Smalltalk › Gemtalk › GLASS

[Glass] GemStone's #on:do: could be cull: instead of value: ?

_‹ Previous Topic Next Topic _›

Classic

List

Threaded

4 messages Options

Mariano Martinez Peck

[Glass] GemStone's #on:do: could be cull: instead of value: ?

Hi guys,

Some weeks ago I noticed #on:do: behaved different than Pharo. Basically, in Pharo, the argument of the do: is not mandatory while in GemStone it is. So Pharo has a kind of #value: behavior while GemStone has #cull:.

Does ANSI Smalltalk say something about it? Do you have plans of making it like Pharo behavior?

Maybe there are strong reason for let it as is?

Cheers,

--
Mariano
http://marianopeck.wordpress.com

_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass

Dale Henrichs-3

Re: [Glass] GemStone's #on:do: could be cull: instead of value: ?

Mariano,

I am not an ANSI geek, so I can't comment on that aspect ...

It is a little unrealistic to expect GemStone to change as fast as Pharo:) I have tried to keep pace with the compatibility layer (i.e., #cull:... was added in the compatibility layer), but for the blocks that are created in the exception handling layer it isn't that easy to add change for compatibility since much of the exception handling logic is done in C code and not Smalltalk ...

With that said, I will check into the feasibility of using #cull: for the exception handling blocks...

Dale

On Mon, Jan 20, 2014 at 4:58 AM, Mariano Martinez Peck <[hidden email]> wrote:

Hi guys,

Some weeks ago I noticed #on:do: behaved different than Pharo. Basically, in Pharo, the argument of the do: is not mandatory while in GemStone it is. So Pharo has a kind of #value: behavior while GemStone has #cull:.

Does ANSI Smalltalk say something about it? Do you have plans of making it like Pharo behavior?
Maybe there are strong reason for let it as is?

Cheers,

--
Mariano
http://marianopeck.wordpress.com

_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass

_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass

Martin McClure-5

Re: [Glass] GemStone's #on:do: could be cull: instead of value: ?

In reply to this post by Mariano Martinez Peck

On 01/20/2014 04:58 AM, Mariano Martinez Peck wrote:

> Hi guys,
>
> Some weeks ago I noticed #on:do: behaved different than Pharo.
> Basically, in Pharo, the argument of the do: is not mandatory while in
> GemStone it is. So Pharo has a kind of #value: behavior while GemStone
> has #cull:.
>
> Does ANSI Smalltalk say something about it? Do you have plans of making
> it like Pharo behavior?
> Maybe there are strong reason for let it as is?

Hi Mariano,

ANSI does say that the argument to the do: is a monadic block. I don't
think it would actually violate ANSI, though, for a Smalltalk
implementation to also allow a niladic block there.

Regards,

-Martin
_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass

Richard Sargent

Re: [Glass] GemStone's #on:do: could be cull: instead of value: ?

Administrator

On Tue, Jan 21, 2014 at 1:36 PM, Martin McClure <[hidden email]> wrote:

ANSI does say that the argument to the do: is a monadic block. I don't
think it would actually violate ANSI, though, for a Smalltalk
implementation to also allow a niladic block there.

It is true that a conforming implementation is free to do more than what ANSI mandates, so making the argument optional should be acceptable.

However, it does mean your code that uses a niladic Block will NOT be ANSI compliant, reducing its portability. (And that is exactly what you found with the Pharo code.)

_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass