Re: Indexing new entry with a size greater than SETTINGS_HEADER_TABLE_SIZE

Yup, Tatsuhiro is right-- it does just move the condition around (actually,
it changes the structure of the code a bit-- in my code MakeSpace() returns
a boolean indicating if it can or can't make enough space, which hid the
test. This phrasing would change that).

I still like it less, but in one special case it does better and for no
overall complexity cost.

Tatsuhiro, will you create an issue for it?
-=R