Re: [Vrs-development] vrs momentum

[Eric Altendorf]

<agile development methodology rant>

I mentioned this in private communication, so I'll mention it again 
here publicly.  The project we're looking at is so big, and our 
resources limited, that we're at a severe risk of getting caught in 
analysis paralysis, trying to do too much planning and design and 
never actually developing anything.

I'd like to vote for the agile development principle of always having 
a working piece of software.  What this means is that we should start 
with a minimal design -- a design for a tiny sub-project that we can 
design in an hour.  Write up a design for that, implement it in the 
next week, make sure it works, then start adding functionality bit by 
bit.  We can have design talks on bigger stuff proceeding in 
parallel, but we gotta make sure we're coding and we have a working 
system at all times.

If we try to do up-front design of large pieces of this project we'll 
get quagmired.  Top priority is to have a working system, even if it 
isn't yet exactly what we want.

</obnoxiousness ;-) >

eric

On Friday 19 July 2002 21:29, Ian Fung wrote:
> so i guess i should maybe start writing up some requirements and
> then have people design and implement parts of the system? what do
> you think the ETA is on the gw middleware? i kinda like to check it
> out before i do any design. is there a way i can get web access so
> i can post to the website and also upload irc logs. maybe when i
> get something written we can meet on irc again to discuss it. of
> course if anyone else would like to start writing stuff too that
> would be great.
>
> ian


-- 
"First they ignore you.  Then they laugh at you.
 Then they fight you.  And then you win."             -Gandhi