2.0.4 Beta 4 Bug Report / Feedback Thread
- Thread starter mikelove
- Start date
koxinga said:
mikelove said:
from 2.0.3 Bug Report Thread
讶异 link issue still not working (at least for me)
HW60
状元
Beta4 has "Added an option to randomize the order of cards appearing before the not-yet-due cutoff in Repetition-Spaced sessions". Where can I change that option? Maybe it would be a good idea to add some information to the instruction manual even if this is still a beta (or keep 2 instruction manuals during testing of a new beta), e.g. I did not find the definition of Flashcards / Advanced Settings / Display / Top Fields: Priority and Overdue.
I have some minor problems with HWR (on a HTC Touch HD): With HWR auto-rec after: Instant, the lines I draw are sometimes incomplete (I have not found out under which conditions this really happens; in flashcard mode it seems to be more often than in dictionary mode). I therefore tried "after 1/4 second" or "after 1/2 second", and the lines are always complete, but then if I am not quick enough writing the whole character, the character is recognized (as far as I wrote it) and the input field is erased. So I have to write the whole character again. With "instant" the character is recognized during writing (which may cause the Touch HD to be busy with HWR and therefore not to be able to catch part of the lines?), but the input field is only erased if I choose a character from the character box. Is there a reason for this diverging behaviour? I would prefer erasing the input field only when choosing a character from the character box (and allowing to continue writing the character and counting the time elapsed again).
If I test flashcards, those with 5 or more characters will not come up because they are not supported under certain conditions. It would be helpful to receive a warning by Pleco and to be able to select exactly these not supported flashcards for a Self Scored session.
The new information for #reviewed is very helpful!
I have some minor problems with HWR (on a HTC Touch HD): With HWR auto-rec after: Instant, the lines I draw are sometimes incomplete (I have not found out under which conditions this really happens; in flashcard mode it seems to be more often than in dictionary mode). I therefore tried "after 1/4 second" or "after 1/2 second", and the lines are always complete, but then if I am not quick enough writing the whole character, the character is recognized (as far as I wrote it) and the input field is erased. So I have to write the whole character again. With "instant" the character is recognized during writing (which may cause the Touch HD to be busy with HWR and therefore not to be able to catch part of the lines?), but the input field is only erased if I choose a character from the character box. Is there a reason for this diverging behaviour? I would prefer erasing the input field only when choosing a character from the character box (and allowing to continue writing the character and counting the time elapsed again).
If I test flashcards, those with 5 or more characters will not come up because they are not supported under certain conditions. It would be helpful to receive a warning by Pleco and to be able to select exactly these not supported flashcards for a Self Scored session.
The new information for #reviewed is very helpful!
mfcb - the best fix for that link issue is to go into Preferences / Stylus Actions and change "Link tap action" to "Popup Defn" - that will bring up the correct entry in a pop-up window. Unfortunately, on further investigation of this bug we realized that the only straightforward way to fix it with that setting set to "Copy to Input" would be to make all Pinyin searches case-sensitive, which is kind of a big step - we can eventually work out a way to flag an entry put in the Input Field as the result of a link tap differently than an ordinary user-input text string, but that's beyond what we can do in this update.
HW60 - it's in the Cards tab of Advanced Settings - bottom of the screen on Palm, and at the right of the line just below "Repetition-spaced" on WM. Though we could use some better documentation on this, that's true...
"Instant" doesn't erase the input box until you pick a character because otherwise it has no way of knowing when you're done writing; the time delay ones do erase it because the idea there is that you can finish writing the character, have it recognized, then immediately move on to the next character. (though this is mainly useful if you're also using "Auto-enter best match") If they left the character in the box, assuming "Instant" was working correctly there'd be no reason to ever use a delay, since you'd need to do that extra screen tap either way.
However, it sounds in your case like "Instant" isn't working correctly - do you find that the interruptions coincide with a new set of characters appearing in the box? It might be that the recognizer is somehow getting triggered even though you haven't finished the stroke yet, though it's also possible that there might be an issue with your screen / stylus that's causing it to skip or stutter (which isn't an issue with the timer because the skip is over / you've already continued writing long before the timer goes off).
We actually already have a "filter by length" option on iPhone, but we don't know where we can fit it on WM without heavily redesigning the flashcard settings screens - there's simply no more room on the Card Filter screen to accommodate it, at least not without making part of that screen cut off for the still-pretty-significant number of people using Pleco on square-screen devices. If you're doing repetition-spaced study, though, this shouldn't actually even be necessary - do one test run in free-answer mode to test yourself on all of those cards, then do one in self-scored and the only cards that are still due will be ones that weren't covered by the original test. A warning for those longer cards would be good too, though.
HW60 - it's in the Cards tab of Advanced Settings - bottom of the screen on Palm, and at the right of the line just below "Repetition-spaced" on WM. Though we could use some better documentation on this, that's true...
"Instant" doesn't erase the input box until you pick a character because otherwise it has no way of knowing when you're done writing; the time delay ones do erase it because the idea there is that you can finish writing the character, have it recognized, then immediately move on to the next character. (though this is mainly useful if you're also using "Auto-enter best match") If they left the character in the box, assuming "Instant" was working correctly there'd be no reason to ever use a delay, since you'd need to do that extra screen tap either way.
However, it sounds in your case like "Instant" isn't working correctly - do you find that the interruptions coincide with a new set of characters appearing in the box? It might be that the recognizer is somehow getting triggered even though you haven't finished the stroke yet, though it's also possible that there might be an issue with your screen / stylus that's causing it to skip or stutter (which isn't an issue with the timer because the skip is over / you've already continued writing long before the timer goes off).
We actually already have a "filter by length" option on iPhone, but we don't know where we can fit it on WM without heavily redesigning the flashcard settings screens - there's simply no more room on the Card Filter screen to accommodate it, at least not without making part of that screen cut off for the still-pretty-significant number of people using Pleco on square-screen devices. If you're doing repetition-spaced study, though, this shouldn't actually even be necessary - do one test run in free-answer mode to test yourself on all of those cards, then do one in self-scored and the only cards that are still due will be ones that weren't covered by the original test. A warning for those longer cards would be good too, though.
HW60
状元
"Auto enter best match" is somewhat dangerous in flashcard mode - if I know the answer, but HWR and I disagree, I have to reject the wrong answer; not bad for dictionary mode, however.
Maybe my instant HWR did not always work because I had too many programs running in the background - I keep watching.
Switching from frequency adjust to repetition spaced solves the 5+ character problem, but is a problem itself: I do not understand the rules for what is going on making a new date for reviewing a card. So far I did not care for difficulty and score buttons. Now I wanted to solve the problem by looking at the result, but the overdue value is only visible during a flashcard session, and even a review only flashcard session (and "if rep-spaced is early: don't change") changes the dates. I think it would be a good idea to include overdue in Manage Flashcards display and allow looking at these values without changing them to get a feeling for the procedure. Another good idea might be some examples (in the flashcard tutorial?) to show the way the most typical scoring systems are working - the time to learn the power of the system should not exceed the time to learn the flashcards ...
Maybe my instant HWR did not always work because I had too many programs running in the background - I keep watching.
Switching from frequency adjust to repetition spaced solves the 5+ character problem, but is a problem itself: I do not understand the rules for what is going on making a new date for reviewing a card. So far I did not care for difficulty and score buttons. Now I wanted to solve the problem by looking at the result, but the overdue value is only visible during a flashcard session, and even a review only flashcard session (and "if rep-spaced is early: don't change") changes the dates. I think it would be a good idea to include overdue in Manage Flashcards display and allow looking at these values without changing them to get a feeling for the procedure. Another good idea might be some examples (in the flashcard tutorial?) to show the way the most typical scoring systems are working - the time to learn the power of the system should not exceed the time to learn the flashcards ...
We're thinking about adding a couple of new default / built-in study profiles to help expose people to different test modes / scoring systems / etc, but nothing definite on that yet.
One other alternative you could try would be to keep frequency-adjusted, but after doing your Free-answer session (still using "All" cards, right?), add a Card Filter for cards "reviewed" "not within" "1 days" (or "1 hours" or however long it's been since you started the session) - that should likewise pick up all of the > 4 character cards that were skipped in the original session.
"If rep-spaced is early: don't change" doesn't change the score, but it still updates the "last reviewed" date - otherwise your repetition interval isn't increasing, which is the entire point of repetition-spaced study in the first place. If you really want to stop the system from recording review dates you can do so by setting "Scoring type" to "None" in Advanced Settings / Scoring, but I stand by our decision to update those dates in "review only."
One other alternative you could try would be to keep frequency-adjusted, but after doing your Free-answer session (still using "All" cards, right?), add a Card Filter for cards "reviewed" "not within" "1 days" (or "1 hours" or however long it's been since you started the session) - that should likewise pick up all of the > 4 character cards that were skipped in the original session.
"If rep-spaced is early: don't change" doesn't change the score, but it still updates the "last reviewed" date - otherwise your repetition interval isn't increasing, which is the entire point of repetition-spaced study in the first place. If you really want to stop the system from recording review dates you can do so by setting "Scoring type" to "None" in Advanced Settings / Scoring, but I stand by our decision to update those dates in "review only."
HW60
状元
I found a bug which I cannot reproduce, and which happens - if at all - only at the beginnig of a flashcard session with the first word: being asked for a chinese word with more than one syllable (show audio, ask for headword), I answer (free answer mode, repetition spaced), and when I tap on the right-arrow-button to reveal the solution, I see the flashcard, but less characters in the green or red marked character boxes. Actually I was audio-asked for 听见. Now I see the flashcard (chinese word, pinyin and definition), but only one character in the character box (听), marked green because I had given the right answer for the first character. The second character box for 见 is missing (I had given a wrong answer). As far as I remember this only happened with Beta4 (maybe also Beta3), and not more than 3 or 4 times alltogether. The first time it was a 4 character word, and only 2 character boxes were shown. This time it was the third session today, I had made a pause without stopping Pleco after the second session.
I exited the actual session (tapping ok in the top right corner), saved the session, looked at "advanced settings", started the flashcard session again, resumed the previously saved session, and found the same word 听见, now with 2 character boxes below - the box for 见 was red because I had given the wrong answer.
I exited the actual session (tapping ok in the top right corner), saved the session, looked at "advanced settings", started the flashcard session again, resumed the previously saved session, and found the same word 听见, now with 2 character boxes below - the box for 见 was red because I had given the wrong answer.
ups, almost i forgot: when you turn off the Pinyin by going into Preferences / Display / Mandarin Pronunciation and setting the rendering mode for it to "Skip", then you also lose the "links" that are present in some of the cards (like the 讶异 mentioned above). i think, this is not a feature...
Right, that would be a problem - probably best if we ignored the skip setting in those, in spite of the fact that they might give away clues to pronunciation. So maybe we should just wait until we've (finally) gotten around to modifying the ABC database to provide the option of character- instead of Pinyin-based links.
Dear Mike,
Not a big deal, but I still don't think I'm getting completely random flashcards with beta 4. I did about 40 cards. Then my Palm reset itself as it sometimes does. I restarted flashcards, and the first cards up were all the cards that I had just missed, then the card that was up when the Palm reset itself. I still had about 400 cards to go, so I don't think I should have seen the missed cards first.
Rep spacing, self scored, prioritized by random, random sort box checked, 2.0.4b4, Palm Treo 680.
John
Not a big deal, but I still don't think I'm getting completely random flashcards with beta 4. I did about 40 cards. Then my Palm reset itself as it sometimes does. I restarted flashcards, and the first cards up were all the cards that I had just missed, then the card that was up when the Palm reset itself. I still had about 400 cards to go, so I don't think I should have seen the missed cards first.
Rep spacing, self scored, prioritized by random, random sort box checked, 2.0.4b4, Palm Treo 680.
John
You can email it to me at mikelove@pleco.com.