Challenge 1 — Solution

Task: A kanji page is generated and written to its live-site location
only, with the archive copy skipped "to save a step." Using K2,
explain exactly what real, concrete problem this creates later.

K2 requires every kanji page to be written to two locations at once,
kept identical: an archive/reference copy (deliberately excluded from
the live site's own routing, so it's never actually served to a
visitor) and the live copy that visitors genuinely reach. These serve
two different, non-overlapping purposes - the live copy is the actual
served page, while the archive copy is a permanent, separate reference
collection of every kanji page ever generated, independent of whatever
happens to the live site later.

Skipping the archive write "to save a step" means that permanent
reference collection silently stops being complete the moment this
happens - the archive folder now has a real gap in it that doesn't
match what's actually live. Concretely: if the live site were ever
reorganized, migrated, or a live kanji page were accidentally
overwritten or lost, there would be no separate, trustworthy archive
copy of this specific character's own page to recover it from - the
archive's entire value depends on it staying complete and not silently
falling behind what's actually been generated.

Notes:
- This is structurally the same kind of problem the source of truth
  distinction from Chapter 1 describes - the archive copy is meant to
  be a durable, independent record, and a durable record that quietly
  stops being updated isn't actually durable anymore, it just still
  looks like it is until someone needs it.
- K2 frames writing to both locations as "a cheap extra step, not a
  meaningful cost" specifically because the realistic alternative -
  discovering the archive is incomplete only once it's actually needed
  - is a much more expensive problem to fix after the fact.