Mode19 Engine BUG FIX - Update

new topic     » topic index » view thread      » older message » newer message

It is with my sincere regrets that I must inform you of a memory leak
that has gone undetected within the Mode19 Engine until now.

    The only way to correct the problem was to introduce a new
procedure to the Engine, clean_up().  I also added a feature with this
procedure.  This means this is both an bug fix update and a mild
upgrade.

    Memory was being allocated for our machine code but it was never
freed.  It had been the responsibility of the engine user to free() the
memory he used when using Bmp2Mem() and to call x = set_screens(0)
to free() the memory used by the virtual screens before
closing/quitting/aborting the program.

    Two changes have been made.  These changes "could" but "should not"
break
any existing code.
1. x = set_screens(integer) must be called before any call to Wait_Retrace()
or
any of the display routines. fdisplay(), cdisplay(), etc....
2. You do not need to call x = set_screens(0) before
closing/quitting/aborting
the program. But, you should call clean_up(). clean_up() will free() the
memory
used for Bitmaps, machine code, and virtual screens.

PS: Multiple calls to set_screens() will not cause multiple copies of the
machine
code to exist.

        Lucius L. Hilley III
        Representative of Hollow Horse Software

new topic     » topic index » view thread      » older message » newer message

Search



Quick Links

User menu

Not signed in.

Misc Menu