entries as generic actuals

[ROGERS@uhcl.CSNET ("Pat Rogers, High Tech Lab")]

The question about the legality of passing an entry as the actual
to a generic formal subprogram parameter is easy enough: see RM 12.3/4,
and a couple of other places too.


Here's one for you : why is it legal to declare an entry in a task
declaration, and then not give a corresponding accept ? You have to give
supbprogram bodies for subprograms declared in the visible part of
a package, so why not for tasks ?  Who cares, you ask ? If you call such
an entry, it will suspend indefinitely.  I can find no references to this
in the LRM. Anybody know where one is if I'm wrong, or what the rationale
is if I'm right ?  (For what it's worth (apologies, etc), the DEC compiler
took it without objecting...)

Thanks !
Pat Rogers
PRogers@Ada20.isi.edu