I also kept release dates handy so it was a simple matter to list any updates for a given collection past a certain timestamp - just a normal metadata query.
IIRC, Vignette uses the ICE protocol for its syndication server. This service lets you set up scheduled replication processes at the collection level either triggered by new content or on a periodic basis.
IIRC, Interwoven has nice versioning and merging capabilities, which can be good if you have multiple update sites and you need to merge more often. But I have found most programmers don't really get branching in source control (I know, because I have trained many on CVS and VSS branching). I think it is even harder for most authors to cope w/ too much complexity in versioning (Again, I have had to explain all this stuff before).
There are so many CMSes out there, with so many ways of storing the content that, naturally, replication schemes vary as well. In some systems, this might be configured as an export/import "feed". In the end, you need to get "down to cases" and figure it out for your particular system.
hth,
Charles
At 10:23 PM 11/24/2002 +0000, Drew McLellan wrote:
Charles Reitzel wrote:The short answer is to manage replication by collection. Collection is a good unit for this type of management because of its open ended definition. A collection is just a bunch of stuff that goes together.
So people tend to naturally understand and even expect that a collection as a whole will have the same state.
<snip>The specific mechanisms for tracking the status and actually performing the replication are, of course, product dependent.Thanks Charles, that's very useful.
What mechanisms for declaring a given modification or addition as part of a specific collection have you come across?
We some better than others?
drew
--
http://cms-list.org/
trim your replies for good karma.
-- http://cms-list.org/ trim your replies for good karma.
