2.0 Not Quite Ready On June 5th

mikelove

皇帝
Staff member
I imagine a lot of you have already guessed this, but just to state it officially here: unfortunately it's going to take at least a few more weeks for us to get 2.0 finished, so the June 5th release date mentioned in my announcement e-mail earlier is going to be off by a bit.

We're putting the finishing touches on the next 2.0 beta testing release (Beta 5) now and it looks really good - every feature does in fact seem to work, there are certainly still a lot of bugs but any given checkbox/button/etc should now do something pretty similar to what it's supposed to do. We've also cleaned up a few areas of the interface that were still problematic, like the Manage Cards screen for flashcards (which has gone through 3 or 4 revisions but which we're finally happy with).

Beta 5 should be out within a week or so, and we're planning to accompany it with an actual installer and a very basic instruction manual (basically just a list of each option/command/etc and what it does) so it should be usable by a greater number of people than the earlier betas. And since all we have left now is bug-fixing, the betas should come pretty rapidly after we release Beta 5 - hopefully one every week or so until the product is done.

Sorry about the continuing delays - this is the last time we're ever going to try to do this much in one release; in hindsight we would have been much better off rolling out the new features gradually, one release every 6 months or so, rather than rewriting everything from scratch and making you all wait 2 years for a new version. So at least we've learned something from the experience :)


Oh, and one little bit of good news for Mac users: we've just licensed a handwriting recognizer that works on Mac OS, so we're tentatively planning a simultaneous release of the desktop version of Pleco (the real one, not the "preview") on Mac OS and Windows. It'll require an Intel-based Mac, unfortunately, but if you've wanted an excuse to buy that shiny new MacBook Air this may give you one. Hopefully ready by the end of '08, though given the main subject of this posting I'm not sure if you should really believe that.
 

sfrrr

状元
You might consider following the SBSH,net example: Upgrade every month or so. Keeps people like us forum members coming back to the forum and, presumably offering presumably helpful suggestions or secondary uses for various features.

OTOH, It takes an incredible amount of work to keep up with the forums and be available to your users all the time.

And, besides which, you've gotten the full attention of the die-hards, that's for sure.

For the beta testers, a beta a week will certainly keep us on the hook. :roll:

Sandra
 

mikelove

皇帝
Staff member
The first few 2.0 updates probably will be monthly, or something close to that, since there'll undoubtedly be new bugs discovered after the finished version is released and the number of people using it expands 100-fold. Not to mention at least a few areas of the interface that have to be tweaked after they turn out to be massively confusing for people :D
 

ldolse

状元
Hey Mike, How soon do we get a MakePlecodict binary, also the 1.0 flashcard exporter? It's another area I'm intereted in testing. Lastly, has the database format stabilized enough that we can expect some of the free dictionaries to get 2.0 dbs soon? The missing dictionaries is one of the main reasons I've still been using 1.0 more often.
 

mikelove

皇帝
Staff member
MakePlecoDict won't be out until a bit after the rest of 2.0. We're still debating how to handle that, actually - we've been seriously considering having MakePlecoDict generate databases in our new SQL-based user dictionary format instead of the one we use for our own databases; we've successfully imported CEDICT (which is 35,000 entries) to a user dictionary via the handheld software's import function and aside from some slowness in browsing through long lists of entries (which we think we know how to fix) it seems to be quite usable. The big advantage of this is that it would let us make MakePlecoDict open-source, which along with allowing people to port it to Linux/Mac/etc would also let it potentially integrate more closely with other software (a Pleco dictionary exporter for ZDT, for example).

The 1.0 exporter should be out sooner than that - the big issue in that is creating mapping tables that will convert record IDs from the old versions of the ABC and Oxford dictionaries to the new ones. We may end up doing that with only the ABC, actually, it's really tricky with the Oxford (since there aren't any record IDs in the original source data from them) and there are only two dozen or so entries in the Oxford which aren't uniquely specified by their characters plus Pinyin - since the importer now prompts you in ambiguous cases it doesn't seem like people are likely to be too bothered by that.

The free dictionaries haven't really been a question of our database format stabilizing but rather of our getting around to converting those dictionaries to the new (and much more complicated) input file format that our database-making utility uses - once the files are in that format it's usually trivial to rebuild them if something about the database format changes.
 

ipsi

状元
Glad to hear it's still chugging away :).

Are you planning on releasing a Linux version of PlecoDesktop? I've just picked up an EEE, and it'd be nice having Pleco on there :).

While being able to use it on WINE would be ok, I'd much rather have a native version, using wxWidgets or the like.
 

mikelove

皇帝
Staff member
A native Linux version will depend on how the Mac/Windows ones sell - both of those ports are comparatively easy since they can be based on the iPhone and Windows Mobile versions.
 

ipsi

状元
Hmmm... Tough choice: Buy Windows version to support Pleco, or hold out (and pray) for Linux version... Or, I guess, I could buy them both.
 

mikelove

皇帝
Staff member
Well hopefully if we did release a Linux version we'd be able to allow free platform switches, as we have on Palm/Windows Mobile. There's no way to know that for certain until the licenses are negotiated but there's a pretty good chance anyway.
 

mikelove

皇帝
Staff member
Good question, hence the "or so" part. We really want people to start using flashcards - virtually no bug reports or anything else have come in on them - so we're taking a little extra time to get the 1.0-to-2.0 converter working well and try to make sure the database formats and dictionary entry IDs are solid/fixed.
 

sych

榜眼
Mike, I'm waiting for you to tell us that the dictionary IDs and DB formats, etc, have been finalised, before I start using flashcards on a Beta. So far the continual warnings about all of that not being finalised (and hence the beta flashcards data not being usable in the final version) has put me off. But I'm anxious to give it a go, as soon as this is sorted out.
 

mikelove

皇帝
Staff member
Yeah, that should finally be taken care of in B5. Though of course that doesn't eliminate the possibility of a bug killing your flashcard data (or causing it to be recorded incorrectly), particularly given the lack of widespread public testing of the flashcard system so far, so there'll still be some warnings attached, but with regular backups it should be pretty safe, and thus far it's performed extremely well in our testing - importing 30,000 flashcards does take a while (maybe 5-10x as long as with the old version if they're exported from 1.0 and have category/score/etc data attached), but it doesn't crash and the resulting databases seem to be clean and uncorrupted.
 

weight

举人
Mike, I'm one of those nutty guys doing sentences ... Maybe 1500 so far in the user database. I however ignored your call earlier to send in sample databases for you to test. Are you fairly confident there shouldn't be any problems with the conversion, and if your not confident do you want to see my data?

Thanks

mikelove said:
Yeah, that should finally be taken care of in B5. Though of course that doesn't eliminate the possibility of a bug killing your flashcard data (or causing it to be recorded incorrectly), particularly given the lack of widespread public testing of the flashcard system so far, so there'll still be some warnings attached, but with regular backups it should be pretty safe, and thus far it's performed extremely well in our testing - importing 30,000 flashcards does take a while (maybe 5-10x as long as with the old version if they're exported from 1.0 and have category/score/etc data attached), but it doesn't crash and the resulting databases seem to be clean and uncorrupted.
 

mikelove

皇帝
Staff member
Pretty confident, yeah - the files we've already gotten with sentences have worked fine, and there are no longer any hard limits on the length of user dictionary entries thanks to SQLite's easygoing attitude towards record sizes - in theory if you wanted to create an entry with an entire chapter of The Analects as the headword I don't think the exporter or importer would stop you.
 

sych

榜眼
Mike, if you're still looking for flashcard databases to test, let me know and I'll send mine. I have about 4000 flashcards in my DB, from ABC, Oxford, my user dictionary, and an early version of CEDICT (I haven't upgraded CEDICT for quite a while, because my CEDICT-based flashcards are all from the original version).
 

mikelove

皇帝
Staff member
I think we're OK - in a couple of days you'll be able to test them yourself :)

(incidentally, we just added a feature to the converter that will embed definitions from free dictionaries like CEDICT in the export file, so they'll be available even if we haven't converted CEDICT yet)
 

radioman

状元
Mike,

For version 1 flashcard "words" that are mapped to Oxford definitions (as an example), I understand you cannot remap in version 1 to ABC (as an example). When I import them into version 2, will I then have the option to remap the definitions them as desired? Any insights appreciated.

Ed.
 

mikelove

皇帝
Staff member
Yes, there's a command built-in to remap a list of cards to a particular dictionary, so you could do a search for all Oxford cards and use that command to remap them to ABC.
 
Top