Jan. 5, 2017
7:50 a.m.
Tim Parenti wrote:
These cases have been sufficiently rare that I think it's worth documenting them as we go, so they're easier to find once we feel '%z' is sufficiently well-supported we can rip them out.
Thanks. A downside of these detailed comments is that we'll have to maintain the comments, along with the data, if the data changes before we start assuming %z. So how about if we instead use simpler comments that don't contain detail that we'd have to maintain? That should be good enough to document the situation. Something like the attached, say.