Andrei Gordienkov

Solution Architect @ ABAX, IT Trainer

[Andrei Gordienkov @violettape]

@BartomiejKaa Finally it works

[Kubek K @kubek_k]

Hej @Citi i #VeloBank

Czy was do reszty pogoęło?
Połączenie banków miało być bezproblemowe, w weekend.

Mamy poniedziałek rano i nie działa nic.
Ani starym użytkownikom Citi.
Ani starym użytkownikom Velo

[Mitchell Hashimoto @mitchellh]

I strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out.

I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really).

It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely.

The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture.

We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying.

I worry.

[Andrei Gordienkov @violettape]

@mattpocockuk except it Business Central xD

[Simon Martinelli @simas_ch]

Software Engineering history should be mandatory in computer science courses. I see people trying to solve the distributed fallacies with the same problems like EJB, DCOM, and CORBA 😱

[Andrei Gordienkov @violettape]

I just published How spec-lint lets a team self-check structural traceability https://t.co/cB3AicsId5

[dex @dexhorthy]

@mattpocockuk this is great and I think this is the right framing but something I struggle with - how do you deal with “StOp GaTeKeEpInG rEaL eNgInEeRiNg” that I can only imagine is piling in the comments because those people have a point but also we need adults or it’s all gonna fall apart

[Andrei Gordienkov @violettape]

Discussed with AI a problem and suddenly it provieded: " Architecture is the brake pedal that allows the AI car to drive 150 mph. Without it, you're just a highly productive crash dummy." - very accurate, imho.

[Kirill Skrygan @kskrygan]

Would you be interested if JetBrains releases a totally local AI agent, working 100% on your laptop, using our code insight engine and deeply integrated into the IDE?

Yes, it will be probably 1 month behind the very recent frontier models, but no token blood bath anymore

WDYT?

[Andrei Gordienkov @violettape]

@dexhorthy that's so much resonates with what I think about AI. Unfortunately, there some frictions in my company to understand where AI is really bad and why engineers skill and architecture is so badly required now.

[vicki @vboykis]

What I most desire doesn’t exist on the market yet: a coding agent model, but it’s only been trained on code reviews by Eastern Europeans.

[Andrei Gordienkov @violettape]

I just published Spec-driven AI development for teams. Attempt to connect What with How. https://t.co/uPYZkaxC4J

[Andrei Gordienkov @violettape]

@softwarevlogger а модификации нужны потому, что все препинаки не зависят от языка, спасибо статье Optozorax

[Andrei Gordienkov @violettape]

@softwarevlogger я использую Capewell 0.9.2 с некоторыми мрдификациями и Диктор. Трудно было что пиздец, но рекомендую. Теперь на qwerty ощущение что пальцы узлом завязываются, причём между собой на двух руках.

[Valentin Ignatev @valigo]

Let's play a game - win32 types vs Polish language:

LPCWSTR
PSZCZYNA
WCSLEN
WCZESNY
LPCTSTR
BYDGOSZCZ
WSTRZAS
HGDIOBJ
DOWOD
HWINSTA
DLUGOSC
LPCSTR
DWORD
KAL
LPWSTR
SZCZECIN
PCWSTR
BLAD
PUHALF
CHUJ
UHALF

[Vlad Khononov @vladikk]

Last week I announced the Modularity plugin for Claude Code. Got so much positive feedback! Thank you all 🎉

From all the messages I received, one stood out: a software engineer who had been trying to surface technical debt, but wasn't taken seriously. Running the /modularity:review skill surfaced the same issues, and the reasoning and risks in the review document helped make the point.

Seeing a machine use the Balanced Coupling model to identify significant architectural issues and design proper modular solutions makes me happy, proud, and somewhat uneasy. I'll leave it at that.

Next version is in the works, hoping to finish this weekend. Currently, the design skill primarily targets greenfield solutions. In the next version it will better support evolving brownfield codebases, and produce implementation plans that coding agents can execute.

If you missed the plugin, give it a try: https://t.co/hjTHncmsPJ

[Vlad Khononov @vladikk]

🌟🌟🌟Modularity Skills for Claude Code🌟🌟🌟

Last week I quietly released my Balanced Coupling-based Claude Code plugin. Today it hit 100 GitHub stars, so time to make it official 🎉

The plugin tackles architectural and design challenges that existing tools don't and comes with two skills:

/modularity:review analyzes your codebase for architectural / design issues and produces actionable recommendations for improving its modularity.

/modularity:design processes functional requirements and designs a modular architecture, complete with architecture documentation, module design docs, integration contracts, and test scenarios.

Based on my own testing and feedback from others, it produces relevant and actionable results. The part I'm still working on: its subdomain identification abilities. But hey, that's the hardest part of DDD anyway. To its defense, it compensates by asking tons of questions to understand your business context.

If you installed the plugin last week, make sure to update, as there were some usability improvements. Thanks, @alexeiled!

🔗Link to the repository with installation and usage docs is in the first comment

P.S. Sharing is caring 😉

[Andrei Gordienkov @violettape]

@vladikk may I contact you in tg to discuss results?

[Andrei Gordienkov @violettape]

@vladikk tested /review for greenfield and ongoing project and I have mixed feelings. But more good one compare to confusing. Slightly touched "hl-design" and that's the point of more confusion or very opinionated results (known arch sickness).