Problems using SIL Converters in Adapt It

[AdaptIt.Bob]

The following is a question about SIL Converters:

Today, I was setting up Russian AdaptIt for Unicode on a local
coworker's machine, and was transliterating Genesis 1 from a non-
western script to Cyrillic (reversing a bidi mapping), and I got
several errors, and odd behavior.

The SIL converter problems were as follows:

#1. ? character at the beginning of the TECkit Unicode editor. I saw
this behavior a month ago at GC on a colleague's machine.
#2. error message: "there is a problem...this might be due to corrupt
memory..."
#3. Some entire words were not converting at all. There would be a
blank box under the word that was supposed to have been
transliterated. Then, with some words, the consonants would
transliterate but not the vowels.

The frustrating thing is, that the same set up under AdaptIt 3.3.1
with EC 2.6.1 on my XP machine is working great. The AdaptIt version
was 3.6.2, EC 2.6.1. The difference I noticed was that her machine was
running Office 2007.

This is one possible answer to the problem:

#1 sounds like the problem of having Paratext installed whereby it
installs an old version of the TECkit DLLs in a place that trumps the
install location for SILConverters.

Search and destroy all copies of the TECkit DLLs (i.e. TECkit_x86.dll
and TECkit_Compiler_x86.dll) that are not in the “C:\Program Files
\Common Files\SIL” folder.