I confirm that it works. I watched the clock for a while and I didn't understand how to read the time, or the day, or anything. I think that I is google the original clock and look for instructions.
Why should it be behind AWS or CloudFlare? Can't anybody maintain their own servers in 2024, especially when everything works and doesn't crash after a HN spike?
Not to pick on you in particular, but I hate this laziness trend from sysadmins that are the cause of the whole centralisation of the Internet. If you have the knowledge to build a SAN and entire architecture yourself, teach that to the youngbloods, instead of just telling them to get AWS credits.
Make the Internet decentralised again.
---
Don't give a junior dev a cloud server, but teach them to administer a UNIX machine and they won't need free credits from anyone.
Not the OP, but the thing that springs to mind is to make it resilient to DOS attack and similar. As you say, it didn't crash after the HN spike, but it seems likely that a motivated attacker would not have trouble bringing it down. Also the request latency is quite bad, which doesn't bother me a bit for the type of site that it is, but that's solved easily enough since it should be super cacheable.
Interesting...my thoughts are that moving to the cloud is not lazyness. Cloudflare is for DDOS protection, AWS is so I can focus on the servers not the hardware.
I really do like your point on teaching young bloods I had assumed much like the knowledge about steam trains it was old knowledge that is becoming not needed.
