…is that it takes forever to achieve and by the time you deliver it, the world you built it for is gone.
Better to make a quick best guess, ship it and then iterate it in concert with the changing of the world.
Technorati Tags: philosophy, architecture, technology, design
Entries (RSS)
May 17th, 2007 at 7:17 pm
Vice versa the trouble with “good enough” is that it’s arbitrary. Once you shipped, why bother improving? (yes, it’s a rethorical question.)
May 17th, 2007 at 10:29 pm
Dan, perfection takes time. It took Toyota half a century to unseat GM.
May 21st, 2007 at 9:19 pm
Is perfection subjective? If it is, whose opinion should we use? If it isn’t, what is it? How can it be encapsulated? It is possible to specify perfection? If not, how do we know whether we have attained it? How are we able to tell when something cannot be improved? We certainly cannot prove that anything is flawless.
Isn’t a solution where one could reasonably say that ‘on balance, it was the best that I/we could have done given the knowledge and practical constraints I/we had at the time’ a kind of satisfactory, albeit transitory, perfection in itself?
May 22nd, 2007 at 9:58 am
James said:
“Is perfection subjective? If it is, whose opinion should we use? If it isn’t, what is it? How can it be encapsulated? It is possible to specify perfection? If not, how do we know whether we have attained it? How are we able to tell when something cannot be improved? We certainly cannot prove that anything is flawless.”
Yes, perfection is almost always subjective - the key issue is to find the appropriate stakeholders who can give you feedback in respect of how well you’re doing. This might be customers (as in paying for your product/service) or executives (who for example want reduced time to market but don’t know how best to achieve it - interestingly this tends to focus people on speeding up the build cycles but few think about speeding up deployment).
“Isn’t a solution where one could reasonably say that ‘on balance, it was the best that I/we could have done given the knowledge and practical constraints I/we had at the time’ a kind of satisfactory, albeit transitory, perfection in itself? ”
May 22nd, 2007 at 2:58 pm
Dan said: “Transitory perfection isn’t perfection”
If perfection is subjective and the opinion probably best used is the stakeholders’ (of course) then you must also be transitively linked to THEIR transitory ideas of perfection (”just a little hindsight or one more fact renders it imperfect” to them also), which of course are the functions of expectations of others (and so on). The very pursuit of the definition of perfection to me a bit like a wild goose chase.
I would have been less puzzled if you had said the trouble with ‘project-based development or long delivery cycles is”.
When it comes to evaluating the delivered artefact:
“Is it fit for purpose*”
Yes: leave it alone
No: instigate action
* extremely difficult to define in itself, no? The answer has to reflect a weighted balance of short and long term goals which may conflict with each other. How is fitness for purpose decided and co-ordinated by whom? What ‘role’ has this remit?
I am not in disagreement with your statement on iterative development.
May 23rd, 2007 at 9:02 am
James said:
“If perfection is subjective and the opinion probably best used is the stakeholders’ (of course) then you must also be transitively linked to THEIR transitory ideas of perfection (”just a little hindsight or one more fact renders it imperfect” to them also), which of course are the functions of expectations of others (and so on). The very pursuit of the definition of perfection to me a bit like a wild goose chase.”
“I would have been less puzzled if you had said the trouble with ‘project-based development or long delivery cycles is”.”
Perhaps because that would give you a concrete example to work with you’d be less confused. However this is but one example, there are lots of others which also deserve thought.
“How is fitness for purpose decided and co-ordinated by whom? What ‘role’ has this remit?”
You’re looking for a definitive/certain position again
Something is only going to be fit for purpose until some fact emerges to indicate it’s not. Perhaps this fact never emerges, perhaps it does. The key is accepting it might so that you say “it’s good enough for now” AND put measures in place to determine when it’s no longer good enough and take action.