Importance of cycling to individual dictionaries?

[mikelove]

If you google WYSIWYG, you get lots of hits for the last 24 hours. The problem for software like Pleco is the increasing complexity. At the same time the manual is of no help, because nobody will ever read it, as you can see when you have a look at some of the less complicated questions in the forum. When Microsoft invented the ribbons, they added a software soon after to help users to find the buttons they knew in the old office, but not in the ribbon office. Pleco's settings are not far away from Microsoft's ribbons as far as complexity is concerned. Therefore WYSIWYG is so important - I do not always want to ask a question in the forum if I want to change one of the settings and do not find it.

Increasing complexity is exactly what this change is meant to help address. Having two buttons is, I would argue, less intuitive for many users; if you've got a search bar and one button next to it and you're not seeing the search results you want, and you think that there are probably better results in there somewhere, the natural thing to do is to tap on that button and see what happens. With two buttons, it's not necessarily obvious to everyone that one button toggles between C-E/E-C and the other one determines which dictionary you're searching in, and when you throw the whole "dictionary groups" concept into the mix it gets even more befuddling; collapsing all of this to a single "try a different type of search" function seems much easier for the typical user to understand.

So if that button will by-and-large encapsulate everything that more advanced users want to do as well, and I think it will (in the two months since we released 2.4, we've had exactly one complaint about the fact that the software now jumps back to the grouped [C] search by default, which would seem to suggest that very few people are attached to single-dictionary searches), then it seems like an all-around design win. It may be less WYSIWYG in the sense that you're no longer dealing with language + dictionary as discrete concepts, but it's more intuitive in that it draws all of the "different type of search" options together in one button.

 

[LantauMan]

Just want to reiterate, Mike, that I'm not attached to single-dictionary searches. I'm attached to SPEED of use. I agree with the above poster that the settings in Pleco are a confusing jungle. But the actual UI, I find to be the epitome of ergonomic excellence.

I can see the attraction of the Facebook app UI for its graphic simplicity. But in Facebook there's no need to do things in a hurry. Like another poster above (who didn't object to the 中英 button, but that he just wouldn't miss it), I also frequently use Pleco to look up Hanzi. But the thing that has made me praise Pleco to everyone I know interested in languages is that it has made the border between Chinese and English CONVERSATION nearly transparent. I can hear an unknown word (for which I obviously don't know the Hanzi), use Swype to rapidly scribble it in, and see the combined results screen, which is always my first choice. If I need more, such as examples, which I often wish to see, it's one tap to go to the individual dictionary at the top of my list, then on to others. If, however, the results are the wrong language (for a pinyin such as "song"), it's a quick single tap on the left to switch languages. It's amazing how quickly I can communicate in this manner.

Combining the buttons so that various tap-combinations are needed, or using a clean menu button like Facebook's, which I assume (I've never used Facebook on my phone), brings up a drop-down list of options, may be graphically simpler, but less ergonomic, in my opinion. How can a single button both toggle languages and scroll through dictionaries without additional gestures? In a hurry, how often will those go wrong?

The beauty of Pleco, to me, is the incredible depth of thought you've put into its usability. Too many app designers sacrifice ergonomics for trendy design. Again, it's still hard to imagine exactly what you're proposing, and I trust your brilliant instincts, but from the sounds of it, your proposal, if it adds a single extra step to toggling language results or scrolling through dictionaries, will add an admittedly tiny extra interval, but those infinitesimal delays add up in rapid conversation. As long as you keep this in mind, I trust that whatever you decide wil be for the best of all users.

Over and out.

 

[mikelove]

Thanks for your continuing feedback on this - having people disagree with me is actually kind of the whole point of this forum

Combining the buttons so that various tap-combinations are needed, or using a clean menu button like Facebook's, which I assume (I've never used Facebook on my phone), brings up a drop-down list of options, may be graphically simpler, but less ergonomic, in my opinion. How can a single button both toggle languages and scroll through dictionaries without additional gestures? In a hurry, how often will those go wrong?

The idea basically is that that button only really needs to have three states anyway - Chinese, regular English and full-text English. So all you're really thinking about is what state that button is in ([C], [E], or [EC]) and what state it needs to be in. If you don't care about English full-text searches you can collapse it to just [E] and [C] and have more-or-less exactly the same function as the 中英 button; if you do care about full-text searches, now instead of having to a) tap 中英 to switch to English and then b) tap on that button to get to the correct state (regular or full-text English search), with b) possibly requiring half a dozen repeated taps to cycle through all of your various dictionaries before you get back to full-text English, now you can get what you want with just one or two taps on that single button.

The only thing that requires an additional gesture is getting search results from a single dictionary; if that particular function is relatively unimportant, as I think it is, then we're making the UI simpler and more intuitive without really adding any work for existing users. Hence my original post title / question about cycling to individual dictionaries: as long as nobody's particularly attached to that function, I don't think too many people will be unhappy with the change, particularly not if we keep the option to bring back the old buttons for people who are affected by it.

 

[HW60]

With two buttons, it's not necessarily obvious to everyone that one button toggles between C-E/E-C and the other one determines which dictionary you're searching in

As the actual version of Pleco changes the 英中 button automatically, this button has become a mere display function and you are probably right to cancel it, though the user now only by looking at his own input and by looking at the results finds out, which of the four directions (E/F/G-C, C-E/F/G, full-text E/F/G and full-text C; F=French, G=German) he is using. This selection can be compared with the input in Google Chrome. Actually the alternative to cancelling the button would be to add 2 more displays #英 and #中 and to move this display to a less dominant position.

But does that necessarily mean to drop cycling through dictionaries? I set all 3 of my dictionary groups to Full-text search only and Skip with Button and see the dictionary groups only in full-text-search or when I select them with tap and hold.

 

[mikelove]

As the actual version of Pleco changes the 英中 button automatically, this button has become a mere display function and you are probably right to cancel it, though the user now only by looking at his own input and by looking at the results finds out, which of the four directions (E/F/G-C, C-E/F/G, full-text E/F/G and full-text C; F=French, G=German) he is using. This selection can be compared with the input in Google Chrome. Actually the alternative to cancelling the button would be to add 2 more displays #英 and #中 and to move this display to a less dominant position.

But does that necessarily mean to drop cycling through dictionaries? I set all 3 of my dictionary groups to Full-text search only and Skip with Button and see the dictionary groups only in full-text-search or when I select them with tap and hold.

You can keep individual dictionaries enabled for cycling under what I'm proposing, but actually in your case I'd recommend disabling the built-in groups and then creating a custom dictionary group for each of your language pairs - that way you're basically cycling through your four types of searches. (so actually it seems like for you this is a considerably more sensible option than the not-particularly-useful 英中)

 

[Tezuk]

I never use that button. No problem to me if it disappears. How will the search function know that we are searching in Cantonese once the update comes through? There are some overlaps between pinyin and jyutping that might cause problems.

 

[mikelove]

I never use that button. No problem to me if it disappears. How will the search function know that we are searching in Cantonese once the update comes through? There are some overlaps between pinyin and jyutping that might cause problems.

We'd actually been planning to do that through a dictionary group anyway, though in its first iteration we may experiment with just searching both and showing all results regardless. (we don't expect there to be much overlap in multi-character words at least, but the best way to see how that works in practice is to try it)


A related idea: we're also thinking about merging normal and full-text E-C searches by:

a) Showing a list of all words / phrases (either in an E-C dictionary or from a full-text index) that match your search query, and
b) Providing a second tab on the definition page for English words that gives you merged full-text results for this word from C-E dictionaries.

With words that only appear in the full-text index, you'd be taken directly to that list tab when you tapped on a result, and instead of showing an E-C definition preview in the search results it would just list the first few words (with Pinyin) from the full-text results. So basically we take away the need for a separate fulltext search type altogether by letting you deal with an English word through a single definition screen just as with a Chinese word now.

Sound reasonable? This would basically drop us to just a single [E]/[C] toggle button, so for most users instead of eliminating the 英中 button we'll have eliminated the dictionary switch one, but retaining the ability to have multiple groups and to search individual dictionaries for those who wish to enable it. It would most definitely be optional, though.

 

[LantauMan]

Sound reasonable? This would basically drop us to just a single [E]/[C] toggle button, so for most users instead of eliminating the 英中 button we'll have eliminated the dictionary switch one, but retaining the ability to have multiple groups and to search individual dictionaries for those who wish to enable it. It would most definitely be optional, though.

I must say that I'm still a bit confused by your descriptions. Also, please define what you mean by "full text search". I'm not exactly clear what you mean by that. All I care about is the ability to retain the current behavior of cycling through the dictionaries I have installed with a single tap each time.

Do I also take it, from a comment above, that you're planning to add Cantonese jyutpin search? Interesting. Now I'll have to learn a whole new system. I learned Cantonese using the Sidney Lau romanization as utilized by the Hong Kong government and many publications, which makes more sense than jyutpin, I think. But that's progress...

 

[mikelove]

I must say that I'm still a bit confused by your descriptions. Also, please define what you mean by "full text search". I'm not exactly clear what you mean by that. All I care about is the ability to retain the current behavior of cycling through the dictionaries I have installed with a single tap each time.

Yeah, it may be better to just implement this stuff and tell people how to go about enabling it... anyway, "full-text search" is the search type where you type in an English word and it gives you a list of Chinese-English entries that contain it. Right now, you have to switch between regular English-Chinese search (where you see a list of English-Chinese dictionary entries, i.e., English headwords with Chinese definitions) and full-text; the idea is that we'd combine those too, so now you just see a list of all unique English words, regardless of whether they come from E-C dictionaries or full-text searches or both, and you tap on one of those to bring up both its E-C definitions and the C-E entries that contain it.

Do I also take it, from a comment above, that you're planning to add Cantonese jyutpin search? Interesting. Now I'll have to learn a whole new system. I learned Cantonese using the Sidney Lau romanization as utilized by the Hong Kong government and many publications, which makes more sense than jyutpin, I think. But that's progress...

Actually, since it seems like we have to support Yale along with Jyutping anyway (too many people using that to ignore it) we'll probably support as many systems as we can - Lau and Guangdong and maybe even HK Government / SL Wong / etc.

 

[scykei]

Sort if I'm late to the discussion.

So this is how I understand it: you're not getting rid of the 中/英, you're getting rid of the cycle through dictionaries part (PLC, TL, ABC, GF, CC). With the implementation of merged results, you only need users to specify it it is supposed to be an English, Chinese, or full-text search. Similar to Googke Chrome's omnibar where it automatically detects your search type if it is a URL or a search term. If in case you entered a URL like link but you actually wanted to search, you can tell it that is not a URL by toggling between it. So that's the use of the E/C/E-C button?

If that's the case then I have no problems with it. However, I'm the type who likes to flip through dictionary entries. Words arranged in order of ABC satisfies my needs and I wonder if this is going to go away? I think you shouldn't get rid of individual dictionary listings and allow the user to scroll through if they so choose. I mean, we bought the dictionary so we should be able to see a list of all it's definitions if we want to, right?

But I'm not sure if I actually got what you said correctly. It's sort of hard to visualise. Maybe you can post a video.

Oh, and I don't think I have too much problems with the sliding GUI of Facebook. It might be weird at first but well get used to it. I trust that Mike will figure out how to make the best of this change. I am a little worried because I have seen some poor implementations of this where the sliding bar only gets in the way. Some even overdo it, like Alien Blue or Jasmine YouTube player.

 

[mikelove]

So this is how I understand it: you're not getting rid of the 中/英, you're getting rid of the cycle through dictionaries part (PLC, TL, ABC, GF, CC). With the implementation of merged results, you only need users to specify it it is supposed to be an English, Chinese, or full-text search. Similar to Googke Chrome's omnibar where it automatically detects your search type if it is a URL or a search term. If in case you entered a URL like link but you actually wanted to search, you can tell it that is not a URL by toggling between it. So that's the use of the E/C/E-C button?

Basically yes - we're 2/3 of the way there already in Android, and since nobody seems to have had a problem with that (as I said, virtually zero complaints about the fact that the system is now defaulting to / pushing you towards merged multi-dictionary searches) we feel emboldened to not only do the same on iOS (where if we screw up we can't simply turn around and release an update a day later to reverse the change) but to make it more integral / harder to avoid.

Oh, and I don't think I have too much problems with the sliding GUI of Facebook. It might be weird at first but well get used to it. I trust that Mike will figure out how to make the best of this change. I am a little worried because I have seen some poor implementations of this where the sliding bar only gets in the way. Some even overdo it, like Alien Blue or Jasmine YouTube player.

Well the basic problem is that we've got too many tabs - the alternative to a Facebook UI would be some sort of a popup multi-row icon grid, but that has its own problems. (the big debate is between those two options, basically)

EDIT: sorry, forgot to address this one:

If that's the case then I have no problems with it. However, I'm the type who likes to flip through dictionary entries. Words arranged in order of ABC satisfies my needs and I wonder if this is going to go away? I think you shouldn't get rid of individual dictionary listings and allow the user to scroll through if they so choose. I mean, we bought the dictionary so we should be able to see a list of all it's definitions if we want to, right?

We already have a replacement for that in the form of our new "Browse Dictionary" screen - the old "list mode" business was kind of confusing so we think that'll be a better bet for most people, plus it's a better home for some other new functions like reading through / batch deletion of user dictionary entries.

 

[Wan]

Basically yes - we're 2/3 of the way there already in Android, and since nobody seems to have had a problem with that (as I said, virtually zero complaints about the fact that the system is now defaulting to / pushing you towards merged multi-dictionary searches) we feel emboldened to not only do the same on iOS (where if we screw up we can't simply turn around and release an update a day later to reverse the change) but to make it more integral / harder to avoid.

Well I did like it better the way it was before the change to auto-merge. I used wildcards according to my needs and cycled through dictionaries or chose them manually, because some dictionaries just aren’t all that accurate or don’t provide information that I’d be willing to rely on (or just aren’t that helpful). Therefore, I really love the dictionary group feature and also exclude some dictionaries from regular search (eg. 古代 which I only need in certain cases, or CF if I forget a French word).

By the way, I find the settings screen to be a little confusing, with countless sub-windows and the important options being spread over different screens. I am very sure I could adjust everything to work exactly the way I want it to, but with new versions every couple of weeks (that’s great, don’t get me wrong. I’d never complain about a healthy update frequency) I’m just a year or so behind at how all the options work. I still think Windows Mobile 6 and loved the very adjustable design and rather tech-savvy approach.

I wouldn’t want you to revert it to the old version though, because a software is never “finished” and if you don’t adapt to modern taste / functionality / design / OS, you can’t sell it, I guess It’s a great product though, and the only reason for me to have a phone with a coloured screen and RAM specs on the packaging…

 

[mikelove]

Well I did like it better the way it was before the change to auto-merge. I used wildcards according to my needs and cycled through dictionaries or chose them manually, because some dictionaries just aren’t all that accurate or don’t provide information that I’d be willing to rely on (or just aren’t that helpful). Therefore, I really love the dictionary group feature and also exclude some dictionaries from regular search (eg. 古代 which I only need in certain cases, or CF if I forget a French word).

Makes sense - that's pretty much what the groups are for, and there's no reason you wouldn't still be able to use them in this system; actually I'd probably recommend that you just keep your primary (less-important-dictionaries-excluded) group accessible and "skip with button" the other ones, so you can get to them with a tap-hold but don't have them coming up constantly.

By the way, I find the settings screen to be a little confusing, with countless sub-windows and the important options being spread over different screens. I am very sure I could adjust everything to work exactly the way I want it to, but with new versions every couple of weeks (that’s great, don’t get me wrong. I’d never complain about a healthy update frequency) I’m just a year or so behind at how all the options work.

Actually the settings haven't changed much in the last year - we try not to rearrange them too often, we'll likely to do again only once we're satisfied with the state of our search system and can therefore start eliminating legacy options we no longer need.

I still think Windows Mobile 6 and loved the very adjustable design and rather tech-savvy approach.

Really? What specifically do you miss about the WM version?

 

[sandwich]

A related idea: we're also thinking about merging normal and full-text E-C searches by:

a) Showing a list of all words / phrases (either in an E-C dictionary or from a full-text index) that match your search query, and
b) Providing a second tab on the definition page for English words that gives you merged full-text results for this word from C-E dictionaries.

E-C sounds like a monster that doesn't belong, so I'd agree that getting it off the dictionary group switcher is a good idea. But how does it effect searchers for Pleco and CC?

Fingers crossed that full-text search will one day, just go away.

 

[mikelove]

E-C sounds like a monster that doesn't belong, so I'd agree that getting it off the dictionary group switcher is a good idea. But how does it effect searchers for Pleco and CC?

Well that's not quite clear - either we'd keep the old behavior or we'd simply display a list of all of the English words in the full-text index that matched your query.

 

[Wan]

Really? What specifically do you miss about the WM version?

Sorry to have bothered you with this: Now that I think of it, it’s just too long ago to clearly remember the features of the WM version. My statement was based on general impression. And I’m perfectly happy with the Android version as is, and I’m also looking forward to the upcoming “facelift” and structural redesign. What you suggested in other threads sounds very promising.