automated checking?

StoriesThis photo is the list of actions/changes/learnings one of my teams came up with in their most recent retrospective. This did not come from someone who went on training or read an article. It also didn’t come from a new Agile coach or Scrum Master. It came from them missing their sprint commitment and goal. This team only managed (on paper) to complete 8 out of 18 points; but they all knew they had delivered and learned a lot more than that measure reflected.  Here are some things that they decided to do going forward:

1. If the team cannot reach consensus about the size of a story, then split it into two stories and size the smaller stories

One of the main reasons the team had such a poor burn-down is that they took in one quite large story which did not quite meet the INVEST requirements. For one, it was a ‘common component’ that was to be used in most of the later stories (so not independent). It also was not small enough – and turned out to be even bigger than the team had thought. During sizing, there had been some debate about its size and eventually reluctant consensus was to make it the smaller size. Turns out the less optimistic team members were right. This was one of the stories that was not done at the end of the sprint.

2. Keep Planning II – and use it to verify the sprint commitment

This is a team that often decides to skip Planning II (I don’t like it, but ultimately it is the team’s decision and so far we’ve muddled along without it). For this sprint, they decided that they did need a session to unpack the stories and how they would be implemented. Everyone agreed that without Planning II we would have been even worse off. They also realised that at the end of Planning II, there were already some red flags that the big story was bigger than everyone had thought and they could have flagged,  at that point, that the commitment for the sprint was optimistic. The team agreed that, in future, if going into the detail during Planning II revealed some mistaken assumptions, then the team would review the sprint commitment with the Product Owner before kicking off the sprint.

3. Feel free to review story-splits in-sprint

Early in the sprint, the team were already aware that the big story was very big and probably could be split into smaller components. Their assumption was that this wasn’t possible once the sprint had started. For me, re-visiting a story split mid-sprint or once you start the work is not a bad thing: sometimes you don’t know everything up-front. It also, in the case where a story is bigger than expected, gives the Product Owner some more wiggle room (negotiation part of INVEST) to drop/keep parts of the story to successfully meet the sprint goal. Of course, where we have got things really wrong, then sometimes the sprint goal cannot be rescued and the sprint would be cancelled.

4. Raise issues as they happen

Pretty much summarises most of the above points. One of the agile principles is responding to change over following a plan, so when change happens make it visible and decide how to adjust the plan at that point. There’s no point in battling on as planned when you know that the planned path is doomed to fail.

Some references:

Dual Track Scrum

Dual Track Scrum

This week I stumbled across an article which referenced something called “Dual Track Scrum” (see the links below). Intrigued, I researched a little more and was fascinated to discover that there was a documented process/approach to product development and delivery that was very similar to what had evolved for a product team I previously worked with.

This isn’t the first time this has happened to me. The last significant deja vu experience was when I found that there was a name for a software development approach that included: time boxing; daily team check-ins; planning and estimation as a team; work defined on a physical card on a white board; a definition of done for the time box that included all delivery activities; and a list of features that could be traded in and out if not yet started. Yes, that was the day I discovered that the approach our team had “created” to successfully deliver an off-shore project was, actually, called Scrum.