The term OverdoseEffect?
was introduced to me by AlistairCockburn
. I believe as the pattern promoters we failed to more prominently state the side effects and possible overdose issues related with pattern use.
Do we need a pattern FDA to give out laws about how to describe patterns or should we use double size print for the PatternConsequences?
I do not know: should I feel guilty or innocent about PatternAbuse
Not enough emphasis is given to the context
in the GoF book. If one applies a pattern only in its appropriate context, PatternOverdose
is not an issue.
IMHO, new pattern books should use a form closer to that of ChristopherAlexander
's books, in which the context is described straight away, and in great detail.
I would guess that most of ChristopherAlexander
's patterns are harder to overdose than the GoF patterns. Unless you have a building with tens of entrances it would be hard to overuse a pattern like "EntranceTransition?
". But it's real easy to go overboard with AbstractFactory
when you first come in contact with it.
Something that is even more prominent, horrible, and irreversable is PatternInconsistency
). This is where a system uses patterns so inconsistently, so out of context (as mentioned above), with different naming conventions (listener sometimes observer others), and sometimes rigorously then other times not that the end result becomes an AmorphousBlobOfHumanInsensitivity
. The problem isn't the dosage of patterns but the resultant chaos from using many inconsistently and haphazardly. --RobertDiFalco