diff options
| author | Simon Tatham <anakin@pobox.com> | 2009-09-16 11:06:02 +0000 |
|---|---|---|
| committer | Simon Tatham <anakin@pobox.com> | 2009-09-16 11:06:02 +0000 |
| commit | 466c398a00de9b6df3d0846d55089ff419fd25ba (patch) | |
| tree | b45f66cbc57dd853856ba727028669f57651f54e | |
| parent | ddfa693b0fd4a0902ba85414dc5bbefc5bb384cf (diff) | |
| download | puzzles-466c398a00de9b6df3d0846d55089ff419fd25ba.zip puzzles-466c398a00de9b6df3d0846d55089ff419fd25ba.tar.gz puzzles-466c398a00de9b6df3d0846d55089ff419fd25ba.tar.bz2 puzzles-466c398a00de9b6df3d0846d55089ff419fd25ba.tar.xz | |
Minor rewording and typo-correction.
[originally from svn r8654]
| -rw-r--r-- | tents.c | 15 |
1 files changed, 9 insertions, 6 deletions
@@ -1964,11 +1964,14 @@ static int *find_errors(game_state *state, char *grid) * Also, that approach fares badly when you introduce the * additional requirement that incomplete grids should have * errors highlighted only when they can be proved to be errors - * - so that a tree surrounded by BLANK squares should not be - * marked as erroneous (it would be patronising, since the - * overwhelming likelihood is not that the player has forgotten - * to put a tree there but that they have merely not put one - * there _yet), but one surrounded by NONTENTs should. + * - so that trees should not be marked as having too few tents + * if there are enough BLANK squares remaining around them that + * could be turned into the missing tents (to do so would be + * patronising, since the overwhelming likelihood is not that + * the player has forgotten to put a tree there but that they + * have merely not put one there _yet_). However, tents with too + * few trees can be marked immediately, since those are + * definitely player error. * * So I adopt an alternative approach, which is to consider the * bipartite adjacency graph between trees and tents @@ -2392,7 +2395,7 @@ static void int_redraw(drawing *dr, game_drawstate *ds, game_state *oldstate, * currently active drag: we transform dsx,dsy but not anything * else. (This seems to strike a good compromise between having * the error highlights respond instantly to single clicks, but - * not give constant feedback during a right-drag.) + * not giving constant feedback during a right-drag.) */ if (ui && ui->drag_button >= 0) { tmpgrid = snewn(w*h, char); |