FMSLogo Code

Make REPCOUNTTEST.TIMERTEST more space efficent by changing a counter that involved string concenation to an incrementing integer. This is equally effective at testing repetition, but it uses less memory and causes less work for the garbage collecter.

Rewrite a test to not include non-ASCII characters directly in the source code and instead use CHAR to create it. This was done so that certain UNIX tools, like grep, don't treat the file as binary.

I confirmed that the bug can be reproduced in FMSLogo 7.2.0 using CHAR, so there is no loss of test coverage.

Sweep through all the tests and update the bug numbers to match what they are in SourceForge today. The old 7-digit numbers are no longer visible after SourceForge renumbered the bugs to be per-project instead of site-wide.

Many small non-functional improvements:
* LOAD "testlib for SIMPLEREPORTTEST
* Update the bug number to match the new number of SourceForge's changes.
* Capitalize the type IDs to make it easier to tell an L from a 1.

Ignore the object files from the Simplified Chinese build.

Add a place-holder for a Simplified Chinese translation provided by "test down" on SourceForge.

Add "backward compatibility" tests for DLLCALL to ensure that future enhancements don't break compatbility with old programs that happen to work.

Clarify that arguments are passed in reverse order in DLLCALL for stdcall convention.
Also explain that PARSE is used on the output.

Add a test case for DLLCALL when the string returned from the native function contains spaces. In this case, DLLCALL returns each word as a separate element in the returned list.

Prepare the release notes for whatever version follows 7.0.0.