What’s the one engineering practice your team claims to value… but consistently avoids when under pressure?
Posts by Modern Software Engineering
We have a new addition to our highly regarded "TDD & BDD: Design Through Testing" course that will revolutionise the way our students learn on our training site... 👇
Below are the 14 markers for successful Continuous Delivery within an organisation...
Little bonus game... can you guess which one of these I think is the most important to get right? 👀
🧵 1/5
I worked on an XP team once where we genuinely had 100% test coverage. It was awesome, we could make changes so fast, and yes we still had the occasional live defect
100% Test Coverage is a LIE, Here’s Why... | Sam Newman
📽️ AVAILABLE NOW
WATCH HERE ➡️ youtu.be/p1xZ-Ni2t8Q
100% Test Coverage is a LIE, Here’s Why... | Sam Newman
📅 TONIGHT @ 7PM
Subscribe & turn on notifications 🔔 so you NEVER miss an upload! ➡️ youtube.com/@ModernSoftw...
Your AI is a machine and the junior dev is a human. So treat the AI like a machine and your junior dev like a human.
@modernswe.bsky.social ❤️
youtu.be/fHfkJRh2veg
I updated my personal site! I've been wanting to give it a "proper" design overhaul for years, but the time investment needed meant that I kept putting it off. Thanks to Claude, and to having a bit more "free" time right now, I finally did it! Take a look:
trishagee.com
The Junior Developer CRISIS: How to Build a Team When AI Does the Entry-Level Work | @tastapod.com
📽️ AVAILABLE NOW
WATCH HERE ➡️ youtu.be/fHfkJRh2veg
The Junior Developer CRISIS: How to Build a Team When AI Does the Entry-Level Work | @tastapod.com
📅 TONIGHT @ 7PM (UK)
Subscribe & turn on notifications 🔔 so you never miss an upload! ➡️ youtube.com/@ModernSoftw...
Every software team will eventually face a heart-stopping mistake, whether it is a developer accidentally deleting live data during a migration or a junior ops engineer unplugging a production server to use a vacuum cleaner... 😬
Nice analysis by @kevlin.bsky.social about being a human using ai coding tools. His focus on the engineering and general principles rather the day-to-day tactics. youtu.be/vpYJMr1pJRY?...
If you had to deploy a single line of code to production right now, how much of that process is automated and how much relies on 'hoping' nothing breaks?
Why "bad things" happening is actually the best chance for a development team to learn.
This is the story of how a costly mistake taught me a BIG lesson about team culture in software engineering.
📽️ Available on @modernswe.bsky.social ➡️ youtu.be/Q0gxen_uXlw
The $1M MISTAKE That Taught Me THIS About Team Culture | @davefarley77.bsky.social 📅 TONIGHT @ 7PM
Subscribe & turn on notifications 🔔 so you NEVER miss an upload! ➡️
youtube.com/@ModernSoftw...
AI is the ultimate amplifier. If your org has Taylorists as senior leaders, you're in for a tough ride and I emphasise
@stevesmithtech.bsky.social goes deeper on this in this episode on the Modern Software Engineering channel, check it out here: youtu.be/kDBeFOscZpc
7/7
Measure AI prompt counts? Watch your developers artificially inflate their usage.
Measure the percentage of AI-generated code? Watch your codebase bloat with unnecessary classes and methods just to hit a quota.
6/7
When you obsess over individual activity metrics, you fall straight into the trap of Goodhart’s Law: When a measure becomes a target, it ceases to be a good measure.
5/7
You could merge 100 pull requests in a single day, but if your release process is bottlenecked by a centralized testing team and you only deploy to production once a month, you have delivered absolutely zero extra value to your customers.
4/7
This is just Taylorism wrapped in a shiny new algorithm.
A recent vendor report proudly claimed: "Daily AI users ship 60% more pull requests than non-users." Who cares?
Pull request throughput is not a unit of delivery. It is not a unit of business success.
3/7
Suddenly, vendors are falling over themselves to sell "developer intelligence" platforms that measure how many hours of AI coding a developer uses, or what percentage of a codebase is AI-generated.
2/7
Is AI actually dragging our industry 20 years backwards? We are currently in a Gen AI hype bubble, and it is bringing out the absolute worst impulses in engineering management.
🧵 1/7
Thanks for your support over the last 6 years 🥳
Rethinking The Software Lifecycle | How To Look After Software For The Long Term | @stevesmithtech.bsky.social
📽️ AVAILABLE NOW
Watch HERE ➡️ youtu.be/CBcdERAv5P8
AI is helping us generate code faster than ever, but how do you ensure you aren't just deploying bugs at record speed? That is exactly why Continuous Delivery remains the state of the art for software development, providing the essential safety net your team needs.
🧵 1/6
Rethinking The Software Lifecycle | How To Look After Software For The Long Term | @stevesmithtech.bsky.social
📅 TONIGHT @ 7PM (UK)
Subscribe & turn on notifications so you NEVER miss an upload! ➡️ youtube.com/@ModernSoftw...
People approach a legacy rewrite like they’re trying to swap an aeroplane engine mid-flight while standing on the wing. It’s madness. Why aren’t we talking more about diverting the flow at the aggregator level to prove the new path works before we touch a single line of the old mess?
What Does Continuous Delivery Look Like WITHOUT Automation? | @kevlin.bsky.social & @davefarley77.bsky.social
📽️ AVAILABLE NOW
Watch HERE ➡️ youtu.be/u2U1kWHuO7I
What Does Continuous Delivery Look Like WITHOUT Automation? | @kevlin.bsky.social & @davefarley77.bsky.social 📽️ AVAILABLE NOW
Watch HERE ➡️ youtu.be/u2U1kWHuO7I