Hacker News Reader

Comment by msie | Hacker News Reader

It's been almost six years since that entry has been published. I don't perceive any big problems with robots.txt and favicon.ico files. Am I mistaken? Bandwidth is better and cheaper now so that's not an issue. Are other solutions just much more complex to implement? I hate over-engineering.

View on HN

It's not exactly the biggest crime on the modern web, but you have to admit that a hardcoded, root-level URI is pretty inelegant.

Pretty inelegant. Exaclty. Those two words sum it up. Not sure what the point of the whole rest of the article was.

One case where they don't work well is when you have multiple users on one domain. Like an old school server with the users at /~user1 and /~user2, etc. user2 might write a cgi program and want to create an entry in robots.txt but he doesn't have access to the global robots.txt, just the stuff in his home directory. Ditto for the favicon.

Nowadays domain names are cheap and I don't see as many sites like that any more...

Yeah, in the past six years I've come to accept that it's the worst possible solution, except for all the others. At this point I would be happy if we could isolate all the new "well-known locations" to a specific sub-path, such as /.well-known/

  http://bitworking.org/news/431/wave-first-thoughts

Reply on news.ycombinator.com