Learn how easy it is to sync an existing GitHub or Google Code repo to a SourceForge project! See Demo

Close

Notepad++ 5.8.1 released

Don HO
2010-09-28
2012-11-13
<< < 1 2 (Page 2 of 2)
  • cchris
    cchris
    2010-10-11

    So the only apparent difference about this is the locale of the OS, not the OS proper - I may have trouble setting it to traditional chinese. Can anyone test and report?

    CChris

     
  • Sultan
    Sultan
    2010-10-11

    versions 5.8 and 5.8.1 have a problem compared to 5.7.1 when using the "Find all in current document". In 5.7.1, the selected search result stay highlighted in yellow when you double click on it while this functionality is not there in the later versions. It's important since when you have many results listed you need to know which one you are seeing in the main page.

     
  • Dragon Chuang
    Dragon Chuang
    2010-10-11

    @cchris
    I can reproduce this issue on :
    Windows XP 32 bit SP3 (traditional chinese)
    Windows XP 32 bit SP3 (simple chinese)
    Windows XP 32 bit SP3 (japanese)

    I can not reproduce this issue on :
    Windows XP 32 bit SP3 (english)

    Thank you :)

     
  • cchris
    cchris
    2010-10-11

    @Drajuang:
    I'll check later from home, but can you determine whether the file encoding matters? ANSI 437 vs ANSI GB2312 vs UTF-8 vs UCS2-LE? Perhaps that's where the actual difference lies.

    @Sultanian:
    I'll check into this within a few hours.

    CChris

     
  • cchris
    cchris
    2010-10-11

    @Sultanian:
    I just downloaded a fresh copy of N++ v5.8.1, and, if I
    * select a word
    * hit Ctrl+F
    * choose Find all in current doc

    then both v5.7.1 and 5.8.1 keep the initially selected word selected. This is on XP32 bits fr_FR on an ANSI 850 encoded file. How is your setup different from this?

    CChris

     
  • cchris
    cchris
    2010-10-11

    @Drajuang:
    With french as system locale, I tried to reproduce your issue by encoding the document to UCS2-LE and UTF-8. To no avail.

    @Don:
    Can you reproduce the problem?

    CChris

     
  • Dragon Chuang
    Dragon Chuang
    2010-10-12

    @cchris
    OS: Windows 7 32 bit Traditional Chinese
    Notepad++ Version: 5.8.1

    I can reproduce this issue on below encoding:
    ANSI

    I can not reproduce this issue on below encoding:
    ANSI 437 (OEM-US ?)
    ANSI GB2312 (Simplified Chinese)
    ANSI BIG5 (Traditional Chinese)
    ANSI Shift-JIS (Japanese)
    UTF-8
    UCS2-LE

    Thank you :)

     
  • Dragon Chuang
    Dragon Chuang
    2010-10-12

    I have a suggest, maybe we could provide more encoding choice in 'Settings / Preferences / New Document' for Asia NotePad++ user or non-english user :)

    just like this imagined image:

    Thank you :)

     
  • cchris
    cchris
    2010-10-12

    You may want to vote for http://sourceforge.net/apps/ideatorrent/notepad-plus/ideatorrent/idea/220/ ; this may help this shortcoming to be addressed.

    The pattern you are reporting suggests to me that something is wrong with the basic ANSI mode when system locale is far eastern (traditional chinese or japanese). I simply don't know what could be going on - suspecting a Scintilla bug actually. Can anyone able to test in such settings help?

    CChris

     
  • Dragon Chuang
    Dragon Chuang
    2010-10-12

    @CChris

    I voted solution.
    I don't know what could be going on, too. :/
    at least , I can around the issue by change encoding to BIG5 now.

    Thank you very much :)

     
  • cchris
    cchris
    2010-10-12

    Could you (7-)zip the file with the three lines and email it to me at quasar850 hat rediffmail doubt com? This way I can run a binary compare and see whether it is something in the file encoding that goes wrong. If not, it may mean that Scintilla's code page in ANSI encoding is wrong someow. I asked Don whether he could reproduce the issue and investigate.

    CChris

     
  • Dragon Chuang
    Dragon Chuang
    2010-10-13

    I did a binary compare.


    result was binary same.

    I think Scintilla's 'Find Previous' behaviour in ANSI encoding is wrong somehow.

    (note: I could not reproduce this issue on  Notepad++ v5.7.1)

     
  • Dragon Chuang
    Dragon Chuang
    2010-10-13

    To correct:
    (note: I could not reproduce this issue on Notepad++ v5.7.1 v5.7)

    I could not reproduce this issue on Notepad++ v5.7
    But, It could be reproduced on v5.7.1 RC2 and 5.8.2.

     
  • cchris
    cchris
    2010-10-13

    Ok, bug located. It is a problem with the newer Scintilla v2.21, which explains why 5.7 didn't have the bug.

    Basically, what happens is that, on a lexer change, Scintilla resets the code page to plain ANSI. When the system locale is SBCS, all is well. When it is not, like for all ideographic scripts, there is a mismatch between the system locale and Scintilla, causing the issue you are experiencing.

    The bug has been reported to the Scintilla project. N++ will fix (about) as soon as they do.

    CChris

     
  • Don HO
    Don HO
    2010-10-13

    Scintilla resets the code page to plain ANSI. When the system locale is SBCS, all is well. When it is not, like for all ideographic scripts, there is a mismatch between the system locale and Scintilla, causing the issue you are experiencing.

    Well, it's not what it does happen exactly.
    The searching previous bug comes with sciintilla 2.21, while the codepage is set to 950/936/932/949 (CJK: Chinese/Japanese/Korean).
    If the codepage is set to 0, everything is OK : you can set CAML as language in Notepad++ under Chinese XP, then you will see the search previous works fine (Notepad++ set codepage as 0 for CAML due to the crash under CJK enviroment with this lexer).

    The different codepages are set according to environments CJK by Notepad++ due to the nature of MBCS - it allows Scintilla to process each ideogram as one unit.

    I will report this bug to Scintilla project. This bug will be fixed in Notepad++ as soon as it is fixed in Scintilla.

    Don

     
<< < 1 2 (Page 2 of 2)