blob: 90073aad506e8eadec385fa241e48882d57d9c56 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
|
#+TITLE: ADR-3: Drill direction
#+DATE: 2026-04-30
#+STATUS: Accepted
* Context
=gloss-drill-export-all= tags every entry for =org-drill=. The card
type controls which direction(s) the drill quizzes — term-to-definition
(recognition), definition-to-term (recall), or both.
* Decision
Set =:DRILL_CARD_TYPE: twosided= on every exported entry.
=org-drill='s twosided card type alternates direction on consecutive
visits, so a single entry quizzes both directions over time.
No per-entry override in v1. Every entry exports as twosided.
* Consequences
*Positive.*
- Tests both recognition and recall over time without doubling the
deck size — one entry, both directions, alternated by =org-drill='s
scheduler.
- No per-entry decision burden during =gloss-add= or
=gloss-drill-export=.
*Negative.*
- Some terms make more sense one-directional. A user who saved
"SBIR" → "Small Business Innovation Research" probably wants
recognition only (acronym → expansion), not recall (expansion →
acronym). v1 doesn't accommodate this.
- A future v2 may want a per-entry =:DRILL_CARD_TYPE:= override or a
defcustom for the default. Both are additive changes.
* Alternatives Considered
*=:DRILL_CARD_TYPE: simple= (term → definition only).* Rejected:
half the value of the cards, and the user can always skip-back
during drill if they don't want the reverse direction.
*Two cards per entry, one each direction.* Rejected: doubles the
deck size; adds maintenance burden when an entry is edited (now
two cards diverge).
*Cloze cards.* Rejected: needs a marked deletion in the body, which
fights the "save once, drill forever" promise.
|
