Question: Block and ST-76

Marcel Weiher marcel at system.de
Sun May 2 12:00:19 UTC 1999


Hi folks,

just finished reading the 'Early History of Smalltalk' paper in  
HOPL-II a couple of weeks ago, and now I have some question regarding  
blocks and open-color keywords.

>From the short snippets of code in the paper, it looks like the  
square-bracket/block notation was used generally in Smalltalk-76 to  
delimit chunks of code, instead of only for delayed evaluation as it  
is now.  Closed-colon keywords evaluated the expression (be it a  
block or something else) immediately (pass-by-value), open-colon  
keywords simply passed the expression itself (generalized  
pass-by-name).  Is this correct?

It also seems that some of the problems described in 'Smalltalk-80:  
Bits of History, Words of Advice' are actually part of the bigger  
problem of dsitinguishing between names/references and the objects  
they refer to when both are first class object, the same going for  
proxies, collections etc.  Have there been any attempts at a general  
solution?

Marcel





More information about the Squeak-dev mailing list