Beta 5 Bug Report / Feedback Thread

[mikelove]

westmeadboy - no way to kill that notification message, unfortunately; the best workaround if you can't use the hardware button is go to with the third option, "on program activate," and use some other application launcher program to bring up Pleco with IA. (we probably shouldn't have made that preference default to On)

weight - that's odd, trying it out just now those all seem to come up correctly - have you changed any of the search-related preferences from their default settings?

sych - thanks for the note on that, looks like a few of the audio files are still mis-tagged - we'll do another run-through at some point between now and the finished release and hopefully kill the last few errors like this.

haraldalbrecht - glad you eventually managed to get it working. The TX in general doesn't seem to be getting along well with 2.0, strangely enough it's quite a bit more stable on late-model Treos in spite of the fact that they're running virtually the same OS. It's unfortunate that the TX has to be so problematic given how many people have TXes... anyway we're hammering away on our TX now and hopefully it'll be at least a little more usable soon.

We're now seriously considering removing Palm OS altogether from our hardware recommendations page, as it's looking less and less likely we're even going to attempt a Palm OS port of anything past version 2.1 or so. But we should at least be able to get 2.0 stabilized on Palm with a few more weeks of debugging.

llammamama - yeah, we're still trying to get those resizers working consistently; may have to implement them at some sort of higher level than regular controls so that they get first dibs on all pen movements. Copy/paste from the edit card screen is a bit of a copyright problem, unfortunately, since the text isn't encrypted/protected once it's in your flashcard database.

torrero - we dropped support for Sony's proprietary VG system in 2.0, it was unfortunately way too difficult to get it working correctly.

 

[malcdelnorte]

Mike,
the document reader is what I've been waiting for - actually, it's what I imagined I was going to be able to
do when I bought my plecodict license

Anyway, problems so far on my Palm TX after a reasonably straight forward straight install with no error messages:

1. when the TX powersaves and turns itself off and Pleco 2.0b5 is running, trying to turn the TX back on has turned
ugly three times now resulting in a blank white screen. My main on/off button is knackered so I click the leftmost
hardware button (with a wee house on it) which should show me a screen of my favourite apps.

2. no sound as yet but I think I read something about having to do something special to get that to work which I
magically forgot.

3. seeing the offset for the pen position worse than normal I think(I get it a bit anyway as a standard mature TX "feature").

Am loving the new version - thanks again,
Malc.

 

[weight]

I've created a default category which I've added 5 dictionary entries to. I leave the settings to default. I test the five cards marking two as right and three as wrong. I go back to test again and I get those same five cards. Why are those 2 cards I marked correct not advancing in score and disappearing to come up tomorrow or the next day.

" weight - that's odd, trying it out just now those all seem to come up correctly - have you changed any of the search-related preferences from their default settings? "

Mike I just did a factory reset and reinstalled pleco. Even with everything set to default it's not giving it to me.

For the record it's a TX

 

[ldolse]

First round of feedback. Most of below is negative, but overall this release seems like a major improvement.

Crashes
Loading an xml file that's in another format instantly crashes Pleco - I accidentally loaded a Chinesepod xml flashcard export and this immediately crashed Pleco.

Hangs (bad)
These two are hanging Pleco in some sort of infinite loop, but in turn Pleco is preventing any new Windows Mobile screens from coming up, so the only way to recover is soft reset.

• Hitting stroke order Play on a rare character which doesn't exist in the stroke order DB hangs pleco. Stroke order tab should probably be disabled for characters not in the database, BUT to really do that the Kangxi radical number on the main Char Info tab still needs to be replaced with the actual radical instead of the radical number. Radical character + number of additional strokes would be ideal to keep the info that's there.
• Looked up 慰 in GuiFan. Highlighted the headword and then clicked the highlighted character to bring up pop-up definition. Popup Definition was Adso. Hitting Play int he pop-up definition box consistently hangs Pleco/whole phone. Normally playing the audio from the main window works for this entry, but it also crashed once just hitting the play icon.... I also successfully played the audio from the popup by using ABC instead of Adso

After crashing and resetting the phone it defaulted to Guifan as the first dictionary again
History is still really flaky, doesn't survive across crashes/resets..


Flashcard Import
This was from an old school 1.0 import file, which had been hand edited. A number of entries were just the character headword, no Pinyin. I set Adso as my first dictionary, Oxford second, ABC Third, removed Guifan

• got multiple matches for 我 － one was the headword 我軍jun1, meaning our army. Seems like this is a bad entry in Adso? Either way it shouldn't have matched, as there were two characters in the headword.
• Similar issue to above - 大 came up with '大? Daihatsu' in the list of options. Aside from the fact that this is a bad entry and should possibly be erased, The ? essentially makes this a two character headword, so it shouldn't match.
• ’美國 meiguo‘ came up twice during the import process, apparently it's a dupe in my file, but the 'skip duplicates' option seemed not to work.

Adso
many duplicate entries - the definition text itself is identical, but the qualifier in parentheses changes, and many of these qualifiers seem to have no useful meaning, such as (xing). I don't have a problem with every sense of a word having it's own entry, as that's the way Adso works, but the duplication of the same sense is problematic. It would be really nice to clean that up before this goes GA.


Instant Access

• IA seems a little slower than B4, but on activate works as desired now, so I'm very pleased overall.
• The taskbar icon still doesn't work with large start menu enabled on an Touchflo device - brings up the program instead of IA. This will probably just have to be filed as a known issue, just noting it.
• Taskbar icon no longer crashes the device when there is also a new message notification, but now it takes over the 'Notification' softkey which used to be tied to alerting and showing a user new messages, so attempting to click notification to view messages just launches Pleco. The message notification icon also no longer appears in the taskbar. I noticed other users expressed other problems with the taskbar option - maybe just best to retire it.

Nits

• Audio seems too quiet - I know the phone is capable of generating more volume - I have it cranked all the way.
• Resizer bar issue noted in known issues, but there is a highlighting issue associated with that. When you miss the resizer most of the time the nearest entry becomes highlighted. Problem is it doesn't become unhighlighted when you select an alternate entry.
• Char Info should be in context menu on single character highlight, particularly if one decides they don't like tap on a character to bring it up - which is generally nice but often causes problems with other features like magnify, see below.
• No good way to get out of magnify - clicking anywhere on the screen is flaky, especially clicking anywhere in the definition text box causes all sorts of bad things to happen, most frequently bringing up CharInfo by accident. Magnify font size was set to 60.

 

[Alexis]

A few bugs/oddities I've found on my Tungsten T3 related to the stroke order features:

1) The behavior of tapping a character for stroke order seems to be quite buggy (in different ways for different characters):

ie. Go to English-Chinese and select ABC. Select:

A-ban3 (works fine)
Click on character, select stroke order, press done. Repeat. OK.

A-gu3 (soft reset 2nd time)
Click on character, select stroke order, press done. Click on character, "Fatal Exception"

A-ku (uses a low res font 2nd time)
Click on character, select stroke order, press done, Click on character, Now see a low res font in the stroke order diagram (have font-18,24,48,96 installed)

2) If I press the palm "HOME" icon while in the stroke order diagrams (also saw it while in the doc order), causes a soft reset.

PS. This release of Pleco 2 is looking amazing. I really like the new "fade" feature on the stroke order diagrams.

 

[Alexis]

Regarding the Flashcards: I'll describe what I tried to do that brough up a couple points of confusion for me with the new system:

Importing and Categories:

I imported all my Pleco 1 flashcards from xml and the import was great! Only had to a answer a couple conflict questions which it handled very well. It created all my seperate flashcard files as categories, which is exactly what I wanted. I also figured out how to "group" categories into sub-categories. Very useful.

Manage Flashcards:

I couldn't figure out how to use this. I understood v1 where it has the two seperate columns to allow you to transfer/copy cards between groups. Now with the new stroke order diagram feature, I wanted to make a special "stroke order" diagram category and put in selected characters from my existing categories into this category. I couldn't figure out how to copy selected characters from one category to another. Is this possible?

In the "more" page, there is a "remove" cat function which eemed to mess up the index in the corresponding drop down field, but didn't seem to do anything else. Not sure what it's supposed to do though...

Profiles:

My big mistake was copying my regular profile to create a strokeOrder profile, then deleting a whole bunch of categories that I didn't want in that profile. I then discovered that this deleted the categories in all profiles.

It seems that my understanding of a profile was a bit off. At first I thought it was a large container for seperate sets of flashcards, but now my impression is that it is a way to keep seperate sets of scores and test settings; Is that correct? If that is the case, then I would suggest calling this "test profile" insted of just "profile" to make it clear to the user what it is. I do like how it remembers the test settings for each profile That is very handy.

 

[mikelove]

macdelnorte - glad we could finally do something for you on the document-reader front.

Have you used flashcards at all yet? We haven't seen this power-save failure behavior yet, but we've seen similar behavior switching to Applications after a big import and I'm guessing they might be related - the database caching system on the TX really doesn't seem to get along well with SQLite for some reason, fortunately if all else fails it's actually possible to circumvent it altogether and write data directly to the Palm's flash memory so one way or another we should be able to get this stabilized soon.

Sound only works if you've both installed and paid for / unlocked the audio databases, we're still deciding what portion of them to make available in the demo version. (which in 2.0 may actually include a lot more dictionary content in general) Have you noticed this pen offset problem everywhere or just in specific areas like the resize bar?

weight - check the Info on each of those cards (bring up Manage Flashcards with "all cards," select a card and tap the 'i' button) - have the scores actually changed on the ones that you marked right? Is the system configured to use repetition-spaced or frequency-adjusted mode? (this is in the first tab of Advanced) In frequency-spaced mode that behavior is actually perfectly normal, since if there are fewer cards available than the total number you ask it to test on it's going to have to pick even less-frequent cards, but in repetition-spaced mode they should disappear after the first test run.

I'm seeing the Zhuyin problem you mention on Palm but for some reason not on Windows Mobile in spite of the fact that the search engine code on both is essentially identical - not sure what's going on with that.

ldolse - thanks for the as-always lengthy and detailed report.

Hadn't done any testing with other XML formats but we'll definitely do that and make sure it behaves itself if you accidentally open one. (actually in theory it should be possible to embed Pleco flashcard XML content into another XML document - conceivably you could even create an (XHTML) web page with a word list that included extra tagging for Pleco, though whether that would have any practical use beyond a cool technology demo I don't know)

Not seeing that problem with 慰 in a popup definition - which preferences have you changed from their default settings?

History is indeed very iffy at the moment, we're trying not to slow the system down too much by continuously saving it out every time you jump to a new entry but it does need to be written out a lot more often than it is now. (which is basically only when the program completely exits - at the very least it should probably happen whenever you put it in the background)

The skip duplicates option was actually working fine in that 美國 case, we check for duplicates after you disambiguate just in case you have two cards in the same file that look the same but are actually supposed to be different - a character that's pronounced in two different ways, for example, or that has different entries depending on the traditional character variant. (not likely with a 1.0 export, since that includes traditional characters and Pinyin, but quite possible with other vocabulary sources) If you search your flashcards you should only see one 美國 actually listed.

我军 in Adso actually has just one pinyin syllable, the jun1 is in there but it's missing the wo3, so that's why it's matching - the software thinks it's a one-character word 我 pronounced "jun1." Editorial work on Adso unfortunately isn't something we can really spend any time on now, since we have dictionaries we're charging people money for that need work too. The fact that Adso's constantly being updated and that we'd have to keep syncing those changes back to it makes it even harder for us to go through and manually fix it for a Pleco conversion. So honestly the best way to get problems like this fixed is to fix them on Adso's website and wait us to get them into Pleco that way.

MDBG's CC-CEDICT will probably be a better bet for flashcards once we get that converted, actually, since it has a rigidly imposed one-entry-per-character-pinyin-combination rule; unfortunately it doesn't have unique ID codes for entries, which makes it a bit harder to match up flashcards between versions, but we've already worked around that in a few other databases.

Not sure why IA's feeling slower - might be the extra dictionaries it has to search. Glad "on activate" works at least. We're probably going to stop making the taskbar icon a default option, but I'd rather not retire it altogether since for non-tech-savvy people or people who aren't doing much with their Pocket PCs besides running Pleco it's still a very convenient way to use IA.

We've gotten a couple of complaints about audio volume by e-mail too, actually - probably just something we should add a preference setting for. That list highlighting glitch should have been mentioned in known issues too since it's another thing we've seen a few times. Good point on the context menu for Char Info, we also need to get context menus working better with just one character selected (another known issue I didn't mention) - right now it often goes to Char Info before you even have the chance to see the popup. Did magnify exit more gracefully in 1.0? We use essentially the same code for it in 2.0 that we did in 1.0 so I'm wondering why it would only start acting up now.

Alexis - what do you mean by "low-res font"? Is it showing up like a magnified but pixellated character or does it just look more blocky / squarish than before? Is it showing up in the full stroke order box or just where the magnified character is in the details screen?

I'm seeing that home button issue here too, not sure how we didn't notice that before. Glad you're liking the fade option, we know it's not quite as good as having digits but since we couldn't come up with a good way to place those next to strokes it seemed like the best we could do.

In manage flash, the way you assign cards to categories now is to choose the category you want from the popup menu at the bottom of the screen and then tap on the icon with a folder and a plus sign on it - since cards can now belong to multiple categories the two-column approach no longer made sense, but doing this should take the same number of steps (pick the destination from a popup list and tap once on a button for each card you want to add).

Profiles are indeed just for separate sets of scores and test settings - there's been a lot of discussion here about that, probably at the very least the "Delete" message in Manage Categories should state something to the effect that this affects all profiles to avoid confusion.

 

[Alexis]

Alexis - what do you mean by "low-res font"? Is it showing up like a magnified but pixellated character or does it just look more blocky / squarish than before? Is it showing up in the full stroke order box or just where the magnified character is in the details screen?

Looks like a magnified but pixellated character in the stroke order box. Stays pixellated when doing the stroke order animation.

Profiles are indeed just for separate sets of scores and test settings - there's been a lot of discussion here about that, probably at the very least the "Delete" message in Manage Categories should state something to the effect that this affects all profiles to avoid confusion.

Personally, calling this "Test Profile" instead of "Profile" would clear up any confusion for me.

 

[ipsi]

Radical input has some issues - it works fine, but there's a huge amount of garbled crap after the last entry when selecting 1.0-style menus. Seems ok with 2.0-style. Selecting one of these garbled entries selects the Traditional 'Dragon' radical.

Having the IA button do the pop-up dictionary when in Pleco would be good. Should that be working or is that still something you'll 'get around to eventually'?

I also had some issues with copying pinyin to the input field. Can't remember where, but the problem was that it would put something like pa1n in, rather than pan1, which wasn't very helpful. Might have been a link, but I don't believe so.

 

[bodybag77]

Hi!
Amazing beta Mike! it was worth all the wait, the product is well polished, it could be a final release!
Except few minor bugs already discussed here it all seems to work properly, and the overall performance seems faster than p 1.0.
I've just bought the licence for the 2 new dictionary and i have to say that the 21cn dictionary is massive! a must-have!
Thank you and greetings to all the Pleco user community

 

[ldolse]

Not seeing that problem with 慰 in a popup definition - which preferences have you changed from their default settings?

I changed a number of things, but can't imagine they'd cause an effect. Here's what I remember:

• Reduced Headword Size to 26 - this beta's default seemed really large
• Increased list font size to 16
• Increased zoom size to 60
• Disabled Taskbar for IA, enabled 'On Activate'
• Changed search delay to 1/4 sec
• Increased Radical Input Font size to 20
• Enabled Scroll pages with Nav buttons in Reader
• Changed Defn tap-sel action to Popup Defn - Thanks for getting that in!

The skip duplicates option was actually working fine in that 美國 case

Ok - I manually rejected it, I'll retest

我军 in Adso actually has just one pinyin syllable, the jun1 is in there but it's missing the wo3, so that's why it's matching - the software thinks it's a one-character word 我 pronounced "jun1." Editorial work on Adso unfortunately isn't something we can really spend any time on now, since we have dictionaries we're charging people money for that need work too. The fact that Adso's constantly being updated and that we'd have to keep syncing those changes back to it makes it even harder for us to go through and manually fix it for a Pleco conversion. So honestly the best way to get problems like this fixed is to fix them on Adso's website and wait us to get them into Pleco that way.

I wasn't really suggesting that you should edit Adso, I realized the cause of the error was that Pleco saw 我'something' plus a single pinyin syllable. I just thought Pleco should be matching both on number of characters in the headword as well as syllables in the Pinyin.
The only other point I meant to make with Adso is that duplicate definitions could be handled as a function of your input scripts, editorial work not required. However if 1.0 goes GA with dupes then it's difficult to remove them in the future due to the aforementioned entry ids. I also agree CEDICT will be much better for this in general though....

We're probably going to stop making the taskbar icon a default option, but I'd rather not retire it altogether since for non-tech-savvy people or people who aren't doing much with their Pocket PCs besides running Pleco it's still a very convenient way to use IA.

Fine to keep it - I'd just strongly recommend changing 'On Activate' to default and disabling the taskbar option.

Did magnify exit more gracefully in 1.0? We use essentially the same code for it in 2.0 that we did in 1.0 so I'm wondering why it would only start acting up now.

It exits beautifully in 1.0. I think it's not so much the magnifier that's the problem as is the fact that everything else you click on is still 'active' when you try to get out of magnify. In 1.0 the rest of the screen seems to be inactive until you get out of magnify.

 

[mikelove]

Alexis - that actually sounds like it might just be a normal character - what happens if you turn on Outline, does that smooth out the edges any? We're working on a much prettier stroke order system (using some new data we licensed from Wenlin) for 2.1, which we may end up using to render rare Chinese characters in regular definitions too.

ipsi - the IA button works using a different mechanism than regular Pleco button controls, and we actually turn it off on startup specifically so that you can assign that button to another function if you want to; you should be able to configure that button to bring up a popup definition in the button commands screen though. The 1.0 radical problem should be easy to fix... the Pinyin issue has actually been around for a while, we've got a much better tones-to-numbers converter system which we use for our desktop dictionary converter software but we haven't gotten around to porting it to the handheld version yet as it's kind of slow (basically it recurses through every possible breakdown of a tone-marked phrase to figure out which character breaks are correct)

bodybag77 - thanks! Glad it's looking good. The 21st Century dictionary is indeed massive, we were delighted to get that license particularly given how few entries our previous E-C offerings have had.

ldolse - just realized what the problem is with IA performance: the "search for files on activate" (or something like that) option is turned on, so it does its full startup file search every time you bring it up. You can turn that off in the same panel that you turn on IA - we have it on by default for the sake of new users trying to get all of their files installed, but once Pleco's found everything there's not really any need to leave it on. (it'll still look for new files whenever you restart it)

Still not seeing that audio crash, maybe it's device-specific; we'll try it on some other models. You're right that we could handle those duplicates automatically, but it's actually not that big a deal to fix them later on - the UID system will perfectly happily handle a mapping of multiple UIDs to a single entry, it can't go the other way since it needs to have just one ID to put in a flashcard but we can merge entries together late on and still have both of the former IDs work. We'll see if we can figure out what we did differently in magnify this time around, might have something to do with our new text field controls.

 

[stephanhodges]

(Tapwave Zodiac Palm. landscape mode)

READER ITEMS

I tried to open a document in the reader. Selected Card3 (left external card slot - 1GB). It had a default of "/Palm/Launcher/document.txt". I erased document.txt, leaving the "/" as the last character. Click START causes a crash AND an ugly continuou sound. memoryMgr.c line 3655 NULL handle

The SAME crash happened when I put in "/Palm/Docs/SGYY.doc" which DOES exist at that location (according to FILEZ). This is on the external card also.

I have about 30+ document files on the external card. Is there a way to browse and select the file name? Seems incomplete without this.

Is there any way to change font sizes WHILE in the reader?

I can't figure out what "Magnify" is supposed to do. Tried several combinations of options, but don't notice any differences.

The left/right hardware buttons don't do anything, although the up/down ones do.

I don't like the highlight bar in the middle (below the text). I prefered the old placement at the top of the text.

Even with "large font" for text selected, it's still pretty small. However, the chapter index combo select box is still SMALL text. I expected the same size as the body text.

Using left/right arrows on the highlight bar, every time I click forward, the dictionary reverts to some other dictionary at the word “一” "yi1" (one). I'm at the 42.22% position. “ 噜苏一道也就算了；”

Percent position is fine to judge progress in a book, but awful for describing the exact location of the text. Can you show the current character number position also?

I figured out how to use the bookmark (tap and hold, then select), and it positions the text to the correct words. But, there's no indication on the screen of where the bookmark is. Can it be underlined, or something?

So, bascially, I can't figure out how to read anything loaded on the palm except the example "Story of Ah"

 

[ipsi]

Ah, right. It's all working just fine now . Except PocketTunes still intercepts it. Oh well. I don't imagine there's much you can do about that.

Also, I see a large number of arrows and such in the Guifan - are they meant to be there or what? They seem a bit on the confusing side.

Stephen: Sounds like you left it at /Palm/Launcher/, correct? That'd be expected behaviour for the Beta as there's no filebrowser implemented... I imagine it'll give you an error message in the finished version if they don't have time to implement a filebrowser.

It also can't open DOC files - you'll need to point it towards a .txt file or .pdb file (there should be some somewhere on this forum).

 

[ldolse]

I tried turning that IA option off, seems like it might be slightly faster. Then again I might just have a misconception on Beta 4's speed.

Just found another thing which really surprised me. At first I thought it was a major bug, though after a bit more analysis I see it's a mix of expected behavior and bugs. I was playing with instant access and somehow the characters 话与 wound up in the search box. Since this isn't a word, I was expecting the 1.0 behavior that it would look up only 话 and ignore 与. Instead it flipped to full text search mode, which I hadn't seen before.

What happened was that ABC came up, and it highlighted the where the terms were found. 中 became inverted, which is cool now that I understand it. What DIDN'T happen was that the list mode didn't change to 果, in fact even if it had previously been on 果 it would change back to 典. While I prefer 典 for a normal lookup, for full text search I think 果 is mandatory.

Next I clicked the dictionary icon to see other results - 21C came up as the only other dictionary with results, and in this case it actually flipped to 果 as desired. The weird thing was the icon was still an inverted 中. I would expect it should be an inverted 英. Clicking the language button to flip to 英 basically disabled full text search, which wouldn't exactly be the behavior I'd expect, but not sure what to expect under this automatic scenario regardless.

Next I tried doing the full text searches with the # modifier so that I could actually force it. It seems to me that full text isn't working in a completely language agnostic way, and I think it should. Examples:

• I tried the search #shine. In the E-C dictionaries this works as expected, entries with 'shine' are shown, 英 was now inverted. However, flipping languages and searching the same term with the # qualifier on the C-E side didn't work as expected. Instead it turned into a Pinyin search - what i wanted was to find any C-E definition containing 'Shine' in the text.
• Tried 话与, but now as #话与 to force full text search either direction. Adding # pretty much made no difference to my previous description. The problem is in this case I'm now using # to be unambiguous, and it's still not searching the E-C dictionaries.

Regardless, aside from the 21C anomaly, full text search isn't working for the use cases that would actually be more interesting - What I was really looking forward to using it for was for English text in the C-E entries, and that doesn't look possible right now. Hopefully it's just an interface bug, and not a major change to search both languages from either direction.

Aside from all that I think I see the use method I'd prefer to control the modes - # is fine, but what would be nicer is to give the language icon a tap'n'hold drop down list just like the dictionary icon. The dropdown would allow you to choose E-C, C-E, Full Text E-C, Full Text C-E. I like the inverted icons as is. I think just clicking the icon without holding it down should still just cycle between C-E and E-C (although cycling between all four would help new users learn about it faster). For myself I'd probably just disable the automatic full text search that can come up under specific scenarios, I'd rather have the 1.0 behavior.

 

[neilperks]

Thanks Mike,

Really impressed by the current version, but will give it a good test over the next few days.

Just one thing immediately hit me, which I think has been mentioned by others is the audio is almost so low its nearly inaudible except in a very quiet situation. The audio on the PocketPC is set to maximum which is normally way too high for most programmes. I noticed the maximum volume on previous betas was very low, but it seems even lower this time (or maybe my imagination). Any chance of increasing this as too loud can always be reduced.

Cheers

Neil

 

[llammamama]

Flashcard issues:

Using the following settings: Repetition spacing, hour-based, auto-adjust, free-answer mode, show pronunciation /definition, ask for headword, prompt for answer quality, allow audio for all flashcards (and a few others mabye).

When I open a new session when there are no 'pending cards', instead of getting an error message ("no cards match the query etc.") I get an empty flashcard session. The handwriting recognizer shows up, as well as the cancel, back, information buttons, however there are no cards and no way to do anything.


Also, I seem unable to set the option "reveal fields separately" regardless of the other flashcard settings (in this case I am using repetition spacing, user-scored, show headword). If I check the option and hit "ok', then it doesn't actually take effect and the next time I go into the option menu it is un-checked again.


I've also had some weird persistent quasi-empty session issues which fixed themselves when I removed all flashcards from the category and then put some back in. In this case I was unable to do any review (free-answer ask for headword) at all since the first card to come up was always empty - although in this case there were four "kanji answer boxes" above the handwriting recognizer even though nothing else was present.


Still, its looking pretty good!

 

[ipsi]

Couple of quick questions about the XML Mike: In the <pron> tag, what should the type attribute be for Zhuyin? zyfh? And for numeric pinyin (e.g. ni3), should the tones attribute be numbers or numeric or what?

Thanks.

 

[weight]

Mike,

is there no way to skip a card while testing?

 

[mikelove]

stephanhodges - thanks for your feedback on the reader. Unfortunately we didn't have time to put in all of the features we wanted in the reader in this release, but we'll certainly be expanding it in the future assuming it proves popular.

The crashes look like they're a simple missing error check, should be easy enough to fix - thanks for finding that. There's not going to be a full folder-based file browser in 2.0 - just too complicated to write - but we may at least be able to show you a list of .txt files on a memory card rather than requiring you to manually enter their names; 90% of the code for that is already in place for Pleco to find its data files on startup so it shouldn't be too difficult to add the remaining 10%. For right now, if you enter the full path to the document on the card and the document is in text format it should come right up.

There's no way to change font sizes while in the reader - it's supposed to save your place in a document when you exit (and restore it when you reopen) but that functionality is a bit buggy at the moment. "Magnify" only affects the characters in the highlight bar - we thought it made more sense to put it in the middle since that makes it clear that it's determining which characters are defined in the dictionary below, but if a lot of people prefer it the other way it wouldn't be difficult to add a checkbox to move it back. (by 2.1 I imagine we might have two or three screens of reader settings, just as with flashcards / dictionary searches now)

Do the left/right buttons work elsewhere in Pleco, say for moving the cursor in the Input Field? Might be a Zodiac issue rather than a reader-specific one. The chapter index popup list control had to be designed specifically for that one place in the software (since there was no way to get Chinese characters displaying correctly in a built-in Palm OS popup list) so we didn't put quite as much time into it as other custom UI controls like our text fields - it should get better in the future, though.

I'm not seeing that "一" problem here - which dictionary is it showing you an entry from? I'm not quite sure where we could squeeze a character position indicator into the interface without taking away something else - on a landscape display there might be room but not on a portrait or square one. A bookmark indicator is a nice idea but too complicated to consider before 2.1.

ipsi - yeah, can't do much about PocketTunes, they're much better at low-level hardware hacking than we are The Guifan arrows actually are supposed to be there, and there's a specific logic behind all of them which you'll be able to read all about once we get around to posting a copy of the intro document - basically the triangles mark the starts of example sentences and the arrows mark different degrees of jumps between sub-definitions.

ldolse - the zhong/ying indicator icon has actually been redefined in 2.0 to refer to the current search language, rather than the language of the currently-selected dictionary; thus, English full-text results from both the E-C and C-E dictionaries will come up with that inverted 英 icon. So if you tap on the dictionary switch button a few times with #shine in the input field you should eventually see a list of ABC results beginning with 遍照。The reason flipping languages left you with a Pinyin search is that it assumed in that case that it had auto-detected your search language incorrectly and that you were actually looking for Pinyin (which isn't supported in full-text searches yet, but we wanted the behavior to be consistent and that is the correct behavior if you enter something that could be Pinyin or English and it auto-detects the language incorrectly)

The original design actually had a full-text search toggle button, but we switched to the prefix character because we were very worried about what would happen if people accidentally switched themselves into that mode and didn't know it - they'd enter a very common word and still only get full-text results and think the software was broken. With the current system, anything that puts them into full-text mode will be cleared up as soon as they enter a new search, so even if you do it it by accident you don't have to know anything about it to figure out how to get out of it. We may disable the full-text fallback preference by default in the finished version, haven't decided about that yet.

neilperks - noted on the audio, there should be a volume control preference setting in the next beta.

llammamama - so there's no message at all with a blank session? That's odd, the software double-checks if a session exits without any cards having been displayed and is supposed to show an error message in that case... might be something specific to free-answer mode. Reveal fields separately sounds like a missing line of code in whatever function is responsible for saving out that dialog box - we actually designed a whole fancy system to automatically take care of setting up / saving out all of the dialogs in Preferences (and eliminate errors like this) but haven't gotten around to implementing it for flashcards yet. The empty card might be some INVALID or oddly-converted card from your 1.0 deck - I'd suggest doing an All Cards search in Manage Cards, going to the Settings screen and having it sort by headword - that should make it easy to find and eliminate any blank / garbled cards.

ipsi - there's no support for Zhuyin in the importer yet, the importer doesn't even look at the "type" attribute in fact. (waiting until Zhuyin support is a bit more solid before we add that) Use "numbers" for numeric Pinyin.

weight - the "Remove from Session" button (enable it in the Display panel in Advanced settings) will do that.