Re: [scrumdevelopment] Why Traceability? Can it be Agile?
- Not related to life insurance (which I have, not that it's relevant).Traceability matrix is not an insurance policy.I repeat, I have, in 30 years in the business, in industry, research and a little bit in government, NEVER seen a traceability matrix pay off.If you have a different experience, please tell us when you used it and it paid back the money spent on creating and maintaining it.AlistairIn a message dated 3/11/2004 8:38:28 AM Mountain Standard Time, tiseo.paul@... writes:
From: acockburn@... [mailto:acockburn@...]
Sent: Wednesday, March 10, 2004 3:30 PM
> I have to say I have always found traceability to
> be a waste of money.
> As a result of these sorts of investigations, I challenge
> the usefulness of traceability.
I'm assuming you also don't carry life insurance? :) None of my thirty-ish
and forty-ish friends with life insurance have had the misfortune of having
their life insurance policy exercised. Does that mean they don't need proper
life insurance because it takes some research to set up properly, costs
something and has no immediate monetary benefit?
I find some processes in software development are not about "immediate
cost-benefit", but are for "risk impact mitigation". (I use quotes because
I'm not sure of which is the right buzzword to insert there.)==============================================
President, Humans and Technology
1814 E. Fort Douglas Circle, Salt Lake City, UT 84103
Phone: 801.582-3162 Fax: 775.416.6457
"Surviving Object-Oriented Projects" (1998)
"Writing Effective Use Cases" (Jolt Productivity Award 2001)
"Agile Software Development" (Jolt Productivity Award 2002)
"La perfection est atteinte non quand il ne reste rien a ajouter,
mais quand il ne reste rien a enlever." (Saint-Exupery)
- Brad Appleton wrote:
> Mike Beedle wrote:Brad:
>> Reversibility and traceability are great concepts but are hard to
> Sure - at the level you are claiming is being asked for. Problem is
> that's NOT what's being asked for! It IS POSSIBLE to do
> lightweight/lean traceability. I've done it, Alistair said he's seen
> it done. And I know of many others that done it too. No one here asked
> for "perfect" traceability or "full" - just plain old "good enough"
> and "barely sufficient" traceability. Alistair described one way. I
> described another. So let's get off the "IMPOSSIBLE" kick shall we
> because its already been disproven.
I would like to see a couple of working examples to really assess
Please don't misunderstand me, I like the idea of doing:
just plain old "good enough" and
"barely sufficient" traceability.
(That's I am trying to do with Balanced Agility with Scrum, "good
and "barely sufficient" Agile Software Development ;-) But
I also think we should present the arguments as to *why*
thorough traceability and reversibility are impossible, at least
with our current tools and environments.
Perhaps a good Open Source Eclipse plug-in will do for bare-bones
traceability .... with some minimal automated "registration
system" per class? I don't know of such a thing, unfortunately.
NOTE: If most people are wondering why the "fireworks" are
going off in this thread, is because it was *precisely* this kind
of arguments about software being "traceable and reversible"
from many artifacts that fueled the movement of Agile Software
Through the 90's some proponents in the industry claimed that
their processes and tools would deliver "traceable and
reversible software" across many artifacts, but many of us at
the other side of the fence were asking:
1) at what cost?
2) with what purpose?
3) with what benefits?
That is a significant part of the "Agile soapbox",
"Writing is re-writing."