These two Merlin hacks were proposedly meant to be examples for pro-
gramming in GNU-C and testing my inliner's for the Merlin libs. Both
of it belong to the last adventures on our planet, I can tell you !
So I had a (too) long fight against the various traps and bugs of the
compiler/Merlin-Descriptions, respectively.

One crazy example so you may get the point:
MerlinFli compiled happily with the baserel option while MerlinFract
simply did not, although it is shorter and contains exactly the same
code on start, which was suddenly complained at by the compiler. So
MerlinFli is baserel, MerlinFract absolute (that is, pc-relative due
to the new options). And that's just one of the more harmless bugs !

By the way, baserel was never gone or wrong with the past-2.3.3 versions,
one can happily compile with this option in all versions. You just have
to use the 2.3.3-Assembler to do so. These actual assemblers simply pro-
duce a wrong object-file format that can not be linked correctly.
Try it !

One more remark about AllocDosObject mentioned in the heading comment
of MerlinFli. Although I own the 3.1-NDUK, I couldn't find a ADO_XYZ-
name-tag that seems to be absolutly neccessary. So I could not implement
this feature. Am I too silly to see the right way, or is this Tag really
missing ?? I would be very pleased if someone could answer this problem
in any way.

So Bug-Reports, comments, improvements, suggestions or whatever you like
are very welcome. For I am not reachable by email, here is my s-mail
address:

Rainer Trunz
Pfingstweidweg 10
63526 Erlensee/Germany