"Vibe coding is fine, but only if you pay close attention to what is being built. We can't just pull a lever on your IDE and hope that something good comes out."

That's not software engineering. That's gambling."

#WernerVogels, 2025

https://www.implicator.ai/werner-vogels-hands-out-newspapers-at-his-likely-final-re-invent-the-man-who-built-the-cloud-isnt-done-teaching/

But ... when we say "vibe coding" we're specifically talking about *not* paying attention to what's being built. Usually because of lacking the base knowledge to understand it.

(1/3)

#SoftwareEngineering #VibeCoding

@urlyman

Has the cost of building software just dropped 90%?

https://martinalderson.com/posts/has-the-cost-of-software-just-dropped-90-percent/

#HackerNews #costofsoftware #softwaredevelopment #technews #startupinnovation #softwareengineering

Why Fighter Jets Ban 90% of C++ Features [video]

https://www.youtube.com/watch?v=Gv4sDL9Ljww

#HackerNews #FighterJets #C++Features #SoftwareEngineering #VideoAnalysis #AviationTech

The Anatomy of a macOS App

https://eclecticlight.co/2025/12/04/the-anatomy-of-a-macos-app/

#HackerNews #TheAnatomyOfAMacOSApp #macOS #Development #AppDesign #TechInsights #SoftwareEngineering

This video from a team lead at Google is fascinating.

Here's someone who has clearly seen the negative consequences of LLM adoption in software teams. He seems like a conscientious manager, talking about how it makes some people struggle, or resistant, and damages the social fabric of a team.

Yet, he's stuck in a frame that "AI is the future of work" and "AI makes better decisions," so everything he has to say about it is in terms of how to drive adoption, even when, he admits, it may have serious negative consequences.

The cognitive dissonance is striking, and familiar. He's working very hard to justify this policy of LLM use. He wants to "put his people first." The paradox comes from trying to force both of these views at the same time.

What if we just didn't force LLMs on people? What if it's not a net benefit? What if overcoming people's resistance means getting worse results? It feels like he won't let himself consider that possibility.

#llm #softwareengineering

I ignore the spotlight as a staff engineer

https://lalitm.com/software-engineering-outside-the-spotlight/

#HackerNews #staffengineer #spotlight #softwareengineering #careerdevelopment #professionalgrowth

This video from a team lead at Google is fascinating.

Here's someone who has clearly seen the negative consequences of LLM adoption in software teams. He seems like a conscientious manager, talking about how it makes some people struggle, or resistant, and damages the social fabric of a team.

Yet, he's stuck in a frame that "AI is the future of work" and "AI makes better decisions," so everything he has to say about it is in terms of how to drive adoption, even when, he admits, it may have serious negative consequences.

The cognitive dissonance is striking, and familiar. He's working very hard to justify this policy of LLM use. He wants to "put his people first." The paradox comes from trying to force both of these views at the same time.

What if we just didn't force LLMs on people? What if it's not a net benefit? What if overcoming people's resistance means getting worse results? It feels like he won't let himself consider that possibility.

#llm #softwareengineering

You Can't Fool the Optimizer

https://xania.org/202512/03-more-adding-integers

#HackerNews #You #Fool #the #Optimizer #optimization #programming #technews #softwareengineering

You Want Microservices, but Do You Need Them?

https://www.docker.com/blog/do-you-really-need-microservices/

#HackerNews #Microservices #Development #Architecture #SoftwareEngineering #TechTrends

Good engineers write bad code at big companies

https://www.seangoedecke.com/bad-code-at-big-companies/

#HackerNews #Good #engineers #write #bad #code #at #big #companies #badcode #softwareengineering #bigtech #programming #HackerNews

2/n

* Context compaction: documentation vs. code.
* AI rift: part of the team doesn’t believe in AI, and the other part uses a lot. The two sides hate each other more and more.
* PR is for mental alignment (among humans).
* And Memento is the best movie for context engineering!

Human Layer CEO Dexter Horthy on Advanced Context Engineering and avoiding the “dumb zone”

https://www.youtube.com/watch?v=xmbSQz-PNMM&t=2408s

#softwareEngineering #coding #genAI #talk

1/

Highly recommending this talk from AIE/CODE 2025 on how to code effectively as a team with AI:

* With coding agent, 40% of your context window is “smart zone”. Once it’s passed, expect more slop.
* How do you know you should start a new context (task): when your coding agent says “You’re absolutely correct!”
* Research, Plan, Implementation (RPI).
* Leverage-wise: R >> P >> I. Screwing up upstream costs big time.

#softwareEngineering #coding #genAI #talk

2/n

* Context compaction: documentation vs. code.
* AI rift: part of the team doesn’t believe in AI, and the other part uses a lot. The two sides hate each other more and more.
* PR is for mental alignment (among humans).
* And Memento is the best movie for context engineering!

Human Layer CEO Dexter Horthy on Advanced Context Engineering and avoiding the “dumb zone”

https://www.youtube.com/watch?v=xmbSQz-PNMM&t=2408s

#softwareEngineering #coding #genAI #talk

1/

Highly recommending this talk from AIE/CODE 2025 on how to code effectively as a team with AI:

* With coding agent, 40% of your context window is “smart zone”. Once it’s passed, expect more slop.
* How do you know you should start a new context (task): when your coding agent says “You’re absolutely correct!”
* Research, Plan, Implementation (RPI).
* Leverage-wise: R >> P >> I. Screwing up upstream costs big time.

#softwareEngineering #coding #genAI #talk

#oop #programming #GUI #commonLisp #McCLIM #softwareEngineering
Article in which I show simple multiple inheritance and method qualifiers in ANSI common lisp's common lisp object system, then show how the McCLIM implementation of the common lisp interface manager 2 specification turns the objects I was working with into rich general user interfaces with a minimal declarative effort presupposing that I was presenting lisp objects.

Rich interfaces pervade lisp's history.
https://screwlisp.small-web.org/clim/basic-principle/

#oop #programming #GUI #commonLisp #McCLIM #softwareEngineering
Article in which I show simple multiple inheritance and method qualifiers in ANSI common lisp's common lisp object system, then show how the McCLIM implementation of the common lisp interface manager 2 specification turns the objects I was working with into rich general user interfaces with a minimal declarative effort presupposing that I was presenting lisp objects.

Rich interfaces pervade lisp's history.
https://screwlisp.small-web.org/clim/basic-principle/

Spec-Driven Development: The Waterfall Strikes Back

https://marmelab.com/blog/2025/11/12/spec-driven-development-waterfall-strikes-back.html

#HackerNews #SpecDrivenDevelopment #Waterfall #SoftwareEngineering #TechTrends #Agile

#programming #softwareEngineering #programmingLanguage #lisp #commonLisp #tiobe
https://screwlisp.small-web.org/fundamental/is-common-lisp-an-especially-good-language/
I graphed the number of years since each of the tiobe programming language top 25 current versions were created and found only ANSI Common Lisp (and Visual Basic 1998) aren't babies. I look at @nytpu 's article about lisp as a fulcrum for understanding what a current programming language having a long history is like. There are no examples outside of LISP.

Looking forward to your thoughts.

You Should Write An Agent

https://fly.io/blog/everyone-write-an-agent/

#HackerNews #You #Should #Write #An #Agent #AgentDevelopment #Programming #SoftwareEngineering #TechBlog #HackerNews

The Impossible Optimization, and the Metaprogramming to Achieve It

https://verdagon.dev/blog/impossible-optimization

#HackerNews #ImpossibleOptimization #Metaprogramming #DevBlog #SoftwareEngineering #HackerNews