Just one person taking a coffee break to go get coffee obviates any savings. The free coffee isn't there for the staff, it is there to save the org money.
The federal government is so worried about the appearance of spending excess taxpayer money that employees set up water clubs and coffee clubs where the employees all chip in for a water cooler and a coffee machine every month.
So, this particular contractor was big into efficiency and cost savings, to the point that we were all required to do a project where we made something more efficient and documented how much money we saved in doing so. The whole thing was mostly bullshit, but one of the interesting things I learned by doing it was that saving engineering time was essentially $0, because from the company's perspective we were all salaried, and as long as the time we were saving was charged to the same contract we would be charging our other work to, it didn't matter.
From that perspective, which I do not agree with, the cost of coffee breaks was also $0, while the cost of providing it was not, so no coffee. On one program I was on we at least managed to get facilities to install a commercial Bunn machine for us, but we were still responsible for buying all the supplies.
Under that simplistic model there is no space for higher quality, higher reliability, etc. Then "efficiency" is one dimensional and you have basically no agency. The cost model should be agreed upon collectively.
You're right, and there were definitely situations where we could have done things better, but didn't, because the impetus wasn't there. It was really frustrating, and I wish I'd been able to get out sooner.
> -2000 lines
Not sure what this is a reference to, but one of the fun things I tripped over while there was that code changes were valued on dollars per LOC. Changesets with net negative LOC were... problematic. I think I once had a change that was near -2000 LOC net.
