On Wed, 7th Oct 2020 13:24:38 -0500, Sqwertz wrote:
> .84 fixed a memory problem for me in Win 7 and continues to work
> just fine in Windows 10.
And I wrote 5 years ago (about the matter, which in itself happened another
couple of years earlier): ;-)
| On Sat, 24 Oct 2015 09:26:35 -0500, Sqwertz wrote:
| > I have been using .84 for years without any problems. And it fixes a
| > memory bug/crash that occurred when I switched from XP to Win7.
| If I'm not very mistaken, it is /not/ the code of the .84 that fixed your
| former crashes. It just happens, that the .84 is compiled using a newer
| Delphi compiler. That compiler avoided the crashes, that some internal
| Delphi library code caused on certain AMD processors when compiled with
| older Delphi compilers. (Like the one .41 has been compiled with...) So
| it all boiled down to a decision between Scylla (= buggy Delphi support
| for certain AMD processor and chipset functions) and Charybdis (= using
| a Dialog re-write in Alpha state).
Your matter was the one I referred to in my earlier posting. And it is the
/only/ case that I know of, that somebody declared /consistently/, that
.84 worked, while .41 did not. Most people lauding the .84 would have been
(at least) equally well served by .41.
One should think of .84 as a car, where a mechanic just did some internal
tests, whether a different brake system would fit. The car was never meant
to appear on street. But someone stole it and now a whole bunch of people
drives it regularly on public roads. Fortunately, the mechanic worked well,
in the first place. But a cotter pin /may/ come of, when security counts