Latest spec and boot images

[Gilad Bracha]

The latest spec (0.098) is now live on the site. A new boot image is available as well. The revised spec removes support for the method category string from the grammar, describes a new experimental feature, transient slots, designed to help support orthogonal synchronization, and has other minor corrections.

[Stefan Marr]

Hi Gilad:

On Sunday, January 15, 2017 at 9:33:30 PM UTC+1, Gilad Bracha wrote:

The latest spec (0.098) is now live on the site. A new boot image is available as well. The revised spec removes support for the method category string from the grammar,

Any specific reason for the removal?

I didn't find them particularly useful in text files, so, I won't really miss them if I remove them from SOMns as well. Am just curious.

 

describes a new experimental feature, transient slots, designed to help support orthogonal synchronization,

Hm, reading the spec, to me it sounds as if transient slots work very much like inner classes.

For a moment, I was wondering whether inner classes could be seen as syntactic sugar over transient slots but that wouldn't be correct, because we don't have class literals, only object literals.

Do I understand that correctly?

and has other minor corrections.

Is the spec a Latex document?

Since I am always curious what you change, perhaps it would be possible to get a diff, or a latexdiff (https://www.ctan.org/tex-archive/support/latexdiff?lang=en)?

Best regards

Stefan

 

[Gilad Bracha]

Hi Stefan

On Sunday, January 15, 2017 at 12:55:21 PM UTC-8, Stefan Marr wrote:

Hi Gilad:

Any specific reason for the removal?

I didn't find them particularly useful in text files, so, I won't really miss them if I remove them from SOMns as well. Am just curious.

The removal was always part of the master plan. They were there because the grammar evolved from a Smalltalk grammar. The most important use of categories was to indicate privacy, but now the language supports that. We have found little real use for categories. Someday we'll have metadata fully supported for anyone who cares. 

 

describes a new experimental feature, transient slots, designed to help support orthogonal synchronization,

Hm, reading the spec, to me it sounds as if transient slots work very much like inner classes.

For a moment, I was wondering whether inner classes could be seen as syntactic sugar over transient slots but that wouldn't be correct, because we don't have class literals, only object literals.

Do I understand that correctly?

Yes, one could view nested classes as immutable transient slots initialized with class literals. But until we decide whether to keep the construct and exactly how it behaves, we'll keep these two things separate. 

and has other minor corrections.

Is the spec a Latex document?

Since I am always curious what you change, perhaps it would be possible to get a diff, or a latexdiff (https://www.ctan.org/tex-archive/support/latexdiff?lang=en)?

Well, I finally put the spec under source control, so from now on we'll have better diffs. I can produce one for you for this round.