Re: agile/xp question (formal analysis)

From: AndyW (foo__at_bar_no_email.com)
Date: 11/27/04 

Date: Sun, 28 Nov 2004 10:53:52 +1200

On Sat, 27 Nov 2004 14:36:30 -0500, Ronald E Jeffries
<ronjeffries@acm.org> wrote:

>On Sat, 27 Nov 2004 12:42:12 +1200, AndyW <foo_@bar_no_email.com>
>wrote:
>
>>>But I don't see why nothing could be done end to end /in principle/
>>>though of course once you split out separate framework from
>>>application, you're pretty much screwed for incremental delivery.
>>
>>Not quite - what tends to happen is you have many teams each working
>>on a feature or some componant of the application. Each team would be
>>following a methodology (usually the same across teams), so in this
>>light you could perform agile/mini waterfall.
>>
>>Teams work in parallel to each other, so the size and speed is
>>dependant on the number of teams you have working at any given time.
>>Once a team finishes a feature it moves on to the next one.
>>
>>At this level a team is defined as maybe 20 or 30 people and a feature
>>would be some major component of architecture - for example the
>>database system, the framework, and application business objects such
>>as the customer services object, product definition object etc.
>>
>>Think of one of these teams as being your 'average' inhouse
>>development team for a small project. Each may be split up into
>>sub-teams (of 6 people) with a line manager and there will be a couple
>>of BAs a team manager, an architect etc.
>
>Then if an average inhouse team of six can produce three to six new
>working tested features in a week, why couldn't a bunch of parallel
>teams do that?

Because of all the extra work that needs to be done on top of that.
Both in technical and non-technical.