in reply to Re^7: Chinese font problem in Perl/Tk
in thread Chinese fonts problem in Perl/Tk

No-go with ActivePerl 5.8.6 on Redhat Linux, Solaris 8, nor Solaris 2.6 for me.

Same as with 5.8.4, some characters displayed codes instead of the character.

"uname -srv" for Redhat Linux: Linux 2.4.18-24.7.x_3.5G #1 Mon Jul 26 13:49:55 CDT 2004
"uname -srv" for Solaris 8: SunOS 5.8 Generic_108528-14
"uname -srv" for Solaris 6: SunOS 5.6 Generic_105181-33

Replies are listed 'Best First'.
Re^9: Chinese font problem in Perl/Tk
by Courage (Parson) on Jan 07, 2005 at 23:27 UTC
    Hmmm.

    I can't reproduce.
    But I can beleive you fall to some kind of configuration problems... may be your OS installations lack some internationalizatio(18)n setting? something with underlying C library (as usual, many perl bugs refer to underlying C lib)

    as zentara said you need to post it as bug report, I beleive to bug tracking system. PerlTk mailing lists also wait for your posting :)

    Once again, try running my code of pure-Tk (under "wish") and give me a screenshot on your system(s)...
    We should start digging from here

    In case putting screenshots is pain for you, I can provide you with proper place to put.

    Best regards,
    Courage, the Cowardly Dog

      Thanks for all your time, Courage.

      A friend of mine in Singapore tried my script on a Solaris 8 box and reported seeing the same problem.

      What OS's have you been able to try on?

      I got your Tk script to run on Redhat Linux (Linux 2.4.18-24.7.x_3.5G #1 Mon Jul 26 13:49:55 CDT 2004) and similar problem -- three of the Chinese characters appeared as \uHHHH codes. Where can I put a screenshot? In the meantime, here's screenshot in my spot.

      Yes, I'll get this issue posted -- probably from home so I can use my handle rather than the office id.
        Given a fact of failing pure-Tk script, looks like you should ask Tcl people, but asking from both sides (including perlTk) is better

        My estimate is 97% probability to fact that the problem lies in underlying C library, which does not recognizes some characters, due to some locale setting on that OS.

        I advice for you to ask in those lists: ActiveTcl@listserv.ActiveState.com, tcl-core@lists.sourceforge.net for pure-Tk code, they probably will give you some OS-level advices

        OSes I can try on are Linux, Win2000, WinXP, and all they have non-english default locale, which helps with unicode.
        Due to my non-native English, as you already could notice...

        addition I think failing with only few characters from Unicode means you do not have recent enough version. Can you check somehow which Unicode version your system uses? Also which version of Tcl/Tk do you hav eon those systems?

        For tcl version, feed puts $tcl_version to tclsh

        Best regards,
        Courage, the Cowardly Dog