"Noah L. Meyerhans" wrote:
> No, ASCII is ASCII, no matter what platform you're on. The only
> difference is that on DOS/Windows, ASCII lines are terminated with 2
> characters, the newline and carriage return. On Linux/Unix, only one of
> these is used. You can use the dos2unix program (from the sysutils
> package) to strip off the extra characters at the end of the lines.
>
> Are you certain the format of these files is something that Minivend knows
> how to handle? It might expect a totally different layout than what
> you're giving it in the text files.
>
> noah
_________________________________________________________
The database sample that is distributed in the demo Minivend system is a
Unix, ascii text file, with tab delimiters. The search functions are
performed by glimpse. All of the people on the Minivend Mailing list say
that this is the correct format to use, though the field delimiters may
be of several different types. The question about DOS ascii vs. Unix
ascii was raised in a different context on this list, and I wondered if
that might be what is causing my problem. I have tried porting the
working sample database to Lotus Approach on a Windows 98 system and
then exporting it as a ascii text file that should work with the Linux
system, however when I select the delimiter (tab) and export it, the
file that is ported to Linux has extra visible quotation {"xxxxxx"}
marks around each field. If there are extra not visible line controls
that Linux/Minivend is not expecting, this may be causing the problem
with the search engine. I will try your suggestion about stripping the
DOS file. Many thanks for the idea.
--
John Foster
AdVance-Computing Systems
[EMAIL PROTECTED]
ICQ# 19460173
