[mdx] mdq issue #9: etag in conditional GET

[Cantor, Scott]

On 6/16/14, 4:31 PM, "Ian Young" <ian at iay.org.uk> wrote:
>
>For Scott: does the Shibboleth SP implementation make use of ETag if it's
>provided? Would it take advantage of anything else if ETag was not
>provided?

It does not. This was moderately surprising, but less so given that the
code I wrote for this was a long time ago, and probably predated my adding
caching support to the standard metadata file loading code. Also, I
probably didn't think much about it given that the amount of data was so
minor and it only attempts to load it when the cacheDuration attribute
indicates to (or it's starting from scratch).

>For everyone: can anyone think of any reason not to require clients to
>support ETag (which is the same as saying that the server only needs to
>support it, and not anything else, which probably simplifies it a little).

I would make it a SHOULD, since clients supporting caching is an
optimization but doesn't really break anything if they don't.

-- Scott
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 5712 bytes
Desc: not available
URL: <http://lists.iay.org.uk/pipermail/mdx-iay.org.uk/attachments/20140616/307cf7cd/attachment.bin>