5 ms·

>I understand some of our code is under tested, but unlike TDD wankery, there is a real business cost to spending a bunch of time writing tests and, since those

>I understand some of our code is under tested, but unlike TDD wankery, there is a real business cost to spending a bunch of time writing tests and, since those systems are either mostly finished / slowly changing, or on their way out, is this candidate going to be difficult or suck it up and deal? (Yes, this happened. See also scut work and/or the realities of living in our fallen world.) I had a feeling that somewhere along the list of "candidate must not make us put up with any of their bullshit" there would be something saying "candidate must be willing to put up with a metric ton of our bullshit".

x0x0 11y ago
Allow me to highlight what I wrote: either mostly finished / slowly changing, or on their way out Most of those old systems support a part of the business that is dying and are maintenance only, but occasionally they do need some extra duct tape. We minimize any investment at this point. At this point it's a day or two of work a month spread across a team, but everyone takes a turn.
- crdoconnor 11y ago
  I don't see the point. Either shut it down completely because it's not worth the investment to maintain or do it properly. Test driven development is a practice that is most valuable on legacy code. Only doing it on new code and not old code is ass-backwards. If anything it should be the other way around. A half assed hack job to support remaining customers will likely only end up alienating them when you inevitably break stuff they rely upon with duct tape. Alienated customers tend to drop you forever and run into a competitor's arms. YMMV but that's always been my experience.
  - x0x0 11y ago
    The point is money. It shouldn't be hard to imagine a business that is earning money but will never again grow, is worth no further investment, and is therefore being run out. I reread what I wrote; you appear to be deliberately obtuse.

x0x0 11y ago

Allow me to highlight what I wrote: either mostly finished / slowly changing, or on their way out Most of those old systems support a part of the business that is dying and are maintenance only, but occasionally they do need some extra duct tape. We minimize any investment at this point. At this point it's a day or two of work a month spread across a team, but everyone takes a turn.

crdoconnor 11y ago
I don't see the point. Either shut it down completely because it's not worth the investment to maintain or do it properly. Test driven development is a practice that is most valuable on legacy code. Only doing it on new code and not old code is ass-backwards. If anything it should be the other way around. A half assed hack job to support remaining customers will likely only end up alienating them when you inevitably break stuff they rely upon with duct tape. Alienated customers tend to drop you forever and run into a competitor's arms. YMMV but that's always been my experience.
- x0x0 11y ago
  The point is money. It shouldn't be hard to imagine a business that is earning money but will never again grow, is worth no further investment, and is therefore being run out. I reread what I wrote; you appear to be deliberately obtuse.