Known Issues for Internationalization

Multiple-writer functionality fails for IQ servers with Chinese or Japanese locales.

To use multiple writers, Sybase IQ InfoPrimer must have execute permission for the IQ server sp_iqstatistics stored procedure. For IQ servers that are set up for Chinese and Japanese locales, running sp_iqstatus returns an error similar to:

Could not execute statement.
Syntax error near '2010' on line 1
SQLCODE=-131, ODBC 3 State="42000"
Line 1, column 1

Workaround for Linux or UNIX:

Workaround for Windows:

After you use these workarounds, some strings in the .iqmsg files and the output of sp_iqmpxinfo, sp_iqstatistics, and sp_iqstatus will be in English instead of Chinese or Japanese.

A login attempt to Adaptive Server Enterprise may fail if the specified user name or password contains multibyte characters.

Workaround: Reconfigure the Client Charset database option:

Unencoded multibyte characters cannot be used in XML datasource.

The XML file specified as a datasource for the Text Data Provider and XML via SQL Data Provider components cannot contain multibyte characters that are not percent encoded.

Workaround: You must convert any multibyte characters into UTF-8-encoded byte sequences, and then convert each byte into its hexadecimal representation in the %HH format (percent-encoding).

For example, this URL contains three Japanese characters:

01100101  11100101
01100111  00101100
10011000  11011111

Converted by byte, the percent-encodings are %65%E5, %67%2C, and %98%DF. The URL entered should therefore be:

ftp://myhost/%65%E5%67%2C%98%DF.txt

Grid engine cannot detect IPv6.

If you have installed Sybase IQ InfoPrimer Development on Windows Vista and enabled IPv6, the grid engine may not detect IPv6. As a result, Sybase IQ InfoPrimer Development may not be able to connect to the grid engine.

Project execution fails if the locales.dat file is not configured.

On UNIX and Linux, if you execute a project without providing a value representing the language for your platform in the $IQIPDinstall/ocs/locales/locales.dat file, project execution may fail.

Workaround: Add the correct language setting in the locales.dat file. For example, if you are using a Solaris machine, and your locale is ja_JP.utf8, open the locales.dat file and search for the [sun_svr4] section, then set:

locale = ja_JP.utf8, japanese, utf8

IPv6-only environment does not work with Microsoft Vista.

By default, Microsoft Vista supports IPv6. However, if you are using Vista in an IPv6–only network environment, you see this error:

ERROR : Cannot create socket. An address incompatible
with the requested protocol was used.

This is mainly due to an incompatibility between the JDK and Windows Vista. However, the same error does not occur if you are working in a mixed (dual-stack) environment of IPv4 and IPv6.

Workaround: None.

Accessing Chinese file or folder names on some UNIX platforms.

On most UNIX platforms, the LANG environment variable is set to "zh_CN.GBK" or "zh_CN.gbk" to support Simplified Chinese characters in the file or folder names. However, on some UNIX platforms, you may encounter errors unless you set the environment variable LANG to the locale supported on your machine.

You can set the environment variable LANG to zh.GBK or zh_CN.GBK to support Simplified Chinese.

If you cannot see the list of locales, install the required language package on your machine.

Source files with byte-order mark are parsed incorrectly.

If you are using the Fixed by Bytes property to parse your file, make sure the source file does not include the byte-order mark. If it does, the file will be incorrectly parsed.

Workaround: Use a text editor to remove the byte-order mark from the source file before parsing it.

Japanese character does not convert correctly from Oracle 10g.

The Japanese character '&#65374', does not convert correctly from the source Oracle 10g database if you use the Oracle 10g or 11g native interfaces, or the Oracle 10g or 11g ODBC interfaces in Sybase IQ InfoPrimer Development on Windows XP.

Workaround: In Sybase IQ InfoPrimer Development, use the Microsoft ODBC driver for Oracle that ships with Windows XP and set the NLS_LANG environment variable for locale to “AMERICAN_AMERICA.JA16SJIS” in the registry of your Sybase IQ InfoPrimer Development machine running on Windows XP.

Insert location component cannot load Unicode data from SQL Anywhere 10.

The IQ Loader DB via Insert Location component cannot load non-Western and multibyte character data from SQL Anywhere 10 using the nvarchar datatype.

Workaround: None.

Cannot enter multibyte characters in password fields.

You cannot directly type in non-Western or multibyte characters in some IQ InfoPrimer password fields; for example, in the Database Configuration window for components. You can type only ASCII characters directly into password fields.

Workaround: You can paste non-Western or multibyte character passwords directly into the password field.

Insert Location Component Cannot Load Unicode Data from Microsoft SQL Server and IBM DB2.

The IQ Loader DB via Insert Location component cannot load non-Western and multibyte character data from Microsoft SQL Server and IBM DB2 to Sybase IQ using Sybase Enterprise Connect™ Data Access (ECDA).

Workaround: None.