Re: UTF8 dn_normalize() and when to load ucdata

["Kurt D. Zeilenga" <Kurt@OpenLDAP.org>]

At 05:01 PM 10/26/00 +0200, Stig Venås wrote:
>Since my new dn_normalize() uses the unicode functions, the ucdata
>must be loaded before the first call to dn_normalize(). Currently
>dn_normalize() is called first time in read_config(), while the
>ucdata is loaded afterwards in schema_prep().

And, of course, we don't know ucdata-path until after we read
the configuration file.

The easy fix is to use an command line argument and/or environment
variable to determine ucdata-path and to load ucdata fairly early
in the startup sequence.

Kurt