Zernat

Thank you, I appreciate it!

Truth to be told, I don't have any hard evidence to back up the privacy part. But I like to think that at least as a bonus, users get much higher speed of feedback. They don't need to wait for upload/download so at least it's that much faster.

Hello everyone!

I am building a website with hundreds (currently 13) of client based tools (I started with images). My primary goal is a simple and clean UI/UX. For example I am kinda proud from the "Before/After" preview, I think it should make people lives easier. I hope to help folks that are not interested in technical details, that just want to quickly get the job done with as little options or noise as possible. The core is privacy and speed. Zero uploads or backend servers. Everything runs locally in the browser via WebAssembly.

A few technical "fun facts" about my stack: - Astro (Hybrid SSG/SSR for i18n) + React islands

- All heavy lifting happens in Web Workers. Rust compiled to WASM for standard operations. For exotic formats (HEIC, AVIF) and Display P3 to sRGB conversions, it lazy-loads magick-wasm (~5MB).

- I had a lot of "fun" with Color Profiles, I am managing them mostly "by hand" to ensure correct colors. I noticed that even some bigger websites have sometimes problems in this regard.

- Everything is deployed to Cloudflare Pages (after working mostly with Python/Django, I am in love with Cloduflare staging/prod deployments and quick e2e testing)

In the future I hope to add much more tools like PDFs, Audio, Video and maybe some "cooler" stuff like ML models for specific problems (also of course fully on client side).

I would love to hear your feedback, especially if you manage to break the WASM workers with some "weird" files. And of course if you would notice problems with ICC Prifles. Thank you!

Well, I have a hard time to interpret it as something different than sarcasm :D

Hello! I wanted to ask you for advice. Few weeks ago I had a "brilliant" idea that I will build a database of online tools, saas'es, etc. After some research (and questions on HN - Thank you! <3) I thought that maybe I could build mini (veryyyy mini) crunchbase.

After this 3 - 4 weeks I built this - https://www.ravenregistry.com/ (it's free, no ads, no paywall)

The problem is, actually useful information are not freely available. So I was able to get "regular" things like url, title, description, contact/pricing url, socials.

And I think that it's not enough for the "power users", so people who would like to use my website for a market research or competition research won't find it useful (?)

So I was wondering if maybe I should try to pay for services like SimilarWeb and try to enrich my data with their data?

Or maybe I should pivot and try to target regular users who are looking for online tools? But then what, my competitor is google?

I also thought that I could make it for regular users and at the same time downscale significantly and instead of going into huge number of results - start filling them manually with way more information?

I would be very grateful for any feedback and ideas! Thank you!

That's amazing, thank you very much for sharing!

I had no idea about crunchbase. Hmm, so I guess the idea here is that I would be building crunchbase for poor people. It looks like they are charging 99$ per month or 588$ per year, for 4M entries (with far more info that I would have, and far more features). So the question is, could I steal some of their customers by offering much smaller amount of data with much smaller price?

I am wondering, if I would have 50 000 entries looking similar to what I already have in my demo (but with better search capabilities), would it be worth, let's say 5$ per month (x searches per day or something) and/or maybe 50$/100$ for a whole database dump in a .csv format?

My demo looks like that currently - https://lazarustest.eu.pythonanywhere.com/

Hi! At first I wanted to build something similar for regular folks that would be looking for online tools. But then I thought that while it could work, it would require a lot of additional work (researching how price tiers works for each website, does tool have free tier available etc.) and then presenting it in a nice, fancy way (screenshots, logos etc.).

So looking at the amount of "frontend" work I thought that I could pivot into targeting people that are not "regular" users. But rather: - people looking for ideas for their business - existing business researching competition or businesses to partner up with - some general way to track trends, see what was done and in what way etc. - maybe marketing people?

That is my current idea. Do you think it makes sense?

I agree, after some discussion I decided to click "hide" on this post and post another with "Shown HN", sorry about wrong flagging.

Done! Thank you for spending the time explaining this to me. I clicked "hide" on this post and wrote one with "Shown HN" - https://news.ycombinator.com/item?id=43239066

Thanks!

Ohhhh, I had no idea about this! To be honest this is my first post on Hacker News ever.

Do you think that I should "hide" this post and write it again with this "Show HN"?

Anyway, thank you for sharing!

Hi! I wanted to share my little "adventure" that I had in the past ~2 weeks. Basically, I got tired of constant thinking and I decided to build something, anything.

By accident I learned about thing called WebAssembly (basically you can execute pretty complicated programs locally in the browser without sending anything anywhere) so I immediately assumed that it could be used for a set of small, private tools.

Usually I am building with Python, Django, PostgreSQL, hosting it on PythonAnywhere. So this time I decided that I will step out of my comfort zone even further! So I went with: - Rust (for tools that will be compiled to WebAssembly) - Zola (static site generator since I don't need anything "dynamic") - AWS S3 + AWS CloudFront (for hosting this static website) (I even used AWS to generate my SSL certificate)

Let me know what you think, and of course if I can already sell my website for million dollars :D

Cheers!