axis-java-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Rob Jellinghaus <r...@unrealities.com>
Subject Re: IRC Meeting Log
Date Wed, 24 Oct 2001 17:40:07 GMT
At 02:51 AM 10/24/2001 -0400, Sam Ruby wrote:
>> he hasn't sent it to me yet).  Once I get his code, I'll be working on
>> integrating it into Axis, using my at-home Perforce source control system
>> to coordinate his code with the moving Axis baseline.
>
>I'd like to strongly discourage this approach.  The longer this code is
>developed independently, the harder it will be to merge and the less
>opportunity for people to provide feedback as to the impact of changes.
>
>Suggestion: there has to be some parts of Axis, as currently implemented,
>which makes attachments unnecessarily harder than necessary to implement.
>Consider changing those parts first.  True that such changes won't directly
>provide support for attachments, but will allow us to get quickly past any
>system instability  that such changes will make (which by definition would
>be due to insufficient test coverage).

Sam, you know me.  I *implemented* the functional test framework :-)  I couldn't agree more
about the desirability of early and often testing.

But right now, the code isn't testable in the Axis test framework.  Once some of it is, I
will definitely check it in.  This will certainly happen significantly before attachments
actually work.

I envision:
- Establish an Axis-hierarchy location for Rick's code
- Write some simple unit tests for it
- Check in
- Commence stepwise reorganization of code as necessary for attachments
- Check in incrementally as often as possible

I don't want to have a private branch going for any longer than absolutely necessary, have
no fear about that!

Cheers!
Rob


Mime
View raw message