Remove issue cache

As the issue navigator has been refactored and does not hit the database for many issues, the need for issue cache is greatly reduced. The issue cache also is one of the main sources of memory consumptionin JIRA, especially if issues have long descriptsions and if Oracle 10g drivers are used (as they seem to allocate strings in 32K buffers at all times).

It appears that as the get() method of the UtilCache (used by the issue cache) is not synchronised but manipulates the underlying collections to provide LRU functionality. As the underlying collections are not synchronised a race condition could occur.

Removing the issue cache should help avoid all these problems and will not decrease JIRA's performance.