8 bit os_f handles

[Jonathan Coxhead]

   David wrote,

 | In previous versions of OSLib os_f was incorrectly defined
 | as an 8 bit quantity. This has now been fixed and os_f is
 | a 32 bit quantity conforming to the RISC OS API.
 | 
 | Before installing an using this version, check you code only 
 | passes os_f to file functions, and you make no invalid casts 
between
 | os_f and byte wide variables, and dont rely on the previous
 | size in
 | persistant structures. Performing these changes now is essential
 | to ensure your code contines to work with new versions of RISC OS.

   Why is this better than globally replacing |os_f| with |os_w| and 
making the same checks? It doesn't sound like such a no-brainer to
me ...

 |  All code must be fully recompiled and relinked in this version
 |  as is the correct practice when installing ALL new libraries.

   I am getting very tired of hearing you repeat this braindead 
opinion over and over again. Are you incapable of understanding what 
you read, or do you have short-term memory problems?

 |  [Details on how to make a correct makefile]

   You mean "details of how to make a makefile in the style that 
David J Ruck would like you to". You are out on a limb, cutting off 
the branch you are sitting on.

   I'm not going to respond to this kind of nonsense again.

        /|
 o o o (_|/
        /|
       (_/