iPhone Feature Requests

[scykei]

Whoops, sorry for late reply.

I've just found out about the Alpha handwriting feature. It's not bad but I would like a way to only recognize numbers because I doubt that I will use handwriting for alphabets too much.

Sorry, I'm not quite sure if I understand this... why would you want to write numbers but not letters? What sort of text are you entering with that box? Are you composing some kind of document in the document reader that uses a combination of Chinese characters and numbers?

Well, I figured that since the keyboard is just a tap away to access and it would probably be much quicker to type them up rather than draw them, especially since the letters usually come large quantities at a time to form words, and the lack of an easy space character makes it even more troublesome to use.

Numbers on the other hand require more effort to access from keyboard. I can think of several uses of numbers like in telephone numbers and other general uses. From my experience, most text written vertically will have their numbers written in their Chinese characters while those written horizontally use Arabic numerals. This probably only happens for large numbers though.

And support for the 〇 character would be good too.

Also, can you let it support common punctuation marks like a fullstop and a comma? It would be much easier to just tap the bottom right area of the screen for a fullstop than having needing to tap three times to access the character and back. Maybe the space characters can be supported too by drawing the classic line across.

Certainly possible, but not something we've had much demand for; so I guess this is being used in the document reader, then. Other than a fullstop and a comma, which punctuation marks do you regularly need to access?

If it is possible, hopefully all of the punctuation marks that are already in the punctuation button below the handwriting space.

Some other requests too:

• Can you make the keyboard not pop up every time we write? It seems that the keyboard pops up no matter what and is only hidden behind the handwriting/radical interface. Can't it pop up only when the "Key" button is pressed? I don't know but having the keyboard pop up with the animation play and then suddenly disappear behind the handwriting input feels clunky.
• I would like to request more information in the Character info section such as other types of romanization. This might have been requested before but I'm not entirely sure. Besides the mandarin romanization systems and non-romanization (Zhuyin), I think I'd like to use Yale Romanization just because I've been studying with it for some time. Jyutping is good but I'm not really used to it.
• In addition to that, it would be great if the Cangjie input can show the characters to press rather than just the keyboard button. I'm currently learning how to type in Cangjie (It's hard ) and I think I've gotten some basics for it. For example, 我 = 竹手戈 instead of HQI.
• A history section for every character I visit. When I come from an entry to another entry through the (>) buttons, it doesn't get recorded into the history. Same happens when when browsing words through document reader, etc. I think there should be another section for it to just to keep track of what I've learnt. I think the list will build fast if this is implemented so some sort of limits should be applied too, like to only keep newest 500 or 1,000 in history or something. I don't think I'll need to scroll too far back in history anyway. Just the recent ones.
• Full screen document reader. The menus might appear every time a word is tapped on for definition though.

That's all I can remember for now.

There is also some random crashing and occasionally documents get reset back to the top after going through several layers of dictionary entries. I'm still trying to find a solid way of reproducing these crashes and document position reset. I think it has something to do with some of the settings I've changed. I'll post if I ever find out the exact cause.

 

[mikelove]

Strange. IMHO ePub is a wonderful format for i* devices. PDF is only OK if if designed for that page format. Otherwise you need scroll till the fingers bleed. ePub text 'floats' and will adapt to the screen.

BTW, one can convert any text PDF to ePub with 'Calibre' - and if things screw up a little - edit the ePub with 'Sigil'. Both are open source and free.

True, but you can also convert any PDF to TXT with Calibre - ePub doesn't add support for decoding files that people can't already easily convert to text, while with many PDFs there's really no way to extract text except with OCR. There's also the email attachment problem - PDF and DOC are two very common email attachment formats that people would like to be able to decode / respond to on-the-go.

Well, I figured that since the keyboard is just a tap away to access and it would probably be much quicker to type them up rather than draw them, especially since the letters usually come large quantities at a time to form words, and the lack of an easy space character makes it even more troublesome to use.

Numbers on the other hand require more effort to access from keyboard. I can think of several uses of numbers like in telephone numbers and other general uses. From my experience, most text written vertically will have their numbers written in their Chinese characters while those written horizontally use Arabic numerals. This probably only happens for large numbers though.

And support for the 〇 character would be good too.

That makes sense, I'm just not sure how many people are using our HWR to compose complicated documents. Not sure why HWR doesn't already support 〇 but I don't know if there's a good way to enable it if not - can certainly check on that, though.

If it is possible, hopefully all of the punctuation marks that are already in the punctuation button below the handwriting space.

I don't think the system can support all of them, but we'll see what we can do.

Can you make the keyboard not pop up every time we write? It seems that the keyboard pops up no matter what and is only hidden behind the handwriting/radical interface. Can't it pop up only when the "Key" button is pressed? I don't know but having the keyboard pop up with the animation play and then suddenly disappear behind the handwriting input feels clunky.

That relates to the fact that we're still supporting iOS 3, which didn't officially allow for the use of custom input methods; once we drop that legacy support it should be pretty easy to get rid of this.

I would like to request more information in the Character info section such as other types of romanization. This might have been requested before but I'm not entirely sure. Besides the mandarin romanization systems and non-romanization (Zhuyin), I think I'd like to use Yale Romanization just because I've been studying with it for some time. Jyutping is good but I'm not really used to it.
In addition to that, it would be great if the Cangjie input can show the characters to press rather than just the keyboard button. I'm currently learning how to type in Cangjie (It's hard ) and I think I've gotten some basics for it. For example, 我 = 竹手戈 instead of HQI.

That's actually not a question of adding more information to the database, just of adding those as additional Mandarin rendering options along with Pinyin and Zhuyin. Good point on Cangjie.

A history section for every character I visit. When I come from an entry to another entry through the (>) buttons, it doesn't get recorded into the history. Same happens when when browsing words through document reader, etc. I think there should be another section for it to just to keep track of what I've learnt. I think the list will build fast if this is implemented so some sort of limits should be applied too, like to only keep newest 500 or 1,000 in history or something. I don't think I'll need to scroll too far back in history anyway. Just the recent ones.

We've had quite a few requests for that - in general we recommend creating flashcards in those cases, since it's possible to go through 1000 words in the reader much more quickly than you might think, but it might be feasible to add those entries to history too eventually.

Full screen document reader. The menus might appear every time a word is tapped on for definition though.

Much-requested, but I'm still not entirely sold on whether it's worth the time investment and potential bugs.

There is also some random crashing and occasionally documents get reset back to the top after going through several layers of dictionary entries. I'm still trying to find a solid way of reproducing these crashes and document position reset. I think it has something to do with some of the settings I've changed. I'll post if I ever find out the exact cause.

The document position reset is probably a RAM thing - we'll investigate - but the crashes are very worrisome; can you illuminate those any further? What are you generally doing when they happen?

 

[radioman]

I made mention of this before, but thought I would reiterate my interest on the following.

I am very (VERY) much hoping that there will be a PDF reader function where you can bring in a non-OCRed PDF file, and then as a reading function, on the fly, you can do instant OCR as you flip through the pages.

The on-the-fly PDF reader function would accomplish the following:

1) allow the OCRing of difficult pages that have graphics that would otherwise screw up standard OCR functions, and
2) solve the labor and extra time required to have to do screenshots, load pages, etc. and importing as photos. Just flip/advance the page and continue reading/OCRing.

Maybe the reader could perform the second step behind the scenes. Just something to avoid having to load the document or photo again and again.

That's been on our to-do list for a while, but we actually get surprisingly little feedback about it - far more people want direct tap-lookup support for PDFs (which we'd have to do via OCR) or Word documents (which would require some sort of wacky external library to extract the text contents of a .doc file, though it might be doable if we limited it to .docx).

 

[gabor]

iSpeech integration? (tts and recogniton)
its for free
http://techcrunch.com/2011/08/03/ispeec ... roid-apps/

 

[mikelove]

I made mention of this before, but thought I would reiterate my interest on the following.

I am very (VERY) much hoping that there will be a PDF reader function where you can bring in a non-OCRed PDF file, and then as a reading function, on the fly, you can do instant OCR as you flip through the pages.

The on-the-fly PDF reader function would accomplish the following:

1) allow the OCRing of difficult pages that have graphics that would otherwise screw up standard OCR functions, and
2) solve the labor and extra time required to have to do screenshots, load pages, etc. and importing as photos. Just flip/advance the page and continue reading/OCRing.

Maybe the reader could perform the second step behind the scenes. Just something to avoid having to load the document or photo again and again.

In the works, basically we just need to a) start supporting PDF as another "image file format" and b) add UI controls / saved state / etc for dealing with the fact that PDFs are multi-page. The biggest hassle is actually RAM-related - on an original 256MB iPad memory is already kind of tight and we're not sure if we can get PDF display + OCR working without crashing so often as to be effectively useless.

iSpeech integration? (tts and recogniton)
its for free
http://techcrunch.com/2011/08/03/ispeec ... roid-apps/

TTS we've already licensed an offline library for (which we should soon be releasing) - much better to do this without an internet connection, for the sake of responsiveness, data usage and support on airplanes / out of cell coverage / etc.

Speech recognition we don't seem to be able to find a license-able offline library for, but iSpeech currently only seems to support Mandarin from a predefined command list, not for arbitrary / freeform speech recognition, so it doesn't look like they'd work for our purposes unfortunately.

 

[gabor]

thanks for your answer!

another thought: "next image" button on OCR Screen, which simply loads the next image in the list, maybe also a "previous image" button, ah, and while at it, it would be nice to show the filename in the OCR (if the image is from Pleco's library and not from the Photo library).

Most people hold the iPad at its sides, but all the buttons in Pleco are on the top and bottom of the screen, making some of them almost impossible to reach without putting the iPad down, why not put them optionally on the sides of the screen?

I am collecting dozens of ideas how to make the user dictionary entry screen better. I have already created over 3000 custom entries, so I do have some experience. I kinda got used to it, although quite difficult. Have you ever actually used it yourself? For exampled for entering a four character idiom, where all four have both simplified and traditional versions, and you don't know the pronunciation? I think in such a case I switch at least 2 or 3 times between the unfinished new entry and the dictionary to get it right. If it can be found in other dictionaries, I can at least copy the traditional or simplified characters, or even the pinyin, but if its not the case it can take several minutes to create a single entry.
Anyway, I will just take some time, sit down and finally write down my ideas on how to make it better, if even just a third of it gets realized it would make work tremendously faster. Maybe its obvious by now: editing custom dictionaries and using these entries in the Document Reader is THE reason I bought Pleco, even more, why I bought an iPad in the first place.

 

[rhenish]

ben_gb wrote:I've got a suggestion for an additional way of looking up characters, for situations where OCR is not possible, or if it's being done from memory.

My suggestion is that there is a way for users to input one or more components in a character (probably using handwriting, though selecting from a list of possible components might also work, or even pinyin entry), and then when they hit search it lists those characters which contain (at least) the components they've entered. It's a bit like searching by radical, except it's not limited solely to radicals, and you can enter several of them to reduce the 'hits'....

...

I guess this would need a some kind of database of character components (not just the 'official' radicals), which might be a sticking point..?

We actually already have such a database for the Character Info "Chars" tab - the main problem is cleaning it up so that variations on the same radical are all linked together, and coming up with a UI for this. But it's definitely on our to-do list - only question is whether we'll have the database cleanup done in time to make this a 2.3 feature.

I was wondering about this "cleanup", as the "components" and "containing" features have a lot of potential, but are currently of little use because of too much information, and containing characters that you can't even look up correctly. Knowing every single component of a character is overkill...would much prefer just the main components. If more information is needed, one can follow the trail to dig deeper (as with MDBG).

Similarly, listing *all* characters that contain a specific character is of limited value to me--just knowing the characters that directly use a particular character is sufficient. For example, I have no way to look up all characters with just the water radical character, except through the main character lookup by radical. Since some might find the current setup (all characters containing are listed), a toggle button would likely be needed.

Cheers,

David

 

[mikelove]

another thought: "next image" button on OCR Screen, which simply loads the next image in the list, maybe also a "previous image" button, ah, and while at it, it would be nice to show the filename in the OCR (if the image is from Pleco's library and not from the Photo library).

Makes sense once we're supporting "Next Page" / "Previous Page" for PDFs.

Most people hold the iPad at its sides, but all the buttons in Pleco are on the top and bottom of the screen, making some of them almost impossible to reach without putting the iPad down, why not put them optionally on the sides of the screen?

In OCR? In both orientations or just landscape?

I am collecting dozens of ideas how to make the user dictionary entry screen better. I have already created over 3000 custom entries, so I do have some experience. I kinda got used to it, although quite difficult. Have you ever actually used it yourself? For exampled for entering a four character idiom, where all four have both simplified and traditional versions, and you don't know the pronunciation? I think in such a case I switch at least 2 or 3 times between the unfinished new entry and the dictionary to get it right. If it can be found in other dictionaries, I can at least copy the traditional or simplified characters, or even the pinyin, but if its not the case it can take several minutes to create a single entry.
Anyway, I will just take some time, sit down and finally write down my ideas on how to make it better, if even just a third of it gets realized it would make work tremendously faster. Maybe its obvious by now: editing custom dictionaries and using these entries in the Document Reader is THE reason I bought Pleco, even more, why I bought an iPad in the first place.

I look forward to reading that. It's definitely a UI pain point at the moment, but to be honest it's not a very heavily used feature and it's tough to justify putting time into it ahead of something more ubiquitous like flashcards or OCR. We certainly need to streamline it a little at least, though.

I was wondering about this "cleanup", as the "components" and "containing" features have a lot of potential, but are currently of little use because of too much information, and containing characters that you can't even look up correctly. Knowing every single component of a character is overkill...would much prefer just the main components. If more information is needed, one can follow the trail to dig deeper (as with MDBG).

So you want just the first-level breakdown but you only want to dive in further by clicking on those components? That's extremely easy to add as an option, we just didn't think anybody would be bothered by seeing all of them since there aren't that many components and the higher-level ones come first. Another advantage of doing it the way now is that we can take you directly to the "characters containing" list when you tap on a component, rather than digging into another component list - is it worth the extra tap to you to see fewer components initially?

Similarly, listing *all* characters that contain a specific character is of limited value to me--just knowing the characters that directly use a particular character is sufficient. For example, I have no way to look up all characters with just the water radical character, except through the main character lookup by radical. Since some might find the current setup (all characters containing are listed), a toggle button would likely be needed.

"Directly use" is a bit difficult to define - in the case of water, I take it that you want characters where the water radical is by itself on the left like 海 but not characters where it's part of another component character like 匯? Again, easy to implement, we just didn't know there was demand for this.

 

[radioman]

I know these things are already under review. But I'm visiting the US and watching Youku. Both Youku and Tudou are easily viewable on the Macbook Pro or the iPad in China, the U.S., and I figure other places. Some of the content is blocked in the U.S., but there are many videos that are not.

I think it would be great to have the subtitle portion of the screen with Pleco on the fly. So I can easily hit the spacebar or whatever to pause, and then could do a word lookup.

I don't see this as an "on the fly" translation requirement, but similar to OCRing a photo. I envision it working as follows:

1) Hit the Pause button.
2) Pause invokes the OCR function (no need to hit an additional OCR button)

• - the characters would be available for mouseover (or finger selection) dictionary lookup.
  - the ability to save the character you highlighted to flashcards or word list.
  - the full OCRed text on the screen would be available capture to a list, or to custom flashcards.

3) Some gesture to continue the video with OCR dormant until next "pause".

In the case of the block reader, the user can highlight the portion of screen to analyze. I figure this would be similar. A control to advance forward and backward in 10/20/30 second intervals would be useful as well.

We could send them to Google Translate, I suppose, though it's admittedly not ideal... but if you're watching Chinese movies for learning purposes you're really just going to want those single-word lookups, and I think as long as we can get the OCR to happen automatically we can make that fast enough to be usable.

But basically, I'd rather take the extra time to do something like this than just embed an ordinary video player into our app and call it a day.

 

[mikelove]

I know these things are already under review. But I'm visiting the US and watching Youku. Both Youku and Tudou are easily viewable on the Macbook Pro or the iPad in China, the U.S., and I figure other places. Some of the content is blocked in the U.S., but there are many videos that are not.

I think it would be great to have the subtitle portion of the screen with Pleco on the fly. So I can easily hit the spacebar or whatever to pause, and then could do a word lookup.

I'm really not sure how we'd do these - as far as I know there's no reliable way to pull frames from embedded video on a web page, now that Apple has disallowed their old private API for taking screenshots. Video in standard formats / in your iDevice's media library is one thing, but streaming web video is another. But if we can make it work this is certainly a good basic outline for what we'd want to do with it, thanks.

 

[zisfro]

iOS Notification Center
Would it be possible to add a Pleco search screen in the iOS 5 Notification Center? Many times when I turn on my phone, its to look up a word in Pleco. But I have to unlock it, go to the springboard, and then open Pleco. There are many programs that can access the Notification Center. Perhaps it would be possible to make a Pleco search available in the Notification Center (Settings>Notifications>In Notification Center).
I would use that daily.

 

[mikelove]

Would it be possible to add a Pleco search screen in the iOS 5 Notification Center? Many times when I turn on my phone, its to look up a word in Pleco. But I have to unlock it, go to the springboard, and then open Pleco. There are many programs that can access the Notification Center. Perhaps it would be possible to make a Pleco search available in the Notification Center (Settings>Notifications>In Notification Center).

I don't think it's possible for us to add a search bar to Notification Center, so about all we could do is add it as a slightly faster way to launch Pleco. But I think we'd run into Apple approval problems if we did that - they don't generally smile on developers repurposing systemwide features like that for a purpose other than what they were originally designed for.

However, if Pleco is a recently-used application you could save a little time by double-pressing the home button and then tapping on the Pleco icon in the bar at the bottom of the screen.

 

[zisfro]

I knew Apple was picky about what could go in NC, so I'm not surprised that a search feature wouldn't be allowed (although it would be an awesome feature).
It might be possible to add it as a feature through Cydia for those who are jailbroken (like WeeSearch), but that may be digressing too much from your main task to be worthwhile.
Either way, Pleco software is awesome and continues to get better and better, so I'm sure if there is a useful feature out there, you'll find it and incorporate it.

 

[mikelove]

I knew Apple was picky about what could go in NC, so I'm not surprised that a search feature wouldn't be allowed (although it would be an awesome feature).
It might be possible to add it as a feature through Cydia for those who are jailbroken (like WeeSearch), but that may be digressing too much from your main task to be worthwhile.
Either way, Pleco software is awesome and continues to get better and better, so I'm sure if there is a useful feature out there, you'll find it and incorporate it.

Thanks! We've shied away from Cydia-specific features so far, though - I think if we were going to Cydia-optimize our app we'd do it more comprehensively with some sort of popup word-lookup shortcut.

 

[character]

In the OCR Block Recognizer, it would be nice to have some kind of HWR/RAD/Key recognizer overlay for those times when the OCR can't recognize a character. Zoom in on the character, pop up the overlay and write the character, look it up by radical, or try to guess the pinyin.

 

[多伦多人]

For the stroke order diagrams, it would great to have the ability to view the entire character in gray (or a light color) as the strokes are being drawn. Perhaps this could be added into Settings?

For example, something like the following:
http://www.nciku.com/search/zh/searchorder/1318591/
www.archchinese.com/chinese_english_dictionary.html?find=鱼

 

[mikelove]

For the stroke order diagrams, it would great to have the ability to view the entire character in gray (or a light color) as the strokes are being drawn. Perhaps this could be added into Settings?

Good idea; we already do have the option to show an outline of the character but a gray version might be more attractive.

 

[ACardiganAndAFrown]

I would love to see a feature that would allow one to be able select and export (copy, what have you) a list of words from the search history searches/entries. Usually I have a bunch of words in history that I want to make a record of (other than flashcards) to be able to copy them straight from the history would be extremely helpful.

 

[mikelove]

I would love to see a feature that would allow one to be able select and export (copy, what have you) a list of words from the search history searches/entries. Usually I have a bunch of words in history that I want to make a record of (other than flashcards) to be able to copy them straight from the history would be extremely helpful.

Would you want to dump them to a text file or to flashcards?

 

[ACardiganAndAFrown]

I would love to see a feature that would allow one to be able select and export (copy, what have you) a list of words from the search history searches/entries. Usually I have a bunch of words in history that I want to make a record of (other than flashcards) to be able to copy them straight from the history would be extremely helpful.

Would you want to dump them to a text file or to flashcards?

A text file or to the iOS copy heaven so that it can be pasted in notebook or writeroom something like that.