OpenIndiana Distribution

As many Russian users have noted and reported recently (see bugs #1705, #1771, #1782 for example), there was a timezone change relevant for Russian users, and they (we) were vocal enough to raise the issue.
Actually, the problem is bigger than those point-and-shoot patches. The source "tzdata" definitions change several times a year and this influences many users in different parts of the world, so I suggest splitting timezone information into a separate package which would be updated as soon as new definitions are published (maintained by IANA as of recently), as grabbed from: http://www.iana.org/time-zones
"tzcode" of utilities such as "zic" can be fetched there as well, and maintained as another separate package.

Other than this, it would also be nice to streamline updates of timezone information in JRE/JDK as in Sun/Oracle TimeZone Updater Tool (i.e. use Oracle's "tzupdate.jar", or make and maintain a similar one to use current timezone definitions): http://www.oracle.com/technetwork/java/javase/downloads/index.html#timezone

Some other subsystems might define their timezone as well, notably MySQL, but these definitions can be sourced from current UNIX OS image's timezone definitions, see: http://dev.mysql.com/doc/refman/5.1/en/time-zone-support.html

Following the update of UNIX/JVM/DBMS/etc. timezone info packages, their 'users' should get restarted (viz. SMF dependencies). For example, there might be a stub service for tzinfo as an "optional_all" dependency for SMF services of application servers. If it is enabled, and the TZ definitions are updated, this service gets "restarted" causing restarts of other services (Tomcat, MySQL, etc.); if the administrator doesn't want this behavior - he keeps the stub service disabled.