Conversation

Notices

1. Ekaitz Zárraga 👹 (ekaitz_zarraga@mastodon.social)'s status on Saturday, 07-Apr-2018 14:02:21 UTC Ekaitz Zárraga 👹

   • jartigag

   @jartigag LoC but it's not fair.

   What *I do in R&D* is make development in small steps and I document at the same time. When I made some documentation I feel like: Ok, I did some new stuff. And I feel good.

   • jartigag repeated this.
   • Jon Dubovsky (jond@mastodon.social)'s status on Saturday, 07-Apr-2018 14:13:39 UTC Jon Dubovsky

     • jartigag

     @jartigag Our big-picture number is just person-days. Since we track those it's easy enough to ask "how many person-days went into the control software for X?", etc.

     It is poorly-correlated to SLOC. For hard problems you spend so much time researching only to produce like 50 SLOC. Easy problems like "implement this state machine" is a lot of very-nearly-boilerplate.

     (I actually care far more about the distribution of completion time compared to the work estimate range!)

     jartigag repeated this.