It occurred to me, as I was wending my weaving, wobbling way from the Tube station back to the hotel, just how much effort is being duplicated throughout the world by web developers (who probably don't come from a software engineering background) re-inventing assorted wheels.
I faced an interesting problem today... a highly experienced chap (who's gone on holiday for a few weeks) had forwarded a link to somebody's solution to the basic problem, but I could tell straight away that it didn't require that much code.
My researches into the fundamental nature of the problem, and the algorithms available to solve it, led me into a number of recondite areas that wouldn't appear immediately relevant until examined closely - but I reckon, as a result of sifting the wheat from the other wheat, that I can do the whole job, better than spec, in about 1/10th as much code - and my code will be clearer and more easily maintainable and reusable (mainly because there won't be so much of it to understand).
I might see about presenting on this theme at a conference or two... something along the lines of "What you don't have to do, because the computer scientists did it for you years ago."
Or, given that a few searches and a critical mind (rather than access to a university library) are all it takes to find these things nowadays, maybe I should go along the lines of "The Correct Use of the World Wide Wheel Repository."
I faced an interesting problem today... a highly experienced chap (who's gone on holiday for a few weeks) had forwarded a link to somebody's solution to the basic problem, but I could tell straight away that it didn't require that much code.
My researches into the fundamental nature of the problem, and the algorithms available to solve it, led me into a number of recondite areas that wouldn't appear immediately relevant until examined closely - but I reckon, as a result of sifting the wheat from the other wheat, that I can do the whole job, better than spec, in about 1/10th as much code - and my code will be clearer and more easily maintainable and reusable (mainly because there won't be so much of it to understand).
I might see about presenting on this theme at a conference or two... something along the lines of "What you don't have to do, because the computer scientists did it for you years ago."
Or, given that a few searches and a critical mind (rather than access to a university library) are all it takes to find these things nowadays, maybe I should go along the lines of "The Correct Use of the World Wide Wheel Repository."
Comment