Earlier quoted context omitted.
It is true, it's the official pricing of GitHub Copilot.
Why is GitHub sticking to per-request pricing when other providers switched to per-token for the high performing models?
Maybe MS wants people to use Co Pilot.
I run multiple $10K MRR companies on a $20/month tech stack
271–280 of 539 posts
Re: I run multiple $10K MRR companies on a $20/month tech stack
#272This is supposed to be a contrarian opinion yet this is a retoric yapped non-stop in the “build in public” community. Of course lean is a good approach, it makes sense, and most engineers know this. Is not a new concept, we’ve been doing this for years in every branch of engineering. The invented “people start with a k8s cluster for 5 users” doesn’t really exist. This is just a story repeated ad nauseam to fit a narr…
> The invented “people start with a k8s cluster for 5 users” doesn’t really exist. Can confirm it exists, especially with founders self-coding with LLMs now.
Experienced dev with limited hands-on big tech infrastructure experience. Based on the results I get from LLMs in domains I understand how get they even get this stuff running using AI?
Re: I run multiple $10K MRR companies on a $20/month tech stack
#273There are zero reasons to limit yourself to 1GB of RAM. By paying $20 instead of $5 you can get at least 8gb of RAM. You can use it for caches or a database that supports concurrent writes. The $15 difference won’t make any financial difference if you are trying to run a small business. Thinking about on how to fit everything on a $5 VPS does not help your business.
> There are zero reasons to limit yourself to 1GB of RAM. By paying $20 instead of $5 you can get at least 8gb of RAM.
In my head, I call this the 'doubling algorithm'.
If there's anything that's both relatively cheap and useful, but where "more" (either in quality or quantity) has additional utility, 2x it.
Then 2x it again.
Repeat until either: the price change becomes noticeable or utility stops being gained.
Tl;dr -- saving order-of single dollars is rarely worth the tradeoffs.
Re: I run multiple $10K MRR companies on a $20/month tech stack
#274Earlier quoted context omitted.
Sqlite smokes postgres on the same machine even with domain sockets [1]. This is before you get into using multiple sqlite database. What features postgres offers over sqlite in the context of running on a single machine with a monolithic app? Application functions [2] means you can extend it however you need with the same language you use to build your application. It also has a much better backup and replication st…
FYI, the color gradient on your website is an easy tell that it was vibe coded: https://prg.sh/ramblings/Why-Your-AI-Keeps-Building-the-Same...
A blog that's 11 years old and uses a minimalist CSS framework https://picocss.com ?
It's a static blog that renders markdown... there's literally nothing to code, let alone vibe code.
Re: I run multiple $10K MRR companies on a $20/month tech stack
#275I concur with some of the commenters that this read as a bit of a brain dump. It has a thread connecting several loosely-related topics.
Observation #1: You can also solve the tech stack problem with Heroku. I think the author's stack probably has a steeper learning curve, but is a cheaper option. I think it's a bit of an odd comparison (I won't say straw-man, as I don't doubt some people do this) to go from a fully-controlled simple setup to using AWS with a pile of extra crap. You can also, for example, run something similar to what he or she is describing on AWS, Heroku etc. (I.e. without the things in the AWS diagram he indicated like kubernetes and load balancers.)
Observation #2: I have not found WAL mode is an antidote to SQLite locks during multiple concurrent writes. (This is anecdotal)
I think regarding Go vs Python/Ruby etc. I completely get that. I would now like to check out Go on web. I use Rust for most of my software writing, but am still on Python for web servers, because there is nothing I can use for Rust that is as powerful and easy as Django.
Re: I run multiple $10K MRR companies on a $20/month tech stack
#276Similar approach here. I run a side project on Next.js + Vercel (free tier) + Neon Postgres (free tier). Total hosting cost: $0/month.
The one place I'd push back on SQLite: if your app has any write concurrency from external processes (cron jobs, webhooks), WAL mode helps but you still hit lock contention. I have data collection scripts running every 30 minutes that write to the same DB the web app reads from. Postgres handled that cleanly from day one. Neon's free tier is 512MB with connection pooling — more than enough for a side project with real data.
Re: I run multiple $10K MRR companies on a $20/month tech stack
#277Earlier quoted context omitted.
> SQLite on the same machine is akin to calling fwrite. Actually 35% faster than fwrite [1]. > This is also a system constraint as it forces a one-database-per-instance design You can scale incredibly far on a single node and have much better up time than github or anthropic. At this rate maybe even AWS/cloudflare. > you need to serve traffic beyond your local region Postgres still has a single node that can write. S…
How do you manage HA?
No offense, you wait. Like everyone's been doing for years in the internet and still do
- When AWS/GCP goes down, how do most handle HA?
- When a database server goes down, how do most handle HA?
- When Cloudflare goes down, how do most handle HA?
The down time here is the server crashed, routing failed or some other issue with the host. You wait.
One may run pingdom or something to alert you.
Re: I run multiple $10K MRR companies on a $20/month tech stack
#278Earlier quoted context omitted.
That's not really "lying" — ARR is usually understood as your projected "Annual Run Rate". It's a useful metric, as long as it is understood that it is an estimate. But, in all honesty, all RR numbers are estimates. MRR is also a "made up number" from a certain point of view: it is not equivalent to cash received every month, because of annual subscriptions, cancelations, etc.
>But, in all honesty, all RR numbers are estimates. Sure, but I would expect you to have at least one data point or at least near it, before making any estimates for that timescale. I don't see many people make MRR projections based on 2 days of of sales, it's just something I've noticed with startups and ARR.
2 days is optimum, you can fit a nice curve - 1, 2 ... at the current rate we will have 536,870,912 by day 30.
Re: I run multiple $10K MRR companies on a $20/month tech stack
#279Earlier quoted context omitted.
Because it doesn't make a difference, because `SELECT 1` doesn't need to touch the database: Running 100,000 `SELECT 1` queries: PostgreSQL (localhost): 2.71 seconds SQLite (in-memory): 0.07 seconds SQLite (tempfile): 0.07 seconds ( https://gist.github.com/leifkb/d8778422d450d9a3f103ed43258cc... )
Why are you doing meaningless microbenchmarks?
Are you claiming that this does not show the speed difference between socket vs in process communication?
Re: I run multiple $10K MRR companies on a $20/month tech stack
#280Earlier quoted context omitted.
Dynamic scaling is not really even available on a single node kubernetes. I was thinking more of Running multiple websites. i.e. 1 application per namespace. Tooling i.e. k9s for looking at logs etc. Upgrading applications etc.
Namespaces exist in Linux [0], they weren’t invented by K8s. You can view application logs with anything that can read a text file, or journalctl if your distro is using that. There are many methods of performing application upgrades with minimal downtime. 0: https://www.man7.org/linux/man-pages/man7/namespaces.7.html
Thats true. The reason I like k8s is once you've gone up the learning curve you can apply that knowledge to cloud deployments, on prem, or in this case VPS.
The authors stack left me thinking about how will he re-start the app if it crashes, versioning, containers, infra as code.
I've seen these articles before... the Ruby on Rails guys had the same idea and built https://kamal-deploy.org/
Which starts to look more and more like K3s as time goes on.