ICU has become one of the most popular open source Unicode libraries around. This presentation first provide a brief overview of the services available in ICU, then discusses a summary of the improvements in ICU since the last San Jose Unicode conference, including the following topics,

• Unicode 3.2 update
• Supplementary character & Unicode property support
• Versioning, memory usage and modularization
• Additional platform support
• Currency (inc. Euro) handling
• Collation & string search function & performance
• Normalization function & performance
• Character conversion improvements (inc. tools)
• Character/word/line/sentence segmentation improvements

In additional, we will also look at the planned new features such as:

• POSIX style compatibility interface
• Runtime plug-in interface for user linguistic analysis resources
• Flexible data handling for user data

Finally, we will examine some misconceptions people have about ICU as well as technical issues regarding the usage of ICU in existing applications:

• ICU distribution is too big.
• Other products are shipping ICU binaries. There will be problems when dealing with multiple versions of ICU.
• ICU has its own Unicode data type and locale object. It's not compatible with other internationalization libraries (e.g. POSIX)
• There is no interface for printing or reading Unicode characters in ICU.
• ICU can't be used on the less commonly used platforms.
• ICU is open-source, therefore poor-quality / unreliable.