Jeff Turner wrote, On 01/07/2003 11.16:
On Mon, Jun 30, 2003 at 03:34:08PM +0200, Nicola Ken Barozzi wrote:...
Nope, it makes sense to remain separate as now.
Yep, fully agree with you that Cocoon and Forrest have distinct identities and shouldn't merge any time soon.
However, most of my email wasn't about that.
Sorry, I misread I guess.
I'm just suggesting copying some Forrest files into Cocoon. It's like a disk cache in an operating system, where some data gets duplicated closer to where it's used, in order to improve performance. Every now and then, we synch the 'cache' (Cocoon CVS) with the 'disk' (Forrest CVS). That synch price is lower than the cost of Cocoon developers having to constantly access the 'disk'.
Ok, now I get it, I thought the other way around, as you started by assuming now that Cocoon is the disk from where you get updates.
I remember we have already had this discussion. The major point was that documentation has to be able to be written and published regardless of the state of the Cocoon CVS. Having a separate and "stable" doc system was regarded as a plus, and I still stand by this idea.
What's so difficult for Cocoon users to have to
cd xml-forrest cvs update build
for every Forrest update?
For us, we can do a similar thing, by adding a script to the Ant task for cocoon synchro, as you have done with the script. So anyone can test Cocoon changes from Forrest anytime.
Gump should also be able to run a Forrest test and see that it works with latest Cocoon, no?
--
Nicola Ken Barozzi [EMAIL PROTECTED]
- verba volant, scripta manent -
(discussions get forgotten, just code remains)
---------------------------------------------------------------------