Re: [StrongED] StrongED 4.70a5 available

[David Pitt <pittdj@xxxxxxxxxxxx>]

Fred Graute, on 25 May, wrote:

> In message <mpro.n5zcx300cvxs006ut.pittdj@xxxxxxxxxxxx>
>           David Pitt <pittdj@xxxxxxxxxxxx> wrote:
> 
> [snip]
> 
> > I have had issues with recent alphas on OS4.39 but that is not what is
> > described here.
> 
> Could you elaborate on that David?

It is the crash on OS4.39 I reported when a4 was releaded, fixed by
disabling the use of DAs.

<quote>
> It seems that this could be an OS4.39 thing rather than a VRPC thing.

Agreed, and I think I know what the problem is now. Calling OS_HeapSort
with too high an address which can happen on ROL versions of RISC OS.

StrongED will use a dynamic area on them to avoid the wimpslot limit but
a DA can located high up in (logical) memory and the 26-bit OS_HeapSort
can't handle that.

To confirm this, set StrongED$UseDA to 0 in the !Run file to force the
use of application memory then run StrongED with VRPC set to 128MB. If
I'm right this should work fine.

That would just leave me with a decision on how to fix this; bring back
the internal HeapSort routine or force the use of OS_HeapSort32 (and by
extension CallASWI v0.03).
</quote>

'set StrongED$UseDA 0' is a sufficient answer for my purposes, and no one
else has mentioned the matter.

-- 
David Pitt

--
To unsubscribe send a mail to StrongED+unsubscribe@xxxxxxxxxxxxxx
List archives at http://www.Torrens.org.uk/RO/StrongED/index.html