[Hypercontent-users] HyperContent 2.0 beta available now

[Alex V. <av...@co...>]

Hello all,
	I'm happy to announce that the HyperContent  2.0 beta release is now 
available from

https://sourceforge.net/project/showfiles.php?group_id=101745

This distribution has been successfully tested under Windows and Mac OS 
X; the code has also been tested on Tomcat 4.1.x under Solaris with an 
Apache front end and NFS mounted filesystems (our production setup).  
JDK 1.4.2 and Ant are required; JDK 1.5 has not been tested.  An 
embedded version of Tomcat 5 is included, but not required. The 
included READ ME has installation instructions, as well as instructions 
on how to migrate an existing 1.x project into 2.0 (about a 10 minute 
process, not including manual re-entry of groups and permissions!).   
Firefox is the recommended browser for using this release; IE will work 
fine, but there area a couple cosmetic glitches I didn't catch in time.

I must apologize in advance to those of you eager to get started with 
2.0; I happen to be leaving for a one week vacation to remote and 
tropical areas (i.e. no internet access) early Wednesday morning, so 
you have a little over 24 hours to ask me questions before I disappear. 
  Hopefully you will be able to spend that week exploring the system, 
and of course I'd encourage you to help each other out on this list.

Most of the known issues with this release are cosmetic and/or related 
to localization support (or the lack thereof); the underlying web 
serving, workflow and batch processing engines are quite stable and 
have been optimized for performance.  The rendering pipelines and 
build/publish are 100% backwards compatible with existing 1.4 project 
configurations and data.  The editors are exactly the same in both 
functionality and appearance to their 1.4 counterparts, with the 
exception of DTD parsing and prediction improvements for the XML 
editor, which make it possible to work with DocBook DTDs, among others. 
The only 1.4 features that aren't yet present are copy/move files, 
rebuild search index, and the Build Options editor.  The latter will 
not be brought forward, barring public outcry - it will eventually be 
superseded by a capacity for inline XML includes (post 2.0).  Copy/move 
files and rebuild search index will be added in time for the final 
release, and hopefully a scheduling component for the workflow engine 
will be added as well.  Target date for the final release is next 
month.

Q. If so much of version 2.0 is identical to 1.4, what's so great about 
it?
A. Lots!
	- standalone operation as a java web application (not just a portal
	  channel any more)
	- built-in web server and workflow engine
	- integration with external authentication, LDAP, web and file
	  servers
	- redesigned GUI allows users to navigate the output web site
	  to get around inside the system
	- support for multiple open files
	- persistent user history tracks recently viewed and saved files
	- management screens are themselves subject to content
	  management, and can be overridden or extended by any project
	- batch processing that can split large asynchronous jobs across
	  a pool of servers
	- support for binary data processing pipelines (e.g. image
	  thumbnails)
	- support for ** directory patterns in xml & resource definitions
	  (e.g. /**/*.jpg = allow jpg files anywhere)

And now a quick word on workflow, which is the major feature everyone 
has been waiting for.  The workflow engine is script-driven, and every 
project can have its own set of custom workflow scripts; by default, 
new projects get copies of the scripts installed in the bootstrap 
project.  The bundled scripts right now allow

	- non-approving authors to request approval of changes
	- approving users to request changes from a member of the authors group
	- upload and download zips
	- build & publish (individual files or whole site)
	- watch files (receive notification of changes)

Any work generated in a workflow is kept around as long as it is 
"enqueued" somewhere, which means you can keep a history of your 
builds, publications, uploaded and downloaded zips, etc. Right now you 
can manually delete the work records any time after they are complete, 
and when the scheduling component is ready you will be able to 
configure self-deletion after a specified amount of time.

Email notifications require you to configure a valid SMTP server in the 
hypercontent.properties file, and you must add valid e-mail addresses 
for users in their VCard files  (these are linked from the user's name 
when logged in).  When integrated with your campus authentication and 
LDAP, HyperContent can automatically construct a VCard for each user on 
first login, including email address, that will be used to support 
workflow.  When I return from vacation, I'll be happy to give people 
some guidance on this integration; for the adventurous, JAAS is used 
for pluggable authentication, and the LDAP to VCard configuration file 
is the same as in version 1.4, so go for it.

Good luck to everyone, I look forward to hearing your feedback.

Cheers,
	Alex