Hello Seighard.
> And I'v ro cicede that I'm just now too busy with other projects to even
> look over any such problems, even the zip-file problem you informed me
> - sorry.
No problems, me too I am very busy and can give only counted time!
It is why I add here some of my investigations, so I can remember it for
next fight.
About fpc 3.3.1 > rev 2375 of Jul 13 2019, I still try to localize the
problem
Last investigations give:.
In msestat.pas, procedure procedure tstatfile.internalreadstat, see the //
commit, the problem comes from there
________________________
procedure tstatfile.internalreadstat;
var
int1: integer;
begin
if assigned(fonstatread) or assigned(fonstatupdate) or
(fsavedmemoryfiles <> '') then begin
areader.readstat(istatfile(self));
end;
if fobjectlinker <> nil then begin
aclients:= nil;
aclientcount:= 0;
fobjectlinker.forall({$ifdef
FPC}@{$endif}dolinkstatreading,typeinfo(istatfile));
setlength(aclients,aclientcount);
sortarray(aclients,sizeof(statclientinfoty),@cmpclients);
include(fstate,stfs_reading);
try
int1:= 0;
while int1 <= high(aclients) do begin
with aclients[int1] do begin
if intf <> nil then begin // here problem, aclients[int1].intf = nil
with fpc 3.3.1
areader.readstat(intf);
end;
end;
inc(int1);
end;
finally
exclude(fstate,stfs_reading);
aclients:= nil;
fobjectlinker.forall({$ifdef FPC}@{$endif}dolinkstatreaded,
typeinfo(istatfile));
end;
end;
end;
__________________________
But I did not catch why aclients[int1].intf = nil with fpc > 3.3.1. rev
2375.
The array aclients is defined (imho) by fobjectlinker.forall({$ifdef
FPC}@{$endif}dolinkstatreading,typeinfo(istatfile));
I am still lost.
Maybe some light will come later.
Anyway, mse is ready for fpc 3.2.0. and this is good.
Fre;D
--
Sent from: http://mseide-msegui-talk.13964.n8.nabble.com/
_______________________________________________
mseide-msegui-talk mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mseide-msegui-talk
