Skip to content

Latest commit

 

History

History
116 lines (84 loc) · 3.65 KB

bugsenhancements.org

File metadata and controls

116 lines (84 loc) · 3.65 KB

bugs and enhancements

bugs

“ankiimport -RB” puts (some) cards back in “new” state

somehow, some cards are ending up back in new state. maybe half?

so, turkish-2016 is the deck in question. it has 214 new cards (mac version).

this file shows state before reviewing, synching from mobile.

now, synch from mobile. launch anki. it says 104 new cards. here is the state after that synch, launching (and quitting!) of anki.

now, create a new HTML file, and use ankiimport to import it, but don’t yet launch anki.

bash greg-minshalls-mbp-2: {1203} date; i=turkish-2016; ./genhtml-bothsides.sh -t $i.txt >| $i.html
Sun Mar 13 09:58:05 EET 2016
bash greg-minshalls-mbp-2: {1204} ./import.sh -RB turkish-2016
1 note added, 525 notes updated, 0 notes unchanged.
will delete 1 card from deck turkish-2016
proceed with import? [y/n] y

here is the state there. it differs in the “due” field, but not much else:

sed 'sX^new/[0-9]*:Xnew/:X' after.mobile.review.ankiimport.txt > y
sed 'sX^new/[0-9]*:Xnew/:X' after.mobile.review.txt > z
diff y z | less

now, launch anki. it says 105 cards. here is the output.

sed 'sX^new/[0-9]*:Xnew/:X' after.mobile.review.ankiimport.txt > y
sed 'sX^new/[0-9]*:Xnew/:X' after.mobile.review.ankiimport.anki.txt > z
diff y z | less

no change. hmm. download to mobile. mobile shows 105 cards left.

so, this didn’t repeat the problem. sigh.

14.03.2016 – importing into my normal, it happened again that a number

my sequence was to run ankiimport before synching with ankiweb. it’s possible that what’s happening is that the cards i reviewed today, having been marked updated, are not going to later be updated by synching with ankiweb (when i actually launch Anki).

but, e.g., ---- okur | reader ---- (from 08.03.2016) is marked “new”. it wouldn’t be new.

instrument ~/docs/mine/flashcards/import shellscript to log “^new” before and after import.

well, actually, in yesterday’s example, okur was included as “new”.

./ankicards --user User\ 1 --deck turkish-2016 | grep ^new

so, after reviewing everything on iphone, and synching from iphone, then launching and closing anki, no new cards remain:

run ./import in flashcards. still no new cards

on <2016-03-16 Wed> finished reviewing new cards on iphone, synchronized from iphone, entered a bunch of new cards on mac, did “ankiimport -RB …” on mac, then launched Anki on mac. at end of this, the “new” cards i had read on the iphone were still marked new. maybe that’s what happens? i.e., the -RB load says “this is recently changed”, so then the “new –> old” on the iphone-reviewed cards is ignored?

that seemed to happen again (same schedule of actions) on <2016-03-17 Thu>.

on getting imaret wrong ( right w/have been 10.1 mos), “leeched”.

i had just moved it to this deck.

enhancements

when moving old cards in to new place, mark as “new”?

or maybe an algorithm to “age” unreviewed cards?

problem being, that when cards are moved from an old deck, after answering the first time, assuming the answer is correct, they go back to “8.7 months”, or some such.