My professional life depends on PHP being alive (small hosting company). So I just have to comment and I am totally biased.
1st, of course good title for HN. But the article is indeed nuanced.
Sometimes lately, people announce to me, that they have built a SaaS App (vibe coded). When inquiring about the stack, they may not even now. It ends up to be what the AI thought will "get the job done" (Thanks for linking the response by Beberlei!). Indeed, AI seems to prefer Python and JS. That's concerning for PHP.
Yet. What I like about PHP is the simplicity. The backend can spit out HTML and it's server side rendered. When I compare that with all the setup that is required in JS to make that work - I think there is some space for this boring tech.
That is true and may have contributed why LLMS may largely have been developed in the civil sector. Reminds me of Lockheed Martin story from Acquired, where again, with interesting connections between silicon valley and LM.
Oh fancy tech, WASM, DuckDB! Looks awesome with the globe and indeed somewhat similar approach, although little overlap in data it seems. My research base was competitor data I already had floating idle in an internal document + some agent web search + manual checks on the data. I actually wanted to have some more GEO support, but getting actual data turned out to be too complicated (for initial release). Also looking glass was new to me I have to admit, fragments here: https://www.findhost.app/regions/
I had to look up what buyVDS stands for, I understand it's "Virtual Dedicated Server" - I am way more used to VPS.
> I had to look up what buyVDS stands for, I understand it's "Virtual Dedicated Server" - I am way more used to VPS.
Yeah VDS means that. Basically it was my birthday (18th to be exact) and someone from community gifted me that domain and I wanted to do something for this so its just a fun project that I have. I might have preferred buyvps.net domain but honestly I am happy that I implemented the idea.
Haha yeah, you can read through the history. I have tried to be detailed in the procedure for anyone interested to read more through it, hope this helps :)
> I wonder, don't you want the site to be indexed?
I must admit that its not as if I don't want my site to not be indexed but rather I don't know how to advertise it. I am thinking of adding more features to it perhaps in the future as well but most of my ideas are prototypes. I want to prove that I can build things (with AI currently but now in college, I would try to focus more so without AI)
What are some tips that you might give me if I do want my site to be indexed or more popular? I must admit that I know nothing about getting users on my websites! and so I'd like to learn this skill and I am more than happy to know these things.
On a tangent, I am also up for collaboration of data or anything similar if that might interest you. I believe in open collaboration and basically just like small hosting providers in a world of monopolized power like AWS etc., small hosting providers give me a glimmer of hope and I'd like to preach the message and do my part in spreading information about it, let me know what are your thoughts :)
AI can surely help you making the page more indexable. Not sure how much we want that, when it is accessible AI will just eat it. Adding to relevant forums can help, here, lowendtalk is good (I may promote FindHost there too). Currently limited in time, can not proceed with my project for a while.
I do support at a small hosting company: Even good developers often know little about hosting.
How to evaluate a hosting provider? Search for "Best PHP hosting" and you get drowned in affiliate pages. Smaller providers without marketing budget are invisible. It's so much shared hosting and so much boring VPS. There is interesting PaaS.
The project is based on internal research. This summer holidays I burned some tokens and learned some new tricks. The site is based on Astro and features Pagefind.
FindHost holds a few hundred providers. I try to focus on a maintainable list with a few providers in each section and region. There is also a heart on a couple of dozen providers I happen to know and like more. This may or may not give some signal.
I have of course not tested all hosting providers. The data is superfical, shallow in places. Categories are a bit blurry because the market is. Pricing is recorded as bands rather than figures, a price tracker rots in a week.
*Disclosure*: fortrabbit - my company - is featured too! It has its own record and a conflict-of-interest section on it. The project is therefore also a marketing stunt.
The data is CC BY 4.0 with JSON and CSV exports, and every page has a markdown twin. Contributions, corrections and additions welcome!
Some fun facts: Remember DotCloud which failed as a PaaS but shipped Docker on the way out, which is a neat parallel to Loudcloud (AWS before AWS) becoming Opsware.
1st, of course good title for HN. But the article is indeed nuanced.
Sometimes lately, people announce to me, that they have built a SaaS App (vibe coded). When inquiring about the stack, they may not even now. It ends up to be what the AI thought will "get the job done" (Thanks for linking the response by Beberlei!). Indeed, AI seems to prefer Python and JS. That's concerning for PHP.
Yet. What I like about PHP is the simplicity. The backend can spit out HTML and it's server side rendered. When I compare that with all the setup that is required in JS to make that work - I think there is some space for this boring tech.
reply