Unicode Considerations

We need to run trimrun from Unix to start the printer queues (done from shell scripts) and to install database components (from perl scripts). How can this be accomplished?

As long as you don't need ttm (Trifox Terminal Manager) it works fine. A set of routines will interface Unicode chars to 8 bit etc. Just link a regular trimrun and you will be fine.

The applications we start from Unix are reports and standalone triggers. There is no need for any communication with a client. And the parameters given for the programs are numbers and filenames, i.e. the parameters could for our part be given as ASCII. I don't know how to send parameters containing NULLs from a script anyway.

Should work fine. Just remember that when writing to a file (list_file(), for example) it is written as Unicode (it will read 8-bit or Unicode automatically).

Would you describe the approach you take for Gcmnode/Unicode on Windows95. Where will you have Unicode and where will you have Shift-JIS ?

The general plan is to keep all strings in 'chad' (double byte) format until the very last stage before we call a system or Win32 api. When recieving strings they will be converted to 'chad' string ASAP.

There is a set of macros (similar to what we did on the server side) and a set of routines (similar to 'toswide.c' on Unix) that provide the specifics (like the DBCS to Unicode conversions, etc).

The Win95 version of 'gcmnode' code base is identical to the NT version.

© 1985-2020 Updated 26 Sep 2005.