Re: resource immutability

On Tue, Oct 22, 2002 at 02:37:30PM -0700, Mark Nottingham wrote:
> Just thinking out loud;
> There are a number of situations where it would be good to explicitly mark
> a representation of a resource as immutable; that is, state that it will
> not vary over time. For example, it would be nice to have an explicit way
> to let schema processors know that they should download a representation
> into their cache and use it from now on. Or, that a manually downloaded
> and installed schema is effectively good forever.
> Although this can (and ultimately should) be done in a framework like RDF,
> HTTP seems like a more reasonable place to start; there are efforts
> underway to bridge HTTP metadata to RDF anyway.
> I'm thinking something like a Cache-Control directive;
>   Cache-Control: time-immutable
> It can be argued that the caching model in HTTP/1.1 allows UAs to do this
> already (i.e., either give the representation an extremely long freshness
> lifetime, or use stale representations with a warning)

We do this for W3C's TR space: current specs have a lifetime of 6 hours,
but archived documents (/TR/yyyy/*) have a lifetime of 52 weeks.

> but this subtlety seems to pass most implementors by.
> An explicit directive is easy to understand and apply.
> Thoughts?

It might just be a matter of educating people what is already possible.

The other day the RSS world discovered ETags and it seems to have
spread around various implementations quite quickly, after some
people wrote it up clearly and others became motivated enough by
the bandwidth savings to spend time implementing it.

Gerald Oskoboiny
World Wide Web Consortium (W3C)

Received on Saturday, 2 November 2002 13:45:53 UTC