David Crossley wrote:
Diana Shannon wrote: <snip/>
What I discovered in the process was item 8 (which I had accidentally merged with item 7 on a previous edit):
8. The documents that are published by Forrest need to be subsequently checked-in to the xml-site CVS.
Do you mean the live site cvs? Do we **really** have to do this? Goodness, I thought Forrest automation would eliminate this step.
Why is it still needed?
Good point. However, this was an "issue" gleaned from some email [1], that cvs history had to be maintained for xml-site/targets/*.html
Another reason was that the current publishing mechanism. is an automated cron job that just copies the content of xml-site/targets/cocoon/ to xml.apache.org
[1] See Ken email linked from wiki to: http://marc.theaimsgroup.com/?l=forrest-dev&m=103649129528547
It seems that automatic "publishing" to CVS can be done by redoing an import to CVS at each site update, and CVS should be able to understand what has changed and what not (like a commit that also creates dirs).
Does anyone know if this can be a problem or a solution?
--
Nicola Ken Barozzi [EMAIL PROTECTED]
- verba volant, scripta manent -
(discussions get forgotten, just code remains)
---------------------------------------------------------------------