Live data from Hacker News

How to complete your PhD (or any large project): Hard and soft deadlines, and the Martini Method

academicproductivity.com

1–4 of 4 posts

Re: How to complete your PhD (or any large project): Hard and soft deadlines, and the Martini Method

#2
Applying the Martini Method to coding is not so simple: as has been pointed out countless times, programmer productivity can't be measured in lines of code produced.

But maybe something like

   svn diff | wc -l
is a good enough stand-in. Replacing a big section of repetitive code with something smaller and more elegant would then count as productivity. Of course, then it becomes an incentive to write dumb, repetitive code, check it in, and then rewrite it the right way, but probably only for the programmer with more alcoholism than self-respect.

Re: How to complete your PhD (or any large project): Hard and soft deadlines, and the Martini Method

#3
post #2

Applying the Martini Method to coding is not so simple: as has been pointed out countless times, programmer productivity can't be measured in lines of code produced. But maybe something like svn diff | wc -l is a good enough stand-in. Replacing a big section of repetitive code with something smaller and more elegant would then count as productivity. Of course, then it becomes an incentive to write dumb, repetitive co…

I use a variant of the Martini Method where I try for at least 2 substantive svn commits a day (not including minor bugfixes or cosmetic issues). I try to do atomic commits, so this correlates very closely with function points. It seems to work quite well; since the beginning of the year I've gone from r569 to r623. It's had a corresponding effect on observed product quality, having gone from essentially a blank webpage to the full editor skeleton, a working game, and several widgets for customizing it.