3.2.50 Beta (Android Oreo + enhanced TTS)

[Bvo]

1) dict order issue related to the main screen (not flashcards, only TCE and PLC interchanged in FC)
2)

With the option to view them as dictionary entries rather than flashcards?

yes. If you browse flashcards, not run a test, you see mixed content issue.
And 老娘 example is about content, not dict order.

Just want to put an emphasis on dict interfluence in flashcards

Just try to add 老娘2 and (a) you'll see 老娘5 (not 2!) in flashcards (b) the card will have mixed content from both entries.

 

[Bvo]

Another content mixed example: 教学1 and 教学4 (both non-neutral).
Try to add 教学1 to FC and you'll have
(a) mixed content from both 1 and 4 entries
(b) in the card there is 2(!) entries for PLC (PLC, TCE, PLC, etc.).
Moreover, in case if you add 4 (not 1) the content will be mixed too but there'll be only one PLC entry. So it's even impossible to predict which word variant it's better to add as content will'be the same anyway but mess inside the card will be different...
If someone could remember a multi-char word with more than two tone variants - let's check it

Hope v4 is better.

 

[mikelove]

Dict order on the main screen was what I was talking about - we group the most similar words first. The mixed content for 教学 is also intentional - we use a less restrictive search when populating those flashcard entry displays because you're not looking at a list of dictionary search results + if we don't show the result there you can't simply scroll through the other results to find it as you can in the dictionary search screen.

Honestly there are a lot of places here where we deliberately did something in a less strict way because for the average user who isn't going in and scrutinizing everything it delivers a better experience; I'm sorry it's not as satisfactory to you, but pretty much every one of these decisions was made for a reason (some of them actually required a considerable amount of programming time), and candidly, Pleco has been working this way for 4 years now (ever since our 3.0 redesign) and AFAIK you're the first person ever to complain about most of this stuff.

In particular I'm committed very strongly to the idea that it's better for us to err on the side of showing you too much data than too little; I'd rather show people extra results for a slightly different word which they can filter out than exclude results for the same word and thus miss an important definition that's only in one dictionary that happens to use different tones for it. It's easy to ignore irrelevant content, it's a lot harder to go digging for missing content, particularly if you don't even know there's anything missing. (I'd also add that Google - which, like it or not, is the world's leader at searching for stuff - adopts a similarly loose approach to Chinese search, including common and not-always-quite-identical-in-meaning variants without even asking if you want them)

 

[Bvo]

Mike, I see the reasons and will not disturb anymore for things that only I worry about, sorry for taking up of you time
Every software is imperfect (as our world is) and it's impossible to satisfy everyone.
Just realized that Pleco is not a real learning tool, but primarily the biggest Chinese dict + some aims for beginners (though OCR module for reading scanned books I like much).

Just one more thought - I think that creating mixed content for a flashcard is wrong by design, by the same logic you have to combine all dict entries with a word variants. Different variants put different emphasis on a word meaning (sometimes they're absolutely different words, just with the same writing) and at main search list you've got that point (may be unintentionally) but at flashcard list you think it should work differently and say they shouldn't be separated...

 

[Bvo]

In particular I'm committed very strongly to the idea that it's better for us to err on the side of showing you too much data than too little; I'd rather show people extra results for a slightly different word which they can filter out than exclude results for the same word and thus miss an important definition

By the way, the autoswitching issue above ("qiri - ciri") shows hat the idea is far from accomplishment, but at that time you sad to me that losing definitions is OK (just because it works in this way).
Obviously it's another example of wrong design.
I'm not trying to teach "how to do", just pointing out that IMHO you have a different logic for different aspects of the same app.
May be many of them here because they come from the app history and old users like it in this way, but I think for new users it requires too much effort to understand how to deal with the app.

Though as not much users complain, may be I'm too much picky
But I'm just trying to help make the app better.

 

[mikelove]

Flashcards definitely need an overhaul, yes, which is why we're working so hard on one - current system dates back to 2007 and the many many technical constraints of mobile devices back then.

Flashcards *on tests* do show you just the one definition; with the definition view in Organize we're treating this more as a list of words than as something you're studying / testing on (particularly if, say, you import a vocabulary list from somewhere and then browse through it), hence the greater number of definitions.

The qiri/ciri bit was about parsing an ambiguous search string in different ways; all of the definitions were there if you flipped between search types. "ci ri" didn't get you more definitions due to the lack of ambiguity, it got them because putting that space in made it obvious that you were not doing an English search and so Pleco dutifully switched to Pinyin. This is similar logic to why entries in search results are grouped by tone even though entries in the flashcard definition screen aren't; in search results you're right next to the other definitions and can easily flip through them / would naturally flip through them, whereas in a flashcard definition screen if you scroll up/down you're going to totally unrelated words. In the cases where we show you less data we're doing it in a context where it's very easy / obvious how to get to the remaining data, and where it's something one would ordinarily do; with missing entries for tone mismatches, if the entry isn't there it isn't there and you're neither going to realize it's missing nor have an easy way to find it.

 

[mikelove]

Sorry, just to add: I do appreciate all your feedback - it’s always good to be challenged on stuff like this! - and we can certainly consider making some of these behaviors optional in the future.

 

[Bvo]

Sure, it's always better to give a choice if there is no "always right logic".
It seems you're a follower of "smart logic" and I prefer "straight/intuitive logic".
If someone don't like/understand an app default behaviour and can change it - it's always the best of all and will satisfy much more users than one-choice only.
And even more - you can make 2 (or more) sets of options and ask user preferences on the very first app start.