Main Inclusion Report for sourcepackage


  1. Availability:

  2. Rationale:

    • Input method is required by many Asia users, like CJK people and Indian. It's a critical issue for them because they can't input their native language without it.
    • SCIM supports many laungages via scim-pinyin, scim-hangul, scim-anthy scim-tables, scim-m17n and so on.
    • Although there are a few candidates for multilingual input method framework such as IIIMF or uim, there are some critical problems in it.
      • uim lacks well supported chinese input methods.
      • IIIMF lacks well supported OSS input methods, lacks comfortable GUI. Some people pointed out security issues in its fundamental design.
    • See also MainInclusionReportScimAnthy

    • See also MainInclusionReportScimHangul

    • See also MainInclusionReportScimPinyin

    • See also MainInclusionReportScimChewing

    • See also MainInclusionReportScimTables

  3. Security:

    • No security histroy.
  4. Quality assurance:

    • SCIM is well maintained by Jame Su who is employed in Novell.
    • Fedora, SUSE and Mandriva shipped with SCIM, and no critical bugs has been found in them.
    • There are some binary compatibility issues related with libstdc++. Since libscim will be linked with all Gtk+ (and patched version of Qt) applications via its input method module system, and SCIM is written in C++, an application which is compiled against different version of C++ with SCIM will cause clash. This problem isn't reproduced when you use SCIM via XIM because it is a protocol based communication, not shared link.
  5. Standards compliance:

    • The package meets debian-policy and FHS.
  6. Dependencies:

    • All in main.
    • SCIM is just only a framework. Also some major IMEngines (language plugins for SCIM) should be included into main to realize true multilingual input. They are all in universe.


MartinPitt: approved

MainInclusionReportscim (last edited 2008-08-06 16:34:58 by localhost)