
@sprky@mastodon.social - Agile coach, recovering developer and part time tinkerer. Co-founder of the awesome Cambridge Agile Exchange (@Cam_Agile_xchg)
Cambridge, UK
Joined May 2008
- Tweets768
- Following680
- Followers212
- Likes315
Marcus Williams retweeted
A QA engineer walks into a bar. Orders a beer. Orders 0 beers. Orders 99999999999 beers. Orders a lizard. Orders -1 beers. Orders a ueicbksjdhd.
First real customer walks in and asks where the bathroom is. The bar bursts into flames, killing everyone.
Marcus Williams retweeted
This is the Facebook playbook: you lure in publishers by promising them a traffic funnel ("post excerpts and links and we'll show them to people, including people who never asked to see them"), and then the rug-pull: "Post everything here, don't link to your own site"
Replying to @MarioNawfal
Interesting.
Our algorithm tries to optimize time spent on X, so links don’t get as much attention, because there is less time spent if people click away.
Best thing is to post content in long form on this platform.
Marcus Williams retweeted
Instant and delayed rewards from writing tests may disappear when we write tests last.
This means we'll be less likely to write tests.
What are these disappearing rewards? Short thread.
Marcus Williams retweeted
i hooked up chatgpt to a furby and I think this may be the start of something bad for humanity
Marcus Williams retweeted
Unpopular opinion: Pull requests are a terrible idea in almost all contexts.
Focus on finishing things as a team instead of keeping everybody busy. Do code reviews in real-time through pairing or ensembling instead of blocking the flow of work waiting for a pull request review.
Marcus Williams retweeted
In Paris, a new “Everything Is Fine” gif is born teapawt.org/marjo_beb/status/16395…
Marcus Williams retweeted
Can GPT-4 *Actually* Write Code? tylerglaiel.substack.com/p/c…
Marcus Williams retweeted
"Hotfix" is just a euphemism for "our regular process is so slow that we can't use it in a situation where we would actually need it most urgently."
Marcus Williams retweeted
One of the oddest agile software development dysfunctions I see so often: same old big upfront solutioning with a delusion of certainty, and then squeezing that solution through Scrum with no feedback, no working product increments, and no retrospection. Like … what’s the point?
Marcus Williams retweeted
There's a comparison between writing maintainable code and playing pool/billiards.
If you're bad at shooting pool like I am, you make the easiest shot you can, hit a ball into a pocket, and the cue ball rolls wherever it rolls on the table. 1/7
Marcus Williams retweeted
Committing at least once a day, reduces the blast radius of your changes: if it's good you can build on it more quickly; if it fails you find out and fix it more quickly.
Marcus Williams retweeted
Huge thanks to our wonderful CAE community for all your support over the years. There won't be any meetups scheduled for the forseeable...
...but we're still here! Find us on Slack or contact the CAE crew @sprky @simolaw @Labyrinthsprite
Marcus Williams retweeted
In output-first planning, present is clearer than future: we only know roughly what features we will build in Q4 but here are our Q1 user stories
In outcome-first planning, future is clearer than present: our Q4 goals are set, in Q1 we will look for the right paths to reach them

ALT Output driven planning - “We focus our energy on getting clarity from stakeholders about what outputs we need to deliver first, while we have only a rough idea what outputs we might be able to deliver later.” Outcome driven planning - “We start by aligning on what the desired customer outcomes and business impacts look like. Rather than commit to specific deliverables up-front, the team is able to autonomously set their experiments from Day 1.”
Marcus Williams retweeted
Love this! A math based argument for pair programming!!
<paraphrase> "The smaller the PR, the bigger the cost of the wait time", The bigger the PR, the lower the impact on quality" - we need to change the game, this is a great argument for pair, or mob, programming!
Interesting talk from NDC Copenhagen.
Smaller PRs cumulatively take longer when you factor in lost throughput, dependencies, and delays in async code reviews. Little's Law shows up too.
@tdpauw @davefarley77 👋
youtu.be/fYFruezJEDs
Marcus Williams retweeted
To me the "CI/CD" pipeline starts with conversations you have with customers to uncover their stories and ends up with delivering software into their hands. The mechanics of an automated pipeline that takes over once the code is checked in is the least part of that. 1/2
Marcus Williams retweeted
#AgileMaxim 1: It is in the doing of the work that we discover the work that we must do. Doing exposes reality.
Marcus Williams retweeted
Thank you to all who came along to #CAEmeetup today. We well and truly broke the ice, shared a lot of smiles and learned lots of ways to get our teams going and ready 🧊🧊🧊
























