First of all, great job on this new version of Pleco! I had been using version 1 on my Palm up until yesterday. I just started using an unlocked copy of version 2 for Windows Mobile yesterday. A lot of great improvements versus version 1! I especially like the ability to highlight a character to find out other words and phrases that use the character. My only criticism is that, at least on my HTC diamond, the windows can seem very cramped for long phrases. Keep up the good work.
Beta 8(b) and 8(c) / Release Candidate 2 Bug Report Thread
- Thread starter mikelove
- Start date
hleperlier
举人
I got held up and could not submit bug reports for the previous beta; also, I didn't want to post reports without having read all the bug reports. I installed 8c 24h ago. I've tried to make it crash, haven't managed it yet. Crashes I used to have when switching from one mode to the other (from pop-up, to input characters onto mainline, for the ones I could identify but more random than anything) seem to have all gone. I've tried to do everything very fast to confuse the programme. It hasn't crashed yet.
I'll see what happens when I tried to switch apps frequently. I' ve used the reader: no problem, apart from one small glitch (or a feature?): to move up/size up the top line for the Dictionary appearing at the bottom of the window, I've discovered that the only way to do it is to grab the top line to the left-hand side (2/8th of an inch) of the two indentures. Otherwise, only the pop-up menu appears. In the ADSO (if I'm not mistaken) characters are still only displayed in black and white (no tone-associated colours).
Speed is much improved and seems to be as fast as on Pleco 1.0.
Palm TX/4GB card
Henry
I'll see what happens when I tried to switch apps frequently. I' ve used the reader: no problem, apart from one small glitch (or a feature?): to move up/size up the top line for the Dictionary appearing at the bottom of the window, I've discovered that the only way to do it is to grab the top line to the left-hand side (2/8th of an inch) of the two indentures. Otherwise, only the pop-up menu appears. In the ADSO (if I'm not mistaken) characters are still only displayed in black and white (no tone-associated colours).
Speed is much improved and seems to be as fast as on Pleco 1.0.
Palm TX/4GB card
Henry
mikelove said:
Yeah, I do. Since posting that message (during which time I've restarted PD2 a few times) the problem has gone away. Not sure what changed.
One other thing I'm noticing in 8c, which I'm pretty sure wasn't happening in 7d: the user dictionary entries I imported from PD1 aren't showing up when I look words up. I can still see them on my flashcards if I go to Edit Card from the flashcard search UI (plus they show up when I'm being quizzed), so I know they're in there.
Actually, I'm pretty sure even some user dictionary entries I've created in PD2 aren't showing up. When I switch to the user dictionary by erasing the input field and cycling through dictionaries, if I then press the up and down rocker keys I only see maybe 5 dictionary entries, all of them ones that I've added since starting to use PD2, but I'm pretty sure I've added others as well. All told I have well over 1000 user dictionary entries, the majority of them added in PD1.
Dear Mike,
These are not unique to RC2.
1) In flashcards, I'm getting the wrong definition for some Tuttle words. What happens is that some words have less common pronunciations and the less common pronunciation is not in Tuttle. In flashcards I force the definition from dict with Tuttle at the top followed by ABC and others. When my flashcard is a less common pronunciation of a word whose more common pronunciation is in Tuttle, after the flashcard reveal, I get the character, the less common pronunciation and the more common definition. For instance with 合 the more common pronunciation is he2 and the less common is ge3. With my ge3 flashcard, I get the 合 character, the ge3 pronunciation and the he2 definition from Tuttle. It should just skip Tuttle since it doesn't have the ge3 definition and go on to ABC which has the definition for ge3. This also happens with 个儿, 折, 舍, 间, 陆, 将. In all of these I get the wrong Tuttle definition instead of the correct ABC definition for the less common pronunciations of these characters.
2) In flashcards after the reveal, 3 things are getting stripped off of the ABC definitions that I would like to keep. 2.1) ABC shows that some words can have an 儿 or a 子 by putting them in square brackets like this [儿] or [子]. These get stripped off in flashcards but I would like to see them. 2.2) ABC often shows examples of b.f. words. For example, 衍 has the examples yan3bian4 and fu1yan3 but these are being stripped off in the flashcard reveal. I would like to see the examples. 2.3) ABC puts alternative pronunciations at the bottom of the definition after the words "See Also" but these alternative pronunciations are getting stripped off. I would like to see them.
3) Go to 呼声 in ADS. The audio works fine for the first word, but there are also 4 other long combinations that come up and the audio doesn't work for those. I think it is because right after the characters there is (VERB) or (ADJT) or (PHRASE) or (VERB) and this causes the audio to not play. I also tried 学生 and noticed the same phenomenon. The first few longer combinations play fine but as soon as there are (NOUN) after the characters the audio doesn't play.
Problem 1 should probably be fixed. Problems 2 and 3 are slightly annoying and it would be nice to have them fixed, but not high priorities.
John
These are not unique to RC2.
1) In flashcards, I'm getting the wrong definition for some Tuttle words. What happens is that some words have less common pronunciations and the less common pronunciation is not in Tuttle. In flashcards I force the definition from dict with Tuttle at the top followed by ABC and others. When my flashcard is a less common pronunciation of a word whose more common pronunciation is in Tuttle, after the flashcard reveal, I get the character, the less common pronunciation and the more common definition. For instance with 合 the more common pronunciation is he2 and the less common is ge3. With my ge3 flashcard, I get the 合 character, the ge3 pronunciation and the he2 definition from Tuttle. It should just skip Tuttle since it doesn't have the ge3 definition and go on to ABC which has the definition for ge3. This also happens with 个儿, 折, 舍, 间, 陆, 将. In all of these I get the wrong Tuttle definition instead of the correct ABC definition for the less common pronunciations of these characters.
2) In flashcards after the reveal, 3 things are getting stripped off of the ABC definitions that I would like to keep. 2.1) ABC shows that some words can have an 儿 or a 子 by putting them in square brackets like this [儿] or [子]. These get stripped off in flashcards but I would like to see them. 2.2) ABC often shows examples of b.f. words. For example, 衍 has the examples yan3bian4 and fu1yan3 but these are being stripped off in the flashcard reveal. I would like to see the examples. 2.3) ABC puts alternative pronunciations at the bottom of the definition after the words "See Also" but these alternative pronunciations are getting stripped off. I would like to see them.
3) Go to 呼声 in ADS. The audio works fine for the first word, but there are also 4 other long combinations that come up and the audio doesn't work for those. I think it is because right after the characters there is (VERB) or (ADJT) or (PHRASE) or (VERB) and this causes the audio to not play. I also tried 学生 and noticed the same phenomenon. The first few longer combinations play fine but as soon as there are (NOUN) after the characters the audio doesn't play.
Problem 1 should probably be fixed. Problems 2 and 3 are slightly annoying and it would be nice to have them fixed, but not high priorities.
John
ruisi1 - thanks for the feedback. Shrinking characters in long headwords is certainly an improvement we'd consider for a future release.
hleperlier - great! Yeah, the resizer in the reader is a little flaky - haven't managed to figure out what's wrong with it yet.
koreth - well that's odd... any chance you might have deleted your old user dictionary database accidentally at some point? If you look around in your Palm's memory with FileZ, is there one user dictionary file or are there two? Were you ever prompted to create another user dictionary database?
johnh113 - on #1 the lack of Pinyin specificity in the flashcard dictionary remapper is actually intentional, and here's why. There's a lot of disagreement about Pinyin orthography in the Chinese dictionary community; John DeFrancis addresses this a bit in his introduction to the ABC dictionary, it's not just about spacing / punctuation (which we can ignore) but also about how to represent tones; some dictionaries provide the original / non-transformed tones for each syllable in a word, while others apply tone sandhi transformations where appropriate (so bu4 becomes bu2 etc). The ABC falls in the former category while Tuttle falls in the latter; hence, if we only remapped flashcards when the Pinyin matched exactly, a great many entries would fail to map between ABC and Tuttle (and other dictionaries with differing opinions on Pinyin, though Tuttle's the one that disagrees with ABC most among our current C-E/C-C dictionaries) even when they actually are pointing to the exact same word.
We probably should provide a preference setting for this, but really the best fix would be for us to go back in and manually add non-sandhi-transformed Pinyin to each Tuttle entry. Which would be a lot of work but certainly possible at some point if the Tuttle proves popular. A rough workaround would be to remap multi-character entries when the Pinyin doesn't match but not single-character ones, but that's not exactly an ideal solution either. Pulling the Pinyin from the dictionary entry might help in some respects, but it would still have you studying a different entry than the one you intended to study. But the point is that this is a complicated issue and probably is somewhat beyond what we can cover in a bug fix - we can do something about it in a later release but not in 2.0. In your case I'd probably just steer clear of using Tuttle for flashcards for now - thanks to the unmasked headword characters in example sentences (not to mention the simple length of its entries) it's not all that useful as a flashcard data source anyway.
(the importer, I should add, does require an exact Pinyin match, and won't accept an inexact match unless it can't find any dictionary with an exact one, so this isn't an issue that threatens to seriously screw up anyone's data at least)
#2 you should be able to get at least the example sentences back by re-enabling example sentences in the Flashcards panel in Preferences. Adding the (儿)/(子) would be a bit harder because of the way the flashcard database format is designed (having characters in the headword field that aren't actually part of the headword could create a lot of problems), though i suppose we could consider crawling through the linked dictionary entry and showing them at least after the card is revealed - it would be useful information to have. The links / see alsos are tricky because of the possibility of their giving away the pronunciation for a card - I guess we could hide those whenever the pronunciation is hidden and reveal them when the pronunciation is revealed.
#3 we've actually been considering dropping Adso altogether from the free-dictionary set; CEDICT seems much better suited to our purposes (as it doesn't duplicate entries with the same headword/pronunciation and is rapidly catching up in terms of number of entries), and since it takes an order of magnitude more work to do even a rough cleanup of a new Adso database version than it does with CEDICT/HanDeDict we're not sure if it's worth the effort anymore. Adso might reappear linked to some future version of the document reader (where that extra data would actually be downright useful) but given the direction it's gone and the direction Pleco has gone in the last couple of years I'm not sure if it makes sense as a regular Pleco dictionary database right now.
hleperlier - great! Yeah, the resizer in the reader is a little flaky - haven't managed to figure out what's wrong with it yet.
koreth - well that's odd... any chance you might have deleted your old user dictionary database accidentally at some point? If you look around in your Palm's memory with FileZ, is there one user dictionary file or are there two? Were you ever prompted to create another user dictionary database?
johnh113 - on #1 the lack of Pinyin specificity in the flashcard dictionary remapper is actually intentional, and here's why. There's a lot of disagreement about Pinyin orthography in the Chinese dictionary community; John DeFrancis addresses this a bit in his introduction to the ABC dictionary, it's not just about spacing / punctuation (which we can ignore) but also about how to represent tones; some dictionaries provide the original / non-transformed tones for each syllable in a word, while others apply tone sandhi transformations where appropriate (so bu4 becomes bu2 etc). The ABC falls in the former category while Tuttle falls in the latter; hence, if we only remapped flashcards when the Pinyin matched exactly, a great many entries would fail to map between ABC and Tuttle (and other dictionaries with differing opinions on Pinyin, though Tuttle's the one that disagrees with ABC most among our current C-E/C-C dictionaries) even when they actually are pointing to the exact same word.
We probably should provide a preference setting for this, but really the best fix would be for us to go back in and manually add non-sandhi-transformed Pinyin to each Tuttle entry. Which would be a lot of work but certainly possible at some point if the Tuttle proves popular. A rough workaround would be to remap multi-character entries when the Pinyin doesn't match but not single-character ones, but that's not exactly an ideal solution either. Pulling the Pinyin from the dictionary entry might help in some respects, but it would still have you studying a different entry than the one you intended to study. But the point is that this is a complicated issue and probably is somewhat beyond what we can cover in a bug fix - we can do something about it in a later release but not in 2.0. In your case I'd probably just steer clear of using Tuttle for flashcards for now - thanks to the unmasked headword characters in example sentences (not to mention the simple length of its entries) it's not all that useful as a flashcard data source anyway.
(the importer, I should add, does require an exact Pinyin match, and won't accept an inexact match unless it can't find any dictionary with an exact one, so this isn't an issue that threatens to seriously screw up anyone's data at least)
#2 you should be able to get at least the example sentences back by re-enabling example sentences in the Flashcards panel in Preferences. Adding the (儿)/(子) would be a bit harder because of the way the flashcard database format is designed (having characters in the headword field that aren't actually part of the headword could create a lot of problems), though i suppose we could consider crawling through the linked dictionary entry and showing them at least after the card is revealed - it would be useful information to have. The links / see alsos are tricky because of the possibility of their giving away the pronunciation for a card - I guess we could hide those whenever the pronunciation is hidden and reveal them when the pronunciation is revealed.
#3 we've actually been considering dropping Adso altogether from the free-dictionary set; CEDICT seems much better suited to our purposes (as it doesn't duplicate entries with the same headword/pronunciation and is rapidly catching up in terms of number of entries), and since it takes an order of magnitude more work to do even a rough cleanup of a new Adso database version than it does with CEDICT/HanDeDict we're not sure if it's worth the effort anymore. Adso might reappear linked to some future version of the document reader (where that extra data would actually be downright useful) but given the direction it's gone and the direction Pleco has gone in the last couple of years I'm not sure if it makes sense as a regular Pleco dictionary database right now.
I have to chime in here and say that when looking up computer/technology related terminology, which I do frequently, ADSO still appears to me to have the largest coverage of terms.
I seem to be having the same problem koreth mentions above: user dictionary entries aren't showing up in searches. I can scroll through them all, edit the entries, and they all work in flashcards, but when I input a search (character search), they don't show up at all. I imported them all in 8(c), but they do not contain pinyin information (not sure if that is relevant).
hairyleprechaun - true, it's just difficult to justify the labor to keep it maintained - CEDICT/HanDeDict we can take a database extract from their respective servers and package it up for Pleco in about half an hour, so it's quite feasible to keep those updated, but Adso we haven't managed to streamline that much yet; it's not hard to leave the current Adso database up on our server, but possibly separated from / updated less regularly than the other free dictionary files.
tsibbs - could you send me a copy of your user dictionary database files? Same goes for koreth, actually - having a look at them might make it clearer where the problem is. The missing Pinyin shouldn't be the problem... also, in Manage Dicts your user dictionary is enabled for dictionary searches, right? If you switch to your user dictionary and then enter a search, does that make it work? Have you changed any search/results preferences from their default settings?
tsibbs - could you send me a copy of your user dictionary database files? Same goes for koreth, actually - having a look at them might make it clearer where the problem is. The missing Pinyin shouldn't be the problem... also, in Manage Dicts your user dictionary is enabled for dictionary searches, right? If you switch to your user dictionary and then enter a search, does that make it work? Have you changed any search/results preferences from their default settings?
Any possibility of automating this on your website at some point? I mean create an "update on demand" button that would:
mikelove said:
- Check if new version is available
Fetch new version and apply converstion
Cache the new version and update the web page with basic statistics (word count, date converted, etc) and DIFFERENCE between last conversion and this one (for the RSS reader feature below)
Allow download directly (of course)
Provide an RSS feed for that page (I use an RSS reader to monitor about 250+ sites
Fair enough Mike. While in an ideal world I would prefer the most updated database as possible, I can see that from your end of things that it doesn't really make sense to expend that much effort in cleaning up the data for every update ADSO posts, unless of course you can manage to automate the process as stephanhodges mentions. While I don't know whether or not it is possible to automate the process when new entries are being added that may or may not be consistent with the formatting/style of other entries, as long as I can use the current version of ADSO on later versions of Pleco Dict I will be happy.
For me, another dictionary that includes the computer/technology/science related terminology found in ADSO would also suffice
Darrol
For me, another dictionary that includes the computer/technology/science related terminology found in ADSO would also suffice
Darrol
Ah, that's probably it. Today, after having used PD2 with only minor problems for weeks, I finally deleted my old PD1 install and told ZLauncher to delete all the data files too. Didn't know that PD2 was reading the same files; I assumed that the converter had copied everything over to PD2's databases.
I'll restore my PD1 dictionary file from backup and see what happens.
That's probably worth a warning in the install instructions; it'd be good to have an explanation of what parts of PD1 are and aren't safe to remove after upgrading to PD2.
So much for that theory -- restoring my PD1 dictionary didn't seem to help.
I poked inside the PD2 user dictionary file with a text editor and it only contains the 5 or so entries that show up when I browse through the dictionary using the rocker buttons. Nearly all my user dictionary entries are associated with flashcards; It looks like maybe PD2 used to treat flashcards as an extension of the user dictionary and doesn't do so any more. (Or not; I can't say with 100% absolute certainty that PD2 ever did show me my old user dict entries but I'm pretty sure it did.)
I poked inside the PD2 user dictionary file with a text editor and it only contains the 5 or so entries that show up when I browse through the dictionary using the rocker buttons. Nearly all my user dictionary entries are associated with flashcards; It looks like maybe PD2 used to treat flashcards as an extension of the user dictionary and doesn't do so any more. (Or not; I can't say with 100% absolute certainty that PD2 ever did show me my old user dict entries but I'm pretty sure it did.)
Flashcards are seperate from the user dictionary these days, and their definitions are stored in the PlecoFlash file. You can, though, if you want, use the 'Batch' function in the manage flashcards screen to put all the search results into the user dictionary at once, so it's not hard to rectify . Alternatively, you can export them one by one, if you're so inclined (but the batch function is a lot faster! )
EDIT: They're not added to the user dictionary by default. You need to tick a particular box for that when importing ('Store imported definitions in user dict'), or conciously tap the button when creating a card.
. Alternatively, you can export them one by one, if you're so inclined (but the batch function is a lot faster! )EDIT: They're not added to the user dictionary by default. You need to tick a particular box for that when importing ('Store imported definitions in user dict'), or conciously tap the button when creating a card.
Started to do use PlecoDict 2 for real now and real like what I see. It is a big improvement over Version 1. Especially the document reader, combined with the built-in dictionary and the flashcard system in the background is a very efficient learning environment. I did get a Palm just to be able to use Pleco and did not regret it.
However, i noticed a problem when using the document reader, especially with the ABC, which is what I usually prefer: For characters like 行 and 思, the reader will show only the definition for 'hang', or 'sai', which are rather obscure meanings and by no means the usual meaning of the character. Helpfully, ABC does display a 'see also' link, but unfortunately, in the reader I can't follow the link, so there is no way to look at the correct definition. This seems to be a rather annoying problem to me. It would be great if this could be fixed before the release!
All the best,
Ctop
However, i noticed a problem when using the document reader, especially with the ABC, which is what I usually prefer: For characters like 行 and 思, the reader will show only the definition for 'hang', or 'sai', which are rather obscure meanings and by no means the usual meaning of the character. Helpfully, ABC does display a 'see also' link, but unfortunately, in the reader I can't follow the link, so there is no way to look at the correct definition. This seems to be a rather annoying problem to me. It would be great if this could be fixed before the release!
All the best,
Ctop
In the reader there should be, along the bottom of the pop-up window, a bar with an icon denoting the dictionary, an icon for playing the audio, adding to flashcards, and a down arrow (and done). Hit the down arrow, and it should take you to the next item in the search results.
E.g. Going to the 21st Century Dictionary, searching for OK, sending it to the reader, selecting 行, and switching to the ABC dictionary gives me háng as the first result. Hitting down once, I get hàng, then heng, then finally xíng. I assume that's the behaviour you wanted?
E.g. Going to the 21st Century Dictionary, searching for OK, sending it to the reader, selecting 行, and switching to the ABC dictionary gives me háng as the first result. Hitting down once, I get hàng, then heng, then finally xíng. I assume that's the behaviour you wanted?
Re. could you send me a copy of your user dictionary database files? Sent via email
Re. in Manage Dicts your user dictionary is enabled for dictionary searches, right? It is enabled
Re. If you switch to your user dictionary and then enter a search, does that make it work? No, still not working; it just switches to ABC and/or oxford
Re. Have you changed any search/results preferences from their default settings? No, everything is "stock"
Thanks,
tim
Re. in Manage Dicts your user dictionary is enabled for dictionary searches, right? It is enabled
Re. If you switch to your user dictionary and then enter a search, does that make it work? No, still not working; it just switches to ABC and/or oxford
Re. Have you changed any search/results preferences from their default settings? No, everything is "stock"
Thanks,
tim
stephanhodges - the dictionary converter for 2.0 really isn't in any kind of shape to put on a production system - the full-text indexer especially was designed in a lazy / not particularly memory-efficient way, so while the results it generates are fine I wouldn't want to actually run it on anything but a development system; putting it on our web server is just asking for trouble. (also part of why we're not releasing a MakePlecoDict in our proprietary format but rather planning on one for SQLite)
hairyleprechaun - OK. We'll certainly make sure some version of Adso is available whenever we change our proprietary dictionary database format (which we may actually be doing again as soon as 2.1).
koreth - oh, no, sorry, I meant the old 2.0 user dictionary; 2.0 doesn't care about 1.0's dictionary file, but I thought you might have accidentally deleted one from 2.0 at some point while upgrading. The problem as ipsi says is that imported data is stored in the flashcard database by default; a large number of people were complaining about user dictionary entries they created specifically for flashcards getting in the way of regular dictionary searches, and between that and the fact that 2.0 starts up a bit faster without user dictionaries (which it only creates when you add entries to them) it seemed best to store those custom definitions etc in the actual flashcards. But that Batch "Custom -> User" command should get all of those entries transferred to a user dictionary.
tsibbs - thanks for the file. I'm not seeing anything wrong in this, unfortunately - could you send me the actual user dictionary database file as well? You can get that by using PlecoMover to copy PlecoCUser???????? to internal memory, HotSyncing, then retrieving the file from your Palm Desktop backup folder - see the user manual for more on that.
hairyleprechaun - OK. We'll certainly make sure some version of Adso is available whenever we change our proprietary dictionary database format (which we may actually be doing again as soon as 2.1).
koreth - oh, no, sorry, I meant the old 2.0 user dictionary; 2.0 doesn't care about 1.0's dictionary file, but I thought you might have accidentally deleted one from 2.0 at some point while upgrading. The problem as ipsi says is that imported data is stored in the flashcard database by default; a large number of people were complaining about user dictionary entries they created specifically for flashcards getting in the way of regular dictionary searches, and between that and the fact that 2.0 starts up a bit faster without user dictionaries (which it only creates when you add entries to them) it seemed best to store those custom definitions etc in the actual flashcards. But that Batch "Custom -> User" command should get all of those entries transferred to a user dictionary.
tsibbs - thanks for the file. I'm not seeing anything wrong in this, unfortunately - could you send me the actual user dictionary database file as well? You can get that by using PlecoMover to copy PlecoCUser???????? to internal memory, HotSyncing, then retrieving the file from your Palm Desktop backup folder - see the user manual for more on that.
Okay, then I guess I just never tried to use my (flashcard-centric) user database entries in 2.0 until recently. And the entries I do see are probably ones I created using the 1.0-ish workflow of creating a new dictionary entry then hitting "+" to add them to a flashcard list.
So if one doesn't want a user dictionary entry, the right procedure is to make a flashcard from one of the built-in dictionaries then go to the flashcard interface, search for the card, edit it, switch it to a custom definition, and enter the dictionary data there, correct?
So if one doesn't want a user dictionary entry, the right procedure is to make a flashcard from one of the built-in dictionaries then go to the flashcard interface, search for the card, edit it, switch it to a custom definition, and enter the dictionary data there, correct?