On 10/24/05, Eric Scheid <[EMAIL PROTECTED]> wrote: > > On 25/10/05 1:07 PM, "Luke Arno" <[EMAIL PROTECTED]> wrote: > > > Your editor performs a GET on the entry before you > > edit it, thus retrieving your spelling corrections. > > ah yes, that works ... unless your editor doesn't bother with the GET > because it sees the atom:updated in it's local version matches the > atom:updated in the query it just performed, and so uses the local cache. > The problem with that is that it's not multi-user or multi-station friendly. > > Is GETting the entry before editing a MUST or SHOULD requirement? If not, a > Pace needs to be written. >
I don't remember the exact language but I believe that there was a general consensus about GETing to edit. Does anyone remember what that thread was called? Etags would solve your problem (the lost update problem). If your requirements are less multi-user you could probably get away with GET to edit and GET again for conflict resolution right before PUTing if it has been a while since the initial GET. - Luke
