I am frequently getting an error message:
0.68a error count 1: Dynamite area corrupt: anchor or anchor address corrupted at 8026 (or similar)
This has only recently started, and I am at a loss as to why. It is usually when expiring mail.
I either have to reboot, or, as I have discovered, use the command *Dynamite_Clear which fixes it for a while.
Any advice welcome
--
Tony Clayton tony.cla...@pem.cam.ac.uk
Coins of the UK : http://www.coins-of-the-uk.co.uk
Sent using RISCOS using VirtualAcorn-SA running on a PC
... It's not so much how we stand as the direction we're moving.
> I am using Messenger 1.41d and Newsbase 0.68a on VirtualAcorn-SA with 68
> Mb of memory
>
> I am frequently getting an error message:
>
> 0.68a error count 1: Dynamite area corrupt: anchor or anchor address
corrupted at 8026 (or similar)
>
> This has only recently started, and I am at a loss as to why. It is
> usually when expiring mail.
>
> I either have to reboot, or, as I have discovered, use the command
> *Dynamite_Clear which fixes it for a while.
Dynamite is a heap manager, IIRC, that is presumably used somewhere in
NewsBase. The errors suggest memory corruption, which in turn would suggest
that something else is either corrupted or buggy. If it's just started, I'd
guess that your mail database is unhappy somewhere -- possibly corrput
files.
That's just a wild guess, however -- it's years since I last used NewsBase.
--
Steve Fryatt - Leeds, England
Does this mean the file is a corrupt kaput one? :-)
John
--
John Mulrooney
NOTE Email address IS correct but might not be checked for a while.
Success is a journey, not a destination.
> In article <mpro.kscslc04...@stevefryatt.org.uk>,
> Steve Fryatt <ne...@stevefryatt.org.uk> wrote:
> > possibly corrput files.
>
> Does this mean the file is a corrupt kaput one? :-)
Probably. D'oh.
(I really miss the spell-check in StrongED and Zap. M-Pro on Linux -- just
like any of the other clients I've played with -- just does the standard
sequential step-through type of check, which is much less convenient than
the RISC OS editors' "all the queries in one list" style of working.)