Using the URL to build database-free web apps

bryanbraun.com

18 points·bryanbraun··8 comments

Yeah, this is cool. Combined with a url shortener that doesn't have length limits this idea could really have legs.

I made a semi-stateful web page like this but I stored the state in a single object that gets saved to local storage. Then I stringify that object and email it to myself whenever I switch between phone and computer. ...The web page can read in the string. https://smchughinfo.github.io/search.html

It works good enough for me. But the holy grail would be a way to store that state so that different devices could share state without having to do anything. Hmm.

Although URL shorthener essentially brings in the database the article was trying to avoid in the first place, I was thinking on similar terms. It allows identity-less scenarios very easily, e.g. for effortless product evaluation.

I love this idea, but it offloads the task of storage to the user. Of the user stores these URLs inefficiently (e.g. twitter), it would create a lot of digital garbage.

One big positive concerns sensitive info. The app wouldn't have to store it.

Godbolt.org Compiler Explorer works like this. Your code snippets are stored in the URL you share, so none of your code has to be persisted on the server.

Not if you press the share button which generates a short link that doesn’t have enough entropy to store the entire content.

Another example of this technique is https://sequencediagram.org

How is this different from the GET method in PHP?

I am not versed in PHP, but the article assumes HTTP usage, most probably HTTP GET method. I guess PHP GET you mention is a way to issue HTTP GET request.

Reply on news.ycombinator.com