Re: Thunk routines (Gene Wirchenko)
6 Mar 1998 16:50:26 -0500

          From comp.compilers

Related articles
Thunk routines DAVEJ3@Prodigy.Net (DAVEJ3) (1998-02-20)
Re: Thunk routines (1998-03-03)
Re: Thunk routines (1998-03-03)
Re: Thunk routines (Nick Roberts) (1998-03-06)
Re: Thunk routines (1998-03-06)
| List of all articles for this month |

From: (Gene Wirchenko)
Newsgroups: comp.compilers
Date: 6 Mar 1998 16:50:26 -0500
Organization: All USENET --
References: 98-02-103 98-03-013
Keywords: history (Scott Amspoker) wrote:

>DAVEJ3 <DAVEJ3@Prodigy.Net> wrote:
>>Does anybody know where the routine name Thunk came from?
>>[It was coined about 1960 to describe the callback routines needed to
>>implement call-by-name, shortly after the Algol60 committee realized
>>they'd invented call-by-name by mistake. -John]
>Thank you for the reality check on "thunk". For ages I always thought that's
>what it was (used to implement call-by-name) but it seems that the usage has
>changed in recent years - refering to a glue layer between incompatible
>modules. I figured I remembered incorrectly.

          Considering how call by name has to be implemented, the current
definition applies to it as well. There has been a discussion of call
by name on alt.folklore.computers recently.


Gene Wirchenko

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.