Re: Extreme Programming (was: Re: Consultancy company)

2001-01-21 Thread Piers Cawley

David Cantrell [EMAIL PROTECTED] writes:

 On Sat, Jan 20, 2001 at 12:24:24AM +, Piers Cawley wrote:
 
  Now, I freely admit that I have partaken of the Extreme Programming
  Kool-Aid, and dammit I want to do it.
 
 I want to try it too.  I'm not convinced by all of it - pair programming
 for example - but so much of the other stuff seems damned sensible that
 I want to give it a go.  Including pair programming.  I'm trying to keep
 an open mind on that fucking stupid idea.

When they got the permie in who's taking over the project I'd been
working on, we spent a fair amount of time doing the PP thing. And it
was great. A *fantastic* way of getting information shared and passed
on for what was basically a decently engineered but atrociously
documented project. By the time I left, James knew his way around the
system and was confident he could extend it as required. And I was
confident he was right about that. (Did my ego good to know I'd
written something without sanity checking that was relatively easy for
someone to pick up quickly too...)

-- 
Piers




Re: Extreme Programming (was: Re: Consultancy company)

2001-01-21 Thread Piers Cawley

"Dean S Wilson" [EMAIL PROTECTED] writes:

 -Original Message-
 From: Aaron Trevena [EMAIL PROTECTED]
 
 
 I did a little pair programming at emap - I probably wasn't doing it
 right
 tho'. even so we did get thru the hard bits quicker and could split
 up to
 do the easy stuff. I think it made a difference but then I was mostly
 being a backseat coder so either we did okay or stuart was very
 tolerant
 indeed.
 
 
 How did you establish who would make good pairings? Was it done by
 trying to place two equals or was it done more on a mentoring level of
 a very experienced coder and a less experienced one? (I've not read
 that much on XP)

The Dictum in the XP literature appears to be 'nobody is allowed to
say "No"', pairs form and re form on task by task basis. 'Regular'
pairs are to be discouraged. XP Installed has a bunch of stuff on
this.

 Has anyone who's used XP had a client that was willing to make an
 employee available pretty much full time or was it more they come in
 for a chunk of the afternoon three times a week? I have an issue with
 the fact that clients will be willing to pay a member of staff to
 spend all day in the consultants office in case they need to be asked
 questions.

The XP argument goes something like:

This team costs you X000/day. Your liason costs you X00/day. We
believe that having someone available to us, on site, full time (but
able to do however much of their work that can be done remotely), will
dramatically reduce the amount of our time it takes to deliver a
product, and will also increase the final value of that product. Do
the maths. (Well, maybe not quite so bluntly, but you take my point) 

 I'm not saying its a bad thing to have someone on hand, I can see
 its uses but from the clients point of view why not just have
 contact by phone/email. That was the liaison has access to everyone
 in his base office so he can resolve issues faster with more
 authority than if he were in your offices. Also you have a paper
 trail of requests, questions and responses.

There's a whole chapter on this in XP Installed. Paraphrasing,

  "Customer onsite == answer in 30 seconds.
   Customer offsite == answer today."

They also point out that you can make either version work, but the
onsite customer option works best.

-- 
Piers




Extreme Programming (was: Re: Consultancy company)

2001-01-20 Thread David Cantrell

On Sat, Jan 20, 2001 at 12:24:24AM +, Piers Cawley wrote:

 Now, I freely admit that I have partaken of the Extreme Programming
 Kool-Aid, and dammit I want to do it.

I want to try it too.  I'm not convinced by all of it - pair programming
for example - but so much of the other stuff seems damned sensible that
I want to give it a go.  Including pair programming.  I'm trying to keep
an open mind on that fucking stupid idea.

-- 
David Cantrell | [EMAIL PROTECTED] | http://www.cantrell.org.uk/david/

   Any technology distinguishable from magic is insufficiently advanced



Re: Extreme Programming (was: Re: Consultancy company)

2001-01-20 Thread Dean S Wilson

-Original Message-
From: Aaron Trevena [EMAIL PROTECTED]


I did a little pair programming at emap - I probably wasn't doing it
right
tho'. even so we did get thru the hard bits quicker and could split
up to
do the easy stuff. I think it made a difference but then I was mostly
being a backseat coder so either we did okay or stuart was very
tolerant
indeed.


How did you establish who would make good pairings? Was it done by
trying to place two equals or was it done more on a mentoring level of
a very experienced coder and a less experienced one? (I've not read
that much on XP)

Has anyone who's used XP had a client that was willing to make an
employee available pretty much full time or was it more they come in
for a chunk of the afternoon three times a week? I have an issue with
the fact that clients will be willing to pay a member of staff to
spend all day in the consultants office in case they need to be asked
questions. I'm not saying its a bad thing to have someone on hand, I
can see its uses but from the clients point of view why not just have
contact by phone/email. That was the liaison has access to everyone in
his base office so he can resolve issues faster with more authority
than if he were in your offices. Also you have a paper trail of
requests, questions and responses.

Is the Monday night meeting still on for those of us who can't make
the lunch time one?

Dean
--
Profanity is the one language all programmers understand.
   ---  Anon




Re: Extreme Programming (was: Re: Consultancy company)

2001-01-20 Thread Dave Hodgkinson

"Dean S Wilson" [EMAIL PROTECTED] writes:

 -Original Message-
 From: Aaron Trevena [EMAIL PROTECTED]
 
 
 I did a little pair programming at emap - I probably wasn't doing it
 right
 tho'. even so we did get thru the hard bits quicker and could split
 up to
 do the easy stuff. I think it made a difference but then I was mostly
 being a backseat coder so either we did okay or stuart was very
 tolerant
 indeed.
 
 
 How did you establish who would make good pairings? Was it done by
 trying to place two equals or was it done more on a mentoring level of
 a very experienced coder and a less experienced one? (I've not read
 that much on XP)

The latter. You mix skills. And the second isn't idle. He's coding up
the test cases.

 Is the Monday night meeting still on for those of us who can't make
 the lunch time one?

Oh, yes.

Leon, are you acting as scribe?

-- 
Dave Hodgkinson, http://www.hodgkinson.org
Editor-in-chief, The Highway Star   http://www.deep-purple.com
  Apache, mod_perl, MySQL, Sybase hired gun for, well, hire
  -



Re: Extreme Programming (was: Re: Consultancy company)

2001-01-20 Thread Leon Brocard

Dave Hodgkinson sent the following bits through the ether:

 Leon, are you acting as scribe?

Yes. Don't expect a masterpiece though.

Leon
-- 
Leon Brocard.http://www.astray.com/
yapc::Europehttp://yapc.org/Europe/

... All new improved Brocard, now with Template Toolkit!