🚀 go-pugleaf

We're currently looking at making our applications somewhat language 
independent. At present we only want some method of providing an 
alternate 'char *'. Multi-byte strings are not currently a 
requirement.

We already have a numeric code based system for our error codes so 
we've decided to extend this to all strings.

The Borland documentation appears a bit sparse in this area. It 
describes the ITE and indicates that resource string lists are the 
way to go but then fails to explain how you do the latter for your 
own strings. I have already set up the ITE for two languages and am 
happy about using that.

This has raised a couple of questions though:

1.Translation Manager
This seems to only deal with strings from components. I can't find 
any way to add my own strings. Is there supposed to be some way to do 
this or is it only concerned with VCL strings?

2.Resource string IDs
We can create our own .RC and add it to the project so that we can 
define our own resource strings. It'd be really nice if we could use 
our existing error codes as string IDs but this raises several 
issues:

  How do we avoid conflicts with other string IDs?
  Some of our error codes are -ve. Can you have -ve resource IDs?
  Most of our error codes are 16-bit but some of the more recent are 
32-bit. Is this going to be a problem?

Your thoughts and comments would be appreciated.

Andrue Cope
[Bicester, UK]