InstallAware

Hi,

After upgrading to IA16, I noticed that for some reason my program does not read the INI-files edited by IA-installer anymore. Reason seems to be that IA converts the INI file to UTF-8 with BOM (http://en.wikipedia.org/wiki/Byte_order_mark) in the beginning of file. Strangely enough, IA does not seem to be able to read such files itself. However, read-problems can be fixed by inserting an empty line in the beginning of file. I guess this has something to do with old Microsoft libraries that do not understand BOM?

So, as a heads up to other users if a fix is not possible:

1. InstallAware 16 outputs INI-files that are UTF-8 even if the original file is ANSI.
2. InstallAware 16 "Edit INI File" -function does not work correctly if the original file is already UTF-8 (with BOM). Fix below.
3. UTF-8 INI file read problems (in any program) can be fixed by inserting an empty line at the beginning of file.

Staff, please try this out and comment.

Dear User,

let me verify it.
Regards

Dear CNC,

if you save the ini file using UNICODE encoding instead of utf-8 ... it works fine.

Regards

Not quite sure what you mean since UTF-8 is Unicode encoding. But I guess you mean Unicode without Byte Order Mark. Then yes, it works as input. But the output file format is still UTF-8 with BOM.

I have to make a correction to my first post. The problem with BOM only affects the first section in an ini-file (for example [FIRST_SECTION]). That is, if an ini-file has its first section at the very beginning of the file, and if the file has a BOM, then IA will not see the first section at all. We had this same problem with our own software but we fixed it. And I did a workaround in InstallAware by inserting an empty line in the beginning of the ini-file before editing it. So, not a problem for us anymore. But I still think that people need to keep their heads up with unicode files. Problems can occur if the first three bytes in an ini-file are: 0xEF,0xBB,0xBF (=BOM).