3 ms·
Cool URI's don't change. Slightly less cool URI's leave 301 redirects behind when they move. At the web scale you can't expect every org publishing resources
by olefoo 17y ago
Cool URI's don't change.
Slightly less cool URI's leave 301 redirects behind when they move.
At the web scale you can't expect every org publishing resources to keep them up indefinitely, you need to be able to deal with breakage.
- gnosis 17y agoThere are several different issues here. One is persistent access, which is what you seem to be referring to. Some others are metadata, along with persistent naming and location. It's important to be able to name something uniquely and get some metadata about it, even if that thing can't be located and/or accessed. Some of the schemes discussed in the article are designed to address these issues.
- olefoo 17y agoYes. And there are reasons for decoupling certain classes of identifiers from the DNS. But, for many of the entities we deal with on a daily basis that is a useless bit of metacrap. Dealing with metadata takes effort, for things that exist outside of the web that effort is worthwhile. For things that exist purely on the web, the URI is all the unique ID that is required, and if you need to move it to a different location in the namespace you can do so. A web page or blog post is uniquely identified at a point in time by being retrievable from a URL. If at some later date that URL becomes nonfunctional or changes meaning that resource no longer exists. You can still refer to it by it's URI even though it is not retrievable using the URI as a URL. I agree with you, it's just that unnecessary indirection offends my aesthetic sensibilities.