[MUD-Dev] distributed objects
J C Lawrence
claw at kanga.nu
Wed Feb 16 01:04:05 New Zealand Daylight Time 2000
On Tue, 15 Feb 2000 23:24:54 -0500
Charles Hughes <charles.hughes at bigfoot.com> wrote:
> I'd vastly prefer that the programmers were punished for leaving
> the bug in (or active after being reported), rather than punishing
> the players for exploiting it. We are, unfortunately, in a very
> small minority.
ObStory: Early in my contracting career I spent the better part of a
decade working for a company that had some simple standards it
required of its programmers and the code they wrote in order for
them to be paid:
-- All code must pass lint at full strength with no warnings or
-- All code must follow the comany naming conventions (a variation
on Hungarian notation that actually predates Hungarian Notation and
is quite a bit simpler and more readable) and code formatting rules
(a simple F&R variant).
-- Functions were written one per source file, with an #ifdef'ed
test driver for that function in the same file that excercised ALL
possible flow paths (minor exceptions were made for UI and other
high level functions).
-- Zero bugs, with the definition of "bug" determined by QA and
marketing, not the programmer.
All the contracts were fixed price with 50% paid after the product
passed QA (code review, standards compliance, and end user testing)
with ALL bugs found fixed, and the remaining 50% paid after 45 days
of market sales with all bugs found fixed.
It made for very careful methodical disciplined programmers. Too
many bugs meant you had problems meeting rent and that your next
contract would be at a lower rate, endangering your living standards
even more. Keep your bug rate down however, and your billing rate
and your bank accounts grew quite nicely.
A nice direct and simple feedback loop.
J C Lawrence Home: claw at kanga.nu
----------(*) Other: coder at kanga.nu
--=| A man is as sane as he is dangerous to his environment |=--
MUD-Dev maillist - MUD-Dev at kanga.nu
More information about the MUD-Dev