2.2.11 Bug Report / Feedback Thread

[mikelove]

Minor bug (New iPad): if you click on a character in a dictionary entry, have the keyboard open, click on the 字 icon, then click on the back button, the HWR/1/2/3/4/5 buttons will all be misaligned.

Thanks - this is one of those fix-it-once-we-drop-support-for-old-versions-of-iOS bugs, since the 1/2/3/4/5 button bar is already getting replaced by a much better system Apple introduced to do the same thing.

 

[radioman]

Oh, and with regard to Pleco and writing, a while back I commented that using a pen to delete characters when writing was inconvenient as you needed to use two fingers to clear the screen. I was thinking a single tap would be good (and still do). However, Skritter uses a slide up motion to clear the screen. I did not think I would like it but in fact do. It leaves you with your pen toward the top of the writing area and ready to go.

Could be worth a try, though in general it seems like very few people use a stylus in Pleco - most of the UI works better with buttons, and the mode-switch to and from a pen is enough of a hassle that people usually don't want to do it too frequently while using an app.

After further use of Pleco and Skritter together, I wanted to reiterate my interest in having Pleco add a Skritter-esk swipe-up-to-delete function. In this way, going back and forth between Pleco and Skritter becomes more consistent when using a stylus. I believe Skritter already uses Pleco's delete method, so this would give maximum flexibility for the user to choose their delete method, depending on what method they are using to write the characters (i.e., stylus vs. finger).

 

[mikelove]

After further use of Pleco and Skritter together, I wanted to reiterate my interest in having Pleco add a Skritter-esk swipe-up-to-delete function. In this way, going back and forth between Pleco and Skritter becomes more consistent when using a stylus. I believe Skritter already uses Pleco's delete method, so this would give maximum flexibility for the user to choose their delete method, depending on what method they are using to write the characters (i.e., stylus vs. finger).

I'd actually been thinking about swipe-up as a possible gesture-based way to invoke the "Done" button, which I know a lot of people have asked for - since three-finger tap is problematic because the fingers don't all hit the screen at the same time, swiping in the only direction that you'd never naturally swipe as the first stroke of a Chinese character seems like about the only alternative. (though personally I prefer "Done" as-is - it makes the switch from two fingers back to one very natural, and positions your finger perfectly to view the initial results and then decide whether to tap on one or scroll through them)

 

[radioman]

I'd actually been thinking about swipe-up as a possible gesture-based way to invoke the "Done" button, which I know a lot of people have asked for - since three-finger tap is problematic because the fingers don't all hit the screen at the same time, swiping in the only direction that you'd never naturally swipe as the first stroke of a Chinese character seems like about the only alternative. (though personally I prefer "Done" as-is - it makes the switch from two fingers back to one very natural, and positions your finger perfectly to view the initial results and then decide whether to tap on one or scroll through them)

I realize that Pleco and Skitter come from different organizations, but recent cooperation of companies provides some nice benefits to the users. And having the drawing interfaces work with a level of consistency makes it easier (e.g., Skritter can clear the screen with a two finger tap).

However, I use a stylus to write and would like to see a similar level of consistency. And this upward-stroke-to-clear makes sense for this, similarly positioning the pen to immediately writing again. I can then write portions of the character over and over again, until I can quickly execute an entire character without thinking about it.

Pleco has a whole lot of option settings. Maybe the ability to choose what the upstroke can do could be one of them.

 

[mikelove]

However, I use a stylus to write and would like to see a similar level of consistency. And this upward-stroke-to-clear makes sense for this, similarly positioning the pen to immediately writing again. I can then write portions of the character over and over again, until I can quickly execute an entire character without thinking about it.

Pleco has a whole lot of option settings. Maybe the ability to choose what the upstroke can do could be one of them.

Fair enough, just a matter of seeing if we can fit this in among the hundred odd interesting-little-feature-improvements-that-only-take-an-hour-or-so we haven't done yet for the next update.

 

[radioman]

Fair enough, just a matter of seeing if we can fit this in among the hundred odd interesting-little-feature-improvements-that-only-take-an-hour-or-so we haven't done yet for the next update.

Yes, I always enjoy the the activity of writing in a forum or blog without all those silly restrictions associated with actually having to execute/code/project manage the idea! Anyway, appreciate the consideration. Prioritize as appropriate.

EDIT: and for sure, please don't let this stand in the way of the PDF OCR function!

 

[mikelove]

EDIT: and for sure, please don't let this stand in the way of the PDF OCR function!

That's already in, though we're still hoping to get the performance up a bit (it's as fast as it can be now for full-page recognition - have to render the whole PDF to an image bitmap regardless for that - but we'd like to speed it up for scrolling-around lookups by only rendering the relevant portion of the PDF).

 

[radioman]

That's already in, though we're still hoping to get the performance up a bit (it's as fast as it can be now for full-page recognition - have to render the whole PDF to an image bitmap regardless for that - but we'd like to speed it up for scrolling-around lookups by only rendering the relevant portion of the PDF).

I assume you are talking about bringing a single page at a time for OCR processing. But I am less clear on what you mean by the "relevant portion of the PDF".

When I started further articulating my interest the PDF / sequential Photo function, I was trying to present my view of the minimum useful feature set. My thought was that the current version of Pleco is plenty fast for OCRing a full page graphic, and all that goes with that. If something (e.g., a diagram) gets in the way of a clear OCR read, then you can quickly redefine the OCR area in the page and carry that defined area setting to the next page.

I am all for optimization and speeding things up. But in my view, the real speed improvement (80/20) being the ability to flip to the next or previous page. My hope is that additional feature optimization will not adversely impact the schedule of getting a working PDF reader with paging function in place.

 

[mikelove]

I am all for optimization and speeding things up. But in my view, the real speed improvement (80/20) being the ability to flip to the next or previous page. My hope is that additional feature optimization will not adversely impact the schedule of getting a working PDF reader with paging function in place.

Well that's what this is about improving - right now we have to rasterize the entire PDF page in order to OCR it, which means that it takes a longer time than we're happy with to flip between pages. Fixing this may simply be a matter of adding code to render the PDF from a different source than the OCR bitmap but then quietly rasterize / OCR it in the background, but the point is that while we've got the basic feature working there's more work required to make it smooth enough to actually feel like you're benefitting from quick/easy OCRing.

 

[lechuan]

In OCR Block Recognizer:

If you long press a character to correct it, the screen that pops up with zoom in of the character allows you to draw characters on top with your finger. However, Once you start drawing the bitmapped image and drawn pen lines get compressed and dragged to the top of the screen until it's all a jumble at the top.

I am using iPad 3 with iOS 5.1

 

[mikelove]

If you long press a character to correct it, the screen that pops up with zoom in of the character allows you to draw characters on top with your finger. However, Once you start drawing the bitmapped image and drawn pen lines get compressed and dragged to the top of the screen until it's all a jumble at the top.

I am using iPad 3 with iOS 5.1

That's the problem - the iPad 3 doesn't really work well with full-screen textures of that size. Workaround should be coming in our next big update but thanks to Apple's UDID shenanigans there are a couple of minor iPad 3 bugs we can't address until then.

 

[estudiando]

Hello,

For a long time I did not update the audio files. Today I decided to update the female audio files only. There is a noisy/hissing background. The male voice which I did not upgrade has a silent background.

Is it possible to revert to the previous version of the female voice file (or is there another fix/setting?)

Thanks in advance

 

[scykei]

I think the hissing sound is because of how the volume of the sound files have been made louder. I think the only solution to this for now is to decrease the volume in the settings until the hissing becomes barely audible.

 

[mikelove]

For a long time I did not update the audio files. Today I decided to update the female audio files only. There is a noisy/hissing background. The male voice which I did not upgrade has a silent background.

Is it possible to revert to the previous version of the female voice file (or is there another fix/setting?)

This is because we made the volume louder, as scykei says - we certainly don't like the hissing, and in fact we're getting ready to mostly replace the female audio with a female TTS system, but we got so many complaints about volume / pleas-to-increase-the-volume-even-if-it-meant-a-reduction-in-sound-quality that we decided the hissing was better than leaving it as-is.

You can manually download the old audio from these links:

Basic Female Audio
Extended Female Audio

and install it via File Manager if you'd like to go back to the previous, quieter-but-less-noisy files.

 

[estudiando]

For a long time I did not update the audio files. Today I decided to update the female audio files only. There is a noisy/hissing background. The male voice which I did not upgrade has a silent background.

Is it possible to revert to the previous version of the female voice file (or is there another fix/setting?)

This is because we made the volume louder, as scykei says - we certainly don't like the hissing, and in fact we're getting ready to mostly replace the female audio with a female TTS system, but we got so many complaints about volume / pleas-to-increase-the-volume-even-if-it-meant-a-reduction-in-sound-quality that we decided the hissing was better than leaving it as-is.

You can manually download the old audio from these links:

Basic Female Audio
Extended Female Audio

and install it via File Manager if you'd like to go back to the previous, quieter-but-less-noisy files.

Thank you very much. On an ipod touch, the hissing sound is there regardless of volume.

I will revert to the old file. I can't believe people did not mind the hissing sound. Does the new male audio also has the hissing sound?

 

[mikelove]

I will revert to the old file. I can't believe people did not mind the hissing sound. Does the new male audio also has the hissing sound?

I don't believe so, though there may be some very faint hissing if you really turn the volume up a lot. We've had very few complaints about either file, actually, though I do think TTS is a better long-term solution in general since it's more space-efficient and adapts better to new words.

 

[radioman]

With a some guidance from Mike, I have a workaround that provides at least some relief with regard to reading documents via OCR. It involves having each page as a PNG file and then merging them using ImageMagick.

I have found that up to 9 png files (300dpi monochrome) can be horizontally merged and brought in the OCR Block Reader.

Although the png preparation might not be convenient, the approach at least eliminates the need to have to go and seek out the individual graphic files and load them.

 

[radioman]

Here are some various feature suggestions.

1) If I am in the dictionary or on a flashcard, tapping anywhere on the screen where there is no writing will cycle the dictionary to the next dictionary in your priority list.

2) When saving flashcards, allow them to be saved per dictionary. So that when a flashcard session is run, the definition that is shown will be the one for which the flashcard was originally saved. (maybe it already works this way, but I was unable to verify that).

3) In the OCR, double tapping will resize the paragraph to the width of the window/device. This would be for OCR on the PDF or a photo.

4) If Pleco has been rebooted completely (not just bringing to the foreground), then if the flashcards are set to not prompt for category each time, then prompt for category. The idea is that once you are in a session you are likely to use the same category for saving your cards. However, a clean start likely means that the category selection is much more uncertain.

5) If you are in the reader already, have some sort of refresh from pasteboard button. In this way if during your session, the clipboard is updated, there will be a way to grab it.

6) An option in the reader that, once you save the flashcard, the dictionary bubble will disappear. The idea is that "yeah, I've looked, seen the definition, like it enough that I want to further review it later, time to get backt to reading".

 

[character]

1) If I am in the dictionary or on a flashcard, tapping anywhere on the screen where there is no writing will cycle the dictionary to the next dictionary in your priority list.

There's already a button for that in the dictionary; you can tap on the character on a flashcard to get to other dictionaries. Personally would rather not have the entire screen active.

2) When saving flashcards, allow them to be saved per dictionary. So that when a flashcard session is run, the definition that is shown will be the one for which the flashcard was originally saved. (maybe it already works this way, but I was unable to verify that).

It would be good to have the current behavior better explained: http://www.pleco.com/ipmanual/flash.html#addflash as it's not clear to me either.

3) In the OCR, double tapping will resize the paragraph to the width of the window/device. This would be for OCR on the PDF or a photo.

Resize so that the width is set to the width of all recognized characters?

4) If Pleco has been rebooted completely (not just bringing to the foreground), then if the flashcards are set to not prompt for category each time, then prompt for category. The idea is that once you are in a session you are likely to use the same category for saving your cards. However, a clean start likely means that the category selection is much more uncertain.

While that example is not terribly inconvenient, would rather not have Pleco's behavior tied to whether or not iOS has had to remove it from memory.

 

[mikelove]

radioman wrote:
1) If I am in the dictionary or on a flashcard, tapping anywhere on the screen where there is no writing will cycle the dictionary to the next dictionary in your priority list.
There's already a button for that in the dictionary; you can tap on the character on a flashcard to get to other dictionaries. Personally would rather not have the entire screen active.

I'd tend to agree, though we are adding the option for a dedicated dictionary switch button to appear once a flashcard is revealed.

2) When saving flashcards, allow them to be saved per dictionary. So that when a flashcard session is run, the definition that is shown will be the one for which the flashcard was originally saved. (maybe it already works this way, but I was unable to verify that).
It would be good to have the current behavior better explained: http://www.pleco.com/ipmanual/flash.html#addflash as it's not clear to me either.

They are saved per the current dictionary, but if there's already a card with the same definition then it keeps that existing card / definition and just links it to the new category. We have added an option in the duplicate card prompt to update the existing card to use the new dictionary, though.

3) In the OCR, double tapping will resize the paragraph to the width of the window/device. This would be for OCR on the PDF or a photo.
Resize so that the width is set to the width of all recognized characters?

This one fits with your earlier OCR suggestions I think - some sort of improved resize behavior might happen, but the main thing we're trying to get added to OCR in the next update is PDF support.

4) If Pleco has been rebooted completely (not just bringing to the foreground), then if the flashcards are set to not prompt for category each time, then prompt for category. The idea is that once you are in a session you are likely to use the same category for saving your cards. However, a clean start likely means that the category selection is much more uncertain.
While that example is not terribly inconvenient, would rather not have Pleco's behavior tied to whether or not iOS has had to remove it from memory.

Yes, in fact in general we're trying to make that happen less.

5) If you are in the reader already, have some sort of refresh from pasteboard button. In this way if during your session, the clipboard is updated, there will be a way to grab it.

That should already be happening automatically - is it not doing so now?

6) An option in the reader that, once you save the flashcard, the dictionary bubble will disappear. The idea is that "yeah, I've looked, seen the definition, like it enough that I want to further review it later, time to get backt to reading".

Reasonable little speed improvement, just a matter of adding an extra checkbox somewhere to enable this.