When speaking about speed losses or speed improvements one could use such an example for discussing the "scale" for comparisons. What is more appropriate, a "linear" scale (differences) or a "proportional" scale (factors)?
Here, fore example, 125 ms against 1888 ms for a special task.
Some of you may say: Two seconds for such a task is fast enough, I use this never or seldom, I don't mind.
Others may say: This is inacceptable, a factor of 14 is an increase of 1400%. I need such tasks or similar ones often, my time budget isn't wide enough to cover this.
For me both arguments are good, depends on the personal situation. Perhaps we should use both, in parallel, display/denote absolute values (differences) and proportional values (factors).
Say the increase of time needed for a special/typical task is from 1 second (6.7) to 14 seconds (7.0). Then everybody can judge whether he's willing to pay this prize for using LC 7:
- (Uses difference). The 13 seconds for some user A once a week or month may not matter.
- (Uses factor). The time increase for another user B for such tasks, from half a working day per year to 7 working days per year will certainly matter.
Currently there is a benchmarking project in progress (IDE contributors subforum). Please support this project started by 'Malte' and 'FourthWorld'.
I already contributed. Makes me feel somehow more "enabled" to criticise -- or praise, if applicable.
