> From: Schaible, J�rg [mailto:[EMAIL PROTECTED] 
> 
> Hi folks,
> 
> the topic was not intended to start any flame wars, but I'll 
> face some team members with the 
> we-won't-depend-on-anyone-else's-code syndrom.
> 
> My current basic points:
> - principal and patterns described in DaW
> - lifecycle management
> - configuration capabilities (1)
> - service for logging, instrumenting, pooling
> - other available services ready to use(2)

Well, a hard-line "we-won't-depend-on-anyone-else's-code" is
useless - you wouldn't be able to depend on Xerces, Java or
anything else - so I assume that what we have here is a 
"let's keep dependencies to a minimum" argument.

My own rationale for adopting Avalon was similar to your
basic points, with the difference that I didn't care much
for "other available services ready to use". But what I did
want was:

 + Good solid design patterns.

 + Configuration support.

 + Pluggable implementations of defined interfaces.

 + Unified logging.

And Avalon Framework provided all that. I motivated
the decision with "Look, if we don't use Avalon, we
have to invent something precisely like it ourselves.
Let's go for it, and if it becomes a bother, fork.
That way, we may gain something, and we're sure not
to lose anything."

And that pretty much worked.

I'd emphasize the smallness of Framework and get people
going with that first - once they see how easy it is
to use (especially configuration - I *love* that package)
I think they'll accept it. Emphasize low-committment and
potentially huge gains.

/LS


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to