+ Reply to Thread
Results 1 to 3 of 3

Thread: Severe regression in handling of Japanese verbs.

  1. #1

    Severe regression in handling of Japanese verbs.

    Until a couple of months ago, I could generally count on the most common conjugations of all but the most obscure verbs correctly converting to the appropriate kanji + okurigana. If they were not the default conversion, they would at least show up in the list of candidates, allowing me to select them.

    This is no longer the case. Even when tapping individual letters (as opposed to swiping), basic forms of many verbs do not show up at all in the candidate list. I end up having to hunt around for some other combination of characters that will provide the kanji I'm looking for, and then add the okurigana manually.

    The following is an extremely non-exhaustive list of examples. I've typed in the romaji one at a time to ensure accuracy. None of the following errors are due to inaccurate swiping.

    Format: [Kana typed in] : [Expected conversion, not found in candidate list] : [examples of candidate conversions that are offered]


    And many, many others. These aren't a handful of problems I've run across over several hours of use. This problem is a frequent and unavoidable annoyance when using QWERTY Japanese input under Swype I don't know if this problem occurs with other layouts.

    I don't know if verb conjugations are generated algorithmically or all stored as individual entries in the dictionary, but if it's the former, something is horribly wrong with the algorithm, and if it's the latter, it looks like somebody just randomly deleted a bunch of entries from the built-in dictionary.
    Last edited by brberg; 05-10-2017 at 7:48 AM.

  2. #2
    I am seeing exactly the same difficulty.
    When I typed "わかる" (wakaru), I wanted to see "分かる", but recent Swype got me "分", "分かり"(wakari instead of wakaru), and irrelevant choices.
    It looks Swype did not perform well if Hiragana input/ conversion result is expected to be composed of/broken down into single Kanji character and Okurigana, and Swype discards Okurigana part in this case, resulting we get only single Kanji character without followed by Okurigana.

    When I typed my friend's family name "かなむら" (kanamura), I wanted to have "金村". But Swype got me "金" only and missed "村" or "むら" part. I suspect this is because of the same reason/cause.

    I did not see the difficulty with Swype almost a year ago or before.

  3. #3
    How is this still an issue after nearly two months? I get that Japanese support is not a high priority for you, and that it's one of the harder languages to implement, but if you're not going to make it better, can you at least a) stop breaking it, and b) when you do break it, roll it back to a version that's merely mediocre instead of borderline unusable?

+ Reply to Thread

Similar Threads

  1. SPANISH Second person verbs are missing!
    By kepatronik in forum Non-English Languages
    Replies: 1
    Last Post: 03-05-2014, 2:02 PM
  2. Possible bad Unicode handling
    By daniele_athome in forum Android Bug Reports
    Replies: 0
    Last Post: 06-08-2013, 5:11 AM
  3. Severe incompatibility bug with HTC apps
    By 6rabbit in forum Android Bug Reports
    Replies: 16
    Last Post: 05-27-2013, 7:58 AM
  4. Feature regression between 3.26.x and 1.x
    By pro-logic in forum Android Feature Requests
    Replies: 26
    Last Post: 10-07-2012, 9:17 PM
  5. Version 3.26.92D.37604 regression.
    By spbrereton in forum Android Bug Reports
    Replies: 1
    Last Post: 02-07-2012, 10:54 AM


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts