On 09/04/2008, at 4:10 PM, Maria Odea Ching wrote:
Is it persisted to disk so it is still correct after a restart?
Yep, it's being written into the disk with different filenames and is
updated everytime new artifacts are discovered in the repo. Now that
you've
mentioned this, I think I should also put a check or limit as to how
many
entries in the feeds should be kept so as to prevent the size of the
files
from bloating.
yep, I agree - and the size should be kept small so that it's quick to
download. You also need to make sure that the when the client requests
the RSS, it gets the right headers to not try and update it if it
hasn't changed for performance.
Hmm, by 'request' do you mean when a user clicks the rss feed icon to
subscribe to a feed? If it is, my only concern here is for the new
feed
entries. Since a user only subscribes to a feed once, if we have new
feed
entries (e.g. new artifacts discovered after another repo scan) then
the rss
would only be generated or updated if another user subscribes to the
feed
:-)
I don't think anything happens when they subscribe to the feed - the
same request is made every time the feed reader checks for an update.
The URLs given inside the feed do need to remain consistent, so
passing them through the request for the RSS makes the most sense.
How is the RSS accessed at the moment? I haven't checked the code -is
there a ww action for it, or is it served as the flat file?
- Brett
--
Brett Porter
[EMAIL PROTECTED]
http://blogs.exist.com/bporter/
