Comments on Entropy Overload: One-liner RAII in Delphi

Somehow this reminds me of this trick with interfa...

2010-02-15T17:21:51.735Z

Somehow this reminds me of this trick with interfaces: http://www.malcolmgroves.com/stories/2004/04/05/writingSolidDelphiCode.html

2010-02-01T14:52:51.930Z

This comment has been removed by the author.

Thanks for clearing this up, Barry. So my "fa...

2010-02-01T14:52:25.521Z

Thanks for clearing this up, Barry. So my "fake RAII" instances will get a bit shorter in the future.

As to your earlier comment, about explicit variabl...

2010-02-01T14:40:27.312Z

As to your earlier comment, about explicit variables: in the hypothetical (and breaking change) case, where we optimized interface variable usage, we would likely not only break the described RAII-like functionality but also the explicit variable approach too; the values assigned to FooNotifier and BarNotifier are not used, so "in theory" they can be freed up sooner, and potentially even reuse the same storage.

But of course, destruction of the interface can have side-effects, and that's what's being relied upon for the effect in the post. Changing the language such that side-effects like these have visible changes is not something we do willingly.

uli - the situation is rather the reverse. We (Emb...

2010-02-01T14:37:41.961Z

uli - the situation is rather the reverse. We (Embarcadero) can't really change the implementation semantics because of the backward compatibility impact for people relying on the behaviour.

There is a situation where the scope is different to that expected in the post: if the method creating the interface is inlined into a different method, then the scope will be extended into the destination method. The result temporaries get turned into locals of the destination method.

@Anthony: Do you have a quote to make me feel secu...

2010-02-01T14:11:48.063Z

@Anthony: Do you have a quote to make me feel secure about this? I don't want to rely on implementation details that might change with the next compiler release. :-)

Interface scoping rules are fairly well defined. ...

2010-02-01T13:33:19.013Z

Interface scoping rules are fairly well defined. If you're not maintaining any references to the interface at all, they'll go out of scope when the procedure exits.

How advisable is it to rely on the fact that the h...

2010-02-01T08:38:33.864Z

How advisable is it to rely on the fact that the hidden interface variables from your first "procedure P" life throughout the whole routine? When implementenig similar things, I already noticed that it seems to work, but to be sure I always used explicit variables, like in

procedure P;
var
FooNotifier: IInterface;
BarNotifier: IInterface;
begin
FooNotifier := DisableEnableControls('Foo');
BarNotifier := DisableEnableControls('Bar');
Writeln('Controls on Foo and Bar are disabled now');
end;

I have a procedure that does something similar I k...

2010-01-30T00:53:42.769Z

I have a procedure that does something similar I keep handy. It's extremely useful for avoiding lots of nested try/finally blocks for things like temporary objects, BeginUpdate/EndUpdate, Open/Close, cursor changing, etc.