Re: naming resources - Slug-Header

Erik, 

I understand your arguments for using Slugs ... I think I would be happy to go with them too. 

Some comments inline, all the same ...

>> we are making assumption about the content though, arn't we?
>> It's RDF .. or a picture, mp3, etc.
> 
> at the slug level, all you're saying is that "i am creating a new
> resource, please use the following string as input to generating a path
> component in the newly created URI." that is entirely independent of what
> content type you're shipping around.
> it's on the web service level of
> conversation.
> 
> http://tools.ietf.org/html/rfc5023#section-9.7
> 
>> The description in the RDF makes the system work, and therefore is part
>> of the protocol in some way ... (?)
> 
> yes, RDF is part of the protocol because for example to find any
> hyperlinks that allow you to interact with the server, clients need to
> understand RDF and LDP. but anything that is already covered by web
> architecture (by a spec) already shouldn't be reinvented.

there are lots of "parameters" which be used to influence creation. A hint on the name, preferred range of colours, requested toppings, etc. I don't see much difference between them. After-all, a human friendly naming of the resource, doesn't change the protocol ?

So, I understand your opinion about reinvention, but, I also quite like having all inputs in the same place. 

That said, I appreciate Henry's argument that slugs are necessary with non-RDF content.

Roger

> 
> cheers,
> 
> dret.
>