Ramana

With zero coding experience, artist building 180 webpages in 180 days “Learning by doing” is taken to the extreme by novice programmer.

arstechnica.com/information-...

This article from @arstechnica.com made me believe I can be software engineer even if I am not pursuing a CS degree.

What a different time 2013 was!

Raycast is super nice.

Using AI for every little thing is stupid. You want to rename a variable, use variable renaming from LSP.

YouTube video by TerminalDotShop Adam Wolff from AnthropicAI - Tomorrow 139

This podcast is great listen. Remove the hype out of AI and keeps it real.

Some takeaways:

- AI hallucinating interruptions
- Programmer disengagement is a problem
- Anthropic instrumented Zed to see how they use editors.

@thdxr.com @adam.dev

www.youtube.com/watch?v=xY7p...

I wish I can see what's happening since the past 20 mins

I have 35GB of duckdb wal file. When I open duckdb cli, it's taking forever to open. Don't know how long it will take to open the file

Screenshot of dashboard in Metabase

Metabase is awesome. Really easy to setup dashboards!

Also extremely easy to deploy from Coolify!

Finally setup my Photo Library on a 2 TB SSD after procastrinating for several years. Now I need to organize 4 years of photos. At least I have decent workflow.

This page by @turbopuffer.bsky.social is awesome. It auto sorts the closest region to you.

turbopuffer.com/docs/regions

Did AWS just sherlock @turbopuffer.bsky.social ?

aws.amazon.com/s3/features/...

I love hackmd to write documents when I am on a video call. It's so smooth. It feels much faster compared to writing on Google Docs.

jvns.ca/til/fzf-prev...

This is so cool. Using fzf as UI layer

Thanks for this post @b0rk.jvns.ca

Sigh! Now the VM hangs at pnpm install. May be time to switch VMs.

I lived with it for more than a year not knowing what it was. Now that I know it, I just could not live with it.

I have to breakup with prisma in side project and switch to drizzle. prisma generate causes the VM to hang randomly. I don't have will power to investigate why CPU usage goes to 100% suddenly.

Yes that’s the other big news. Big win for postgres community either way

Me: Dear GCP, can you increase request quota for t2d instances from 0 to 1?

GCP: Please contact through your account team or submit a request to sales team.

Unless you are an enterprise, GCP doesn't want you.

I remember looking at it a while back. It didn’t support versioning back then.

Great to see @planetscale.com launch a Postgres offering. The best place to store any critical data.

Probably the nth time I got nerd sniped into thinking about running CI in spot VMs. It’s too much work for non-existential gain and also adds a bus factor in the context of my workplace. 😮‍💨

After some more poking around, I was able to use Cloudflare tunnels to replicate the same.

All requests have to manditorily go to the worker to respond even though the asset was cached on Cloudflare. I am also consuming my worker daily limit. It seemed like there should be a better.

Cloudflare CDN is the obvious option. First I have used workers to proxy the requests and cache them. It worked but not the way I expected.

I have side project which uses 200GB of bandwidth per month. I currently host this on AWS Lightsail. Images are probably 99% of this bandwidth used. I spent the day trying to optimize this at zero cost.

Declutter your JavaScript & TypeScript projects Project linter to find unused dependencies, exports and files

knip.dev I keep forgetting that this amazing too exists.

Setting up rules in iCloud mail box made a ton of difference.

If I have read the error messages in console, I would have figured it out much sooner. The fix was to simply update the nonce in CSP.

I spent 3 hours fighting with a bug that occured after we migrated to latest version of vite/plugin-react. Turns out it's a issue with our CSP not allowing react-refresh inline script to be injected into index.html.

Last night was quite eventful. First the GCP outage till 3AM then my kid started crying at 4AM. Thankfully my wife took over. I was semi conscious but I was completely exhausted.

Seems like GCP outage has recovered. Our app has returned to it's normal status. Alert Policies are recovering slowly.