neebz

In pure software architecture and design term, this would have been convincing until the recent models who are let’s be honest better than humans in architecture and design.

we have done something similar at puppydog.io but more targeted towards corporate marketing videos. our experience working with customers is the same 80-20 rule.

For 80% AI does the work in seconds and 20% is manual work in minutes to get exactly what they want.

This is like pre-LLM, you presenting a problem to the team, and they google and send you the top few links thinking they have found the solution and have done their job.

I actually banned this and made it part of onboarding.

I can google too. You need to read yourself, ground the solution to our setting and then propose a plan.

do you also have separate prompts for each domain (security, architecture etc?).

would love to look into it if any part of it is open source

Once the plan stage is done is it fire-and-forget for you afterwards?

what a brilliant blog. the Lotus 1-2-3 screen brings so many memories of my childhood.

My father was a power user of Lotus back in the late 80's. He extensively used it as his job at GE. When we moved back to Pakistan, he setup a girls school and tracked everything from students to accounting to results in Lotus. In many ways, Lotus showed him the power of computers and made him buy a home computer when hardly anyone I knew had it.

Late in his life the world moved onto Excel and reluctantly he had to do it too but his love for Lotus never went away.

I've faced the same but my conclusion is the opposite.

In the past 6 months, all my code has been written by claude code and gemini cli. I have written code backend, frontend, infrastructure and iOS. Considering my career trajectory all of this was impossible a couple of years ago.

But the technical debt has been enormous. And I'll be honest, my understanding of these technologies hasn't been 'expert' level. I'm 100% sure any experienced dev could go through my code and may think it's a load of crap requiring serious re-architecture.

It works (that's great!) but the 'software engineering' side of things is still subpar.

For window management, Aerospace has been a game changer for me. I could never work with multiple monitors before that.

We switched to solar in 2021 expecting a 3.5-year payback. Electricity prices rose so fast that we recovered the investment in under two years.

Also the national grid is notorious for it's frequent blackouts (load-shedding) since the early ’90s. Solar allowed us to have uninterrupted supply in the mornings and longer backups during night.

All these coding agent workflows really drive home how important a solid test suite is but who’s actually writing the tests? In my case Claude Code keeps missing key scenarios and I’ve had to point out every gap myself.

Also reviewing LLM generated code is way more mentally draining and that’s with just one agent. I can’t imagine trying to review code from multiple agents working in parallel.

Finally I’ve shipped a bunch of features to prod with Claude writing all the code. Productivity definitely went up, but my understanding of the codebase dropped fast. Three months in I’m actually struggling to write good prompts or give accurate estimates because my grasp of the code hasn’t really grown (code reviews only help so much). Curious if anyone else has run into the same thing?

Looks super interesting. Has anyone here tried a keyboard-first tiling window manager on macOS? Any recommendations?

Finally someone emphasized upon the cognitive load of AI-assisted coding.

It definitely makes me faster but it's consistent prompt->code-review->prompt->code-review->scratch->prompt-code->review cycle which just requires extreme focus.

We setup a solar on in our home in Pakistan back in 2023. Our primary driver was cost since it was getting extremely expensive during summers when the ACs were on. We expected to get the return in 4 years but inflation went so high in subsequent years that it took just 2 years.

I remember back in the summer of 1996 in Pakistan our household was one of the first few to have to internet.

At that time angelfire.com used to give free webspace. My brother got hold of a pirated version of CorelDraw and setup a fan website of his favorite rock band Junoon, which incidentally is still online: https://www.angelfire.com/pa/JUNOON

And then when my brother met the band at a concert and they actually recognized him due to the website. I guess first time we realized how impactful internet is going to be.

yup - some documentation/legal work remains but they should be able own the company abroad as a subsidiary soon.

how would you suggest we should pursue this once all the paperwork is done? hire an immigration lawyer? any recommendations? we are non-YC.

How are the odds for bringing your remote employees to US using the L1 visa? They've worked for us for more than 5 years in management role. Also the startup got acquired sometime ago, does that impact the L1 application?

Bertrand was one of my favorite people in the software field. I really wanted to do a Phd under his supervision.

awesome. thanks for sharing.

previous discussion: https://news.ycombinator.com/item?id=30674159

Some feedback from our transition:

- automated tests are the pre-req of feature flags. kindly make sure that you have good test coverage. otherwise the team loses confidence in code & feature flags which defeats the whole purpose.

- as part of code review guidelines we added one major feedback item: can this PR be merged right now? why not? what can be done to merge it right away? as part of a cultural shift the code reviewers played a critical role. they kept engineers on the their toes at all times.

- feature flags should be part of user stories. in many cases you can't expect engineers to add a feature flag at the end of implementation. the product team should know this at the time of writing user stories.

- as mentioned in many other comments try to clean up code after feature flags become stale. otherwise the code (and respective automated test cases) becomes a huge mess.

- in our case we decided the granularity of feature flags (per-user, per-customer, per-region, per-server etc). we started with per-customer and went from there. worked out fine for us.

- feature flags have major two benefits: "release often, release small" & "decouple releases with big launches". please make sure to instill this in your team every day. mindset changes takes repetition and emphasis. if at anytime you feel that you are not achieving any of these please take a step back and figure out why.

Yes. Feature flags changed the mindset of my whole team. I absolutely love the fact that it pushes the whole team towards release-small-release-often mentality. It's a technical solution with many benefits but I love the cultural impact the most.

In Safari you can set reader mode turned on for specified websites. I have done it for all the sites which I visit frequently. Unfortunately there is no way to to turn it on at global level.

The only minor gripe I have is the slide-above animation when reader mode turns on. It's jarring when you keep getting that on each navigation.

Personally I learnt to be less judgmental of people. From media you compartmentalize everyone into neat little boxes to make sense of the world. We find other people's values ridiculous when we hear/read about it but only when we go and live with them we find out it makes sense. It may be wrong but it makes sense. And there is a good chance if you were brought up in that same society you may have had the same behavior. I've learned to empathize only when I experienced this first hand.

regarding local environments,

have you faced performance issues? running 5-6 docker images (app,redis,db,mq etc) at the same time has been causing the machine to lag for us. most of my team is on the 16 inch macbook.

It's a bit strange but for me Google's biggest pull is live scores.

I follow cricket, tennis and football. And all the dedicated websites are heavy and slow (cricinfo, espn etc). Google scores are quick to load and update even before the rest.

For me Hipmunk had the best calendar control. You could just type in your dates in pretty much any format and it was able to parse it. No need mess around with pop-ups and small buttons to switch months etc.

How exactly do you use NewRelic to see what is blocking the event loop? I thought we always needed Flame Graphs for it (which NR doesn't provide)

My father in the 80s was working at GE where he extensively used Lotus 1-2-3. He loved it so much that he bought a computer in 85 and used to use it at home as well. I always hated the boring blue and black screen.

We moved back to Pakistan in early 90s and we setup a small girls school. He used Lotus to keep track of all students, results, late fees, accounts etc. It required a lot of data entry but he absolutely loved that he could get required data in seconds. All the fellow owners of schools were super fascinated.

It took quite a while for him to shift to Excel later as he was super comfortable with Lotus. Pretty sure before his death deep down he still preferred Lotus over Excel

If anyone on mixpanel is reading this I would love to know how do they maintain mocks? We've been burnt so many times with outdated mocks in my experience that I've always tried to supplement our front-end tests with Selenium as they run against a real server.

You can do what codecov does; They keep updating the same comment with the latest result.