Feature Suggestions

[radioman]

Without completely deciphering the thread, I wanted insight to my specific approach and if Pleco currently handles this.

Right now I am developing large amounts of specifically categorized flashcards on the desktop. But as I use the cards on then handheld, sometimes I run into errors on individual cards. I make note of it and I basically modify them on the the desktop, and import the cards again. The category file might have 200 different custom cards (phrases or sentences) that were previously imported into Pleco from the desktop. If, for example, 15 different cards are changed (pinyin, definition, Hanzi, or category) in any way, then what I was hoping that only the modified cards would be overwritten, with the rest ignored.

*** Is this how the import is handled today? If so, how would that best be configured?

I thought I had problems with this before so I currently just delete the category on Pleco and then re-import the modified category file. However, if I can in fact configure Pleco to import and ignore the exact custom duplicates, then I could preserve my scoring information.

This obviously does not address backing up of the master flashcard/score file, but the approach it easy to modify any errors and just re-import without having to sort out which ones changed.

As for anything that brings the full power of editing the cards onto the desktop, from syncing to full-scale moving of databases, I think that would be extremely useful.

Yes, that's what I meant. Also, getting just the export/import (copy to desktop and back) working early will open things up for others to develop solutions, even without a desktop version of Pleco. So, I would advocate doing it in steps.

The best way to do that might actually just be to copy the whole database to the desktop, make any necessary changes, and then copy it back (or for small changes just submit a list of SQL commands to apply on the handheld), something we're thinking about even for our own sync system - the XML is really more of a way to back up / share data than to synchronize it, exporting to / importing from that imposes a pretty big performance penalty on a handheld at least.

 

[mikelove]

radioman - the way things work now, if the headword / Pinyin changes in those modified cards, they'll be imported as brand new cards. If the headword / Pinyin remains the same and only the definition changes, they'll be ignored / skipped, unless you have the Duplicate Entries setting in the Import screen set to Allow (in which case all of the cards in that modified file will be imported as brand new cards) or Replace (in which case *all* of the cards in the list, even the ones that haven't changed, will be deleted (including their score info) and replaced by brand new ones).

We've thought about adding an "Update" option for duplicate card handling, which would update definitions with new ones in an import file, but that would only work if the headword / Pinyin hadn't changed - tracking card changes even when the headword / Pinyin are updated would pretty much require a desktop sync system.

 

[jiacheng]

Traditional != Simplified Filter

I would like to suggest that a filter criteria be created that you could use to filter out the words whose traditional form differs from the simplified form. That way, you could create categories or flashcard session filters that would allow you to only on those words.. say if you are like me and need to beef up on your ability to recognize traditional characters, I could create a test just for that, and quiz me on that.

 

[stephanhodges]

Traditional != Simplified Filter

I would like to suggest that a filter criteria be created that you could use to filter out the words whose traditional form differs from the simplified form. That way, you could create categories or flashcard session filters that would allow you to only on those words.. say if you are like me and need to beef up on your ability to recognize traditional characters, I could create a test just for that, and quiz me on that.

That would be a great idea. Would you also want to be able to include both in a single session? But, then you'd need a way to know which was being asked for when doing E->C, etc.

 

[renovator]

In flashcard statistics an option that would give historical Score ranges. Have a check box to chose last 4 weeks or last 4 months. This should not be too hard to implement and would be of great value to give an idea if the rate at which I add new flash cards to learn, is OK, too fast or too slow. If cards are being added too fast, then the higher scores would drop over these given periods, if cards are being added too slowly all card scores would rise over these periods. By looking at this 4 week or 4 month period, it would then be easy to optimize the number of cards that can be learned per day.

Christmas will be here soon. May as well ask for everything and see what Santa delivers in next update

 

[stephanhodges]

Could you make a slight change to the import file formats to support comment lines (lines that would be ignored by the importer?

This would allow meta information, etc. For example, I could keep the ZDT "long" category names in my original file, and add a nested, but shorter named set for the Pleco stuff, similar to what I just did for Conversational 301 in the flashcard list.

Of course, a while ago, I also asked for the ability to add a long comment, but that would require a database change, so I don't expect you'd do that for 2.01. I suppose you could define the format for that also...

For me specifically, I use almost exclusively txt files. They're much easier to generate from both Wenlin and Excel, the two programs I use for study...

 

[mikelove]

jiacheng - interesting idea, and wouldn't be too hard to implement (in fact it'd be incredibly easy, just one little clause in a SQL query comparing the simplified and traditional parts of the card); the big problem really with any new option now is interface clutter, but at some point we're going to have to start more aggressively differentiating between basic and advanced options (even within the Advanced Settings screens there are really a lot of things that don't deserve that prominent a placement) and once we do that we can certainly look at adding this.

renovator - that actually would be tricky to implement, since for space-saving reasons we don't really store much historical data now; storing the dates / previous scores every time a flashcard changes (which is what we'd need to do to generate this data) wouldn't really be feasible with the limited internal memory of most Palms and most WM handhelds. (might be an option on iPhone, though )

stephanhodges - you can already do that with XML files, actually, since the importer ignores any text that isn't in a tag it recognizes - if you just add a "<comment>" tag (better to give it a less likely name like "<stephanhodgescomment>" so there's no possibility of it conflicting with a new tag in some future version) you can store whatever you want in there and it shouldn't affect the importer at all.

I'm a bit worried about adding support for text comment lines because it might break some existing import files - we went through this a couple of times with the 1.0-to-2.0 converter when people were storing weird characters / sequences / etc in their database which we'd been trying to use for some special purpose, whatever character we choose there are bound to be a couple of people who have flashcards starting with that character. I guess we could use something similarly long / unlikely to come up by accident, like "###***THISISAPLECOCOMMENT," but that could get a bit unwieldy

 

[stephanhodges]

Actually, I suggest using a simple javascript comment style. Also, I pretty much expected XML already had support (in that you would ignore non-relevant tags).

Most urgent for me right now is to add a level to the groupings, because my Classes-Course-Chapter still results in much too large of files, and I'd like to be able to segment lists one more level.

 

[llammamama]

I know there are a lot of people out there (myself included) who do a lot of studying with sentence flashcards, and I'm wondering what types of support could be provided to make creating flashcards using example sentences easier. Currently my work flow looks like this:

1. Full text search (in the Tuttle dictionary, usually) for a word I want to learn in a sentence.
2. highlight/copy the sentence + pronunciation + definition (the Tuttle dictionary is GREAT for this, btw)
3. open flashcards
4. create new flashcard
5. paste everything into the definition box
6. cut the chinese sentence
7. paste into the headword field (cut / paste again if the sentence was longer than two lines)
8. cut the pronunciation
9. paste into the pronunciation field (cut / paste again if the pronunciation was longer than two lines)
10. save the card

This doesn't actually take all that much time, but it does feel as if this could be improved considerably. I think there has been some discussion of this already, but it would save a step if were able to go straight to the 'add new blank card' screen from the main dictionary screen. via an 'add new blank card button', or mabye a new entry in the add card tap-hold context menu.

My dream would be to be able to highlight the sentence+pronunciation+definition (or sentence+definition for the Oxford dictionary) click a "add highlighted to flash" button, and then have it be added to a flashcard with the respective fields parsed from the highlighted text (/clipboard).

Assuming I had properly highlighted a section of chinese-pinyin-english or chinese-english, it seems as though parsing this and assigning it properly to those three fields should be easy. The dictionaries always present the information in the same order, and if the user's highlight didn't begin with Chinese characters then an error could be returned. Furthermore, the main dictionary display mode (simplified or traditional) would automatically determine which headword field to fill in. (If you wanted to be even trickier you might be able to automatically fill in both). I imagine this would save tons of time for anyone like me who is interested in studying sentences, and it doesn't seem like it would be a huge programming project either.

If a feature like this was actually added.. ooh man I would probably break down and cry tears of joy (as opposed to the quiet tears of joy which are already trickling down my cheeks - Pleco 2.0 rocks! )


Also, how did you get the full text search to be so fast?! I love it.

 

[renovator]

renovator - that actually would be tricky to implement, since for space-saving reasons we don't really store much historical data now; storing the dates / previous scores every time a flashcard changes (which is what we'd need to do to generate this data) wouldn't really be feasible with the limited internal memory of most Palms and most WM handhelds. (might be an option on iPhone, though )

To regenerate the data would take time and resources but the data is there every day in the statistics, so could you not just save the Score Ranges table by date and then display rather than re-generate the data. I would think that would take very little space.

 

[jiacheng]

It would be nice if you could increase number of buttons that can be assigned in the flashcard and dictionary interfaces. My PDA has quite a number of buttons, so it would be nice to utilize them all.

Also, I'm not sure how to do this cleanly, but it would be cool if some buttons could have different functions in different contexts of the flashcard interface. Say for example, I want to assign "reveal" to a button and also "correct" to the same button. But the problem comes in that when u press it to reveal, the card will also be marked correct at the same time. I find that using the buttons tends to be more efficient, and this would enable you to do more complex test types without having to use the styus.

 

[mikelove]

stephanhodges - well // is already used for category names (in retrospect a poor choice), so that specific code isn't possible unfortunately. Understood about the groupings, but I'd really rather not make the space for category names on Palm any smaller and there's no way to get a 4th level of category groupings without doing that (to create the extra column for a third bank of arrow indicators).

llammamama - that is pretty convoluted, yeah - sentences are something we're thinking a lot about for 2.1, given that this is an electronic dictionary it seems kind of silly that example sentences are still tethered to specific entries at all - what would really be ideal is a feature that brings up ALL of the example sentences from ANY dictionary that happen to contain a particular word (whether or not that word is actually the original focus of the example sentence). And then of course we'd want to provide a way to automatically create flashcards based on any of these sentences, along with perhaps a slightly more streamlined system for creating your own (say with an integrated Pinyin-to-character input system).

renovator - that might work, though generating those reports can be kind of slow and I'd hate to hang up someone's Palm (or drain their battery) by doing it automatically - better if we can do it as a background task, which won't be possible until everything's a lot more multithreaded (not only a Palm OS hangup but a we-haven't-really-designed-the-software-to-be-that-threadsafe hangup even on multitasking-friendly OSes like WM).

jiacheng - there's probably going to be an option to customize button settings for each profile (or each test type) at some point, so that might help with this, though I guess there are some scenarios where even in a specific session type there might be more than 8 commands you want to routinely access. I'd try out the Screen Left / Center / Right options, though, if you haven't already - just assigning those three will take care of most of the common commands during most session types.

 

[llammamama]

...what would really be ideal is a feature that brings up ALL of the example sentences from ANY dictionary that happen to contain a particular word (whether or not that word is actually the original focus of the example sentence). And then of course we'd want to provide a way to automatically create flashcards based on any of these sentences, along with perhaps a slightly more streamlined system for creating your own (say with an integrated Pinyin-to-character input system).

That sounds even better than what I was hoping for, and might even be considered another 'prime' feature of Pleco if you manage to pull it off. I am definitely willing to pay for that sort of functionality, but then again you run into the trouble of having even more confusing packages/prices. Oh well I guess I'd just buy another dictionary instead


Also, one more usability issue is the unavalibility of the main dictionary window (or popup definitions) while editing flashcards. Often times I want to place a definition or other text from the dictionary permanently into the flashcard field, but at the moment you need to exit the flashcard module completely in order to access the dictionary. Just one more thing to put on your endless list of possible improvements ^_^

 

[mikelove]

Yeah, the no-charge-for-version-upgrades-but-please-buy-our-new-dictionaries model may be getting a bit too unwieldy - if Apple proves flexible about iPhone distribution we're hoping to introduce a vastly-streamlined system for buying / upgrading with that.

If you flip back and forth between the From Dict and Custom tabs in Edit Card, you can pull up dictionary information that way; not quite as nice as the full dictionary interface but it's an easy way to grab a few characters at least.

 

[renovator]

It would be great if the instruction manual could have page numbers as well as sections. When new releases are done, to issue a link we could click on that would allow us to print only the pages that have changed since the last update (version). I find it very useful to have a copy of the entire manual in a 3 ring binder where I can make notes and review sections from time to time. As your system includes more options, this will be more important to reduce the number of emails you get on how to do things in Pleco. I understand you will go to videos at some point, but the printed backup will still be important.

I don't want to have to reprint the entire instruction manual at each update, three hole punch it and then have to throw out the old one. That is just too much of a waste of time and paper. This should not be too hard to implement given the version control tools in the marketplace today for manuals. I know this type of housekeeping can be a pain when time is maybe better spent in development than in documentation but in the long run I would think you would save time and frustrations on both your part and the users by doing this.

 

[mikelove]

Good idea, but before we can do that we need to switch over to some sort of document management package that lets us generate the same manual in both web-based and printable form; it's too much work to keep two different editions of the manual synchronized, which is why at the moment we only offer it web-based (since that's more useful these days, and easier to put on a handheld). This would hopefully also allow other improvements, like separating the Palm / WM / eventual iPhone versions into their own manuals rather than requiring you to read through platform-specific screenshots / options / etc.

 

[renovator]

Is there a way to use the Compact QWERTY keyboard in the flash cards rather than the Pleco keyboard? In going from Pleco 1.0 to 2.0 I find that the number of cards I can do in an hour has been reduced by 40% because of the time it takes me to input with the stylus in 2.0 rather than just write the pinyin in 1.0. I thought over time I would get much faster with the stylus, and although speed increases a little every week, I now know writing will always be quicker than entering letters with a stylus. My guess is that if the Compact QWERTY could be used, I could train my 2 thumbs in a short period of time to be much quicker than the stylus.

 

[mikelove]

Compact QWERTY won't work in flashcards because it takes up too much of the screen - in theory we could get Pleco to work with its character input messages (as we've already done in the main screen in the just-released 2.0.1b1, or at least we think we have), but without totally redesigning the flashcard session interface there's no way to work with the narrow little band of space that's left after you bring up that keyboard.

If you don't like the free-answer keyboard you could always go back to regular self-scored sessions - they still work just as in 1.0, and still include a drawbox, so if you find that more convenient then by all means keep doing that.

 

[renovator]

If you don't like the free-answer keyboard you could always go back to regular self-scored sessions - they still work just as in 1.0, and still include a drawbox, so if you find that more convenient then by all means keep doing that.

Mike, I am really getting quite greedy here. I love the free-answer keyboard because it never allows you to "cheat". The draw box is just too easy from time to time to say close enough on a card you think you know and thought you just made a quick key error. The free-answer keyboard actually also trains you in the fashion that you will be using the pinyin in SMS, etc. I had to pay the pied piper and put extra effort in re-learning some of those cards when I went to 2.0 without the drawbox so I don't want to go back to drawbox now.

I was actually looking at third party apps such as Cliquick which is highly customizable as a possible solution in case there was not enough space with the HTC Compact QWERTY. I would not be surprised if a company such as cliquick which appears to share the same dedication to work with their customers as you do to modify their app to be able to be used with your flash cards. It is an inexpensive app so could be a win-win situation for them and you as well.

 

[mikelove]

Heh, that's exactly why we added free-answer tests Another keyboard certainly might behave better - Cliquick looks like it can scale down to the appropriate size, though some of the work to make it compatible with flashcards would probably have to be done on our end.