Whitesmiths C compilers

utzoo!decvax!ucbvax!unix-wizards utzoo!decvax!ucbvax!unix-wizards
Tue Nov 10 05:31:54 AEST 1981


>From minnich.EEVax at UDel Tue Nov 10 04:55:47 1981
   I have used the PDP11 and 8080 versions of the Whitesmiths compilers
for about nine months now. One bug in the 8080 version required 
hand disassembly and replacement of bit-field procedures. There was One bug in 
the 11 version, again related to bit-fields. No other problems, despite 
the fact that I intentionally stretched them a bit. 
   Many bugs in the RT-11 and CPM systems they were running on. 
   Need we go into the many bugs that exist in Unix software? 
On Unix I seem to spend an ordinate amount of time fixing 'working' 
utilities.
   The library differences are quite easily fixed, and in some cases 
Whitesmith has the better design. Especially for argument parsing. 
   As for the differences in the languages, Whitesmith has made 
some positive changes, although old-style C hackers might not like it.
For instance, one can set a switch to c0 that will require structure 
elements be used only on matching types. Much like Pascal.
   My main point here is that Whitesmith is a mixed bag. It is 
a total loser only to the extent that you are closed to a different 
way of doing things. 
   As for the compatibility issue, I am having great fun moving 
some old V6 C cross compilers for the Motorola 68xx family to 
4.1BSD. Almost every line is illegal, the question being which ones
won't work.
ron



More information about the Comp.unix.wizards mailing list