Hi,
Just out of my own experience...
When you install a magnolia, the first time the public and author
instance start up, there is a process called bootstraping, which
basically initialises those two instances with default values that
are almost identical. Bootstrap content can be found in WEBINF/
bootstrap in the form of XML files, but is also sometimes embedded in
JARs (like the default site, which comes bundled in the magnolia-
samples.jar). (Note that sourceforge's WARs come with this process
already done, as far as I remember).
These default values provide a fully working configuration on both
side, with identical content and identical model/view/controller
configuration.
There are still a few differences, which are the normal differences
between the authoring and public environment like :
- Password protection of the authoring environment
- Cache activation on the public side
- Public instance subscription to the authoring instance
- Etc.
But apart from that, intances are in the same state, post bootstrap.
That means that you do not have to activate all nodes from the
authoring instance in order for the public side to work, the
synchronisation work is done by bootstraping the same basic datas on
first start. It's just as if everything was activated for you.
Howerver, when you start modifying content or paragraphs or user
settings (...), then you have to activate in order to send
modifications on the public side.
But what is important here, is to inderstand that the color code does
not reflect the prensence/absence of a node on the public side. It is
not a throughout comparison between the public and the author
instances to see if there is a match between the public and the
author version of the node.
On the contrary, color informations only come from the datas of the
authoring side. Nodes on the authoring site contain metadata that
store dates of modification and publication, and this is where the
"colors" are computed from, not from a dialog between the instances,
and that's because there are several ways to import content into a
Magnolia instance (bootstrap, hand editing, activation, import/export
servlets...) that the author instance can not always know about.
Example...
In a scenario where a user would publish a modification to a web page
through the normal activation process on the authoring environment
(therefore, insuring that the light is "green" for the said node),
and then log in the public instance, make some further modification
directly there, in such a case, the authoring instance wouldn't ever
know about this, and still show up a green light.
In a nutshell : colors reflect what the author side thinks the state
is. Red is never published or heavily modified (in a way that
requires a de activation). Orange is modified since last activation.
Green is not modified since activation.
Dots are red at first because nothing was ever activated. But things
work thanks to the bootstraping process that provided identical datas
to both instances, which is basically what activation does. In any
other case, things don't work if they are not activated.
You can go behind the back of the author instance and modify directly
things on the public side, but the author side will never know about
it and colors become meaningless.
All things considered, on a stable config and during a normal
workflow, colors can be trusted, as significantly different nodes
from the original/default configuration (config nodes as well as
content nodes) have to be activated at one point anyway, therefore
showing up in green, and if a once-green light becomes orange or red,
this IS a sure sign that something happened.
I wrote this a little too fast, I hope I did not make any serious
mistake.
Guillaume.
Le 17 avr. 07 à 09:09, Narinder Kumar a écrit :
I am running Magnolia 3.0.2 on Tomcat 5.5. It shows red
corresponding to Subscriber and even then I am able to activate new
contents from Author to Public instances.
Just try editing one of the sample pages provided by Magnolia
without doing anything at Subscriber node. If you get an activation
request in Publisher Inbox and corresponding action is executed
successfully then you don't need to do anything at Subscriber node
as well.
Perhaps Magnolia developers may throw more light on the behaviour
of Red and Green buttons after a new install against some of core
nodes.
Regards
Narinder
----- Original Message ----- From: "Nicholas Orr" <user-
[EMAIL PROTECTED]>
To: <[email protected]>
Sent: Tuesday, April 17, 2007 12:16 PM
Subject: Re: [magnolia-user] Activation of New Clean Install
After mucking around for a while I've come to the conclusion that it
is just the Subscriber node that needs to be activated. Is that
correct?
----------------------------------------------------------------
for list details see
http://www.magnolia.info/en/developer.html
----------------------------------------------------------------
----------------------------------------------------------------
for list details see
http://www.magnolia.info/en/developer.html
----------------------------------------------------------------
----------------------------------------------------------------
for list details see
http://www.magnolia.info/en/developer.html
----------------------------------------------------------------
