[Sussex] XP stuff [was]Gentoo problems.

Steve Dobson steve at dobson.org
Wed Feb 25 10:06:25 UTC 2004


Mornin' Geoff

On Wed, Feb 25, 2004 at 09:40:06AM +0000, Geoff Teale wrote:
> I think the main discrepancy in our thought is that your are thinking 
> of XP as one rigid collection of methods.

Not quite, but go on.

>                                             I don't think any 
> methodology can be applied unchanged to all projects.  I merely state 
> that some large subset of the XP methods can be applied in most 
> situations and that there are advantages to that approach.  

Agreed.
 
> Most importantly I think that XP makes clear some good solutions to
> problems that dog almost all software developments.

I'm not sure what you mean by "makes clear."  The problem with other
design methods (like JSP) is that they have a linear process.  First you
design, then you code, last you test.

The XP process recognises that external forces are going to change the
requirements, and tries to minimize their impact.
 
> We certainly don't do XP here exactly as the books say - 

I wouldn't expect you to.

> I think it would be unrealistic to think that XP could be applied 
> anywhere precisely as the books say.

I wouldn't argue that it should.  But how much of XP can you tailor
before it stops being XP?  Remove peer programming, short release 
cycles and make the design documentation a requirement and you have
something that looks more like JSP than XP.

Steve




More information about the Sussex mailing list