Re: How to get Ada to ?cross the chasm??

[Niklas Holsti <niklas.holsti@tidorum.invalid>]

On 18-05-10 01:05 , Randy Brukardt wrote:
> "Niklas Holsti" <niklas.holsti@tidorum.invalid> wrote in message
> news:flegbrFj0qjU1@mid.individual.net...
>> On 18-05-04 01:27 , Randy Brukardt wrote:
>>
>>   [snip]
>>
>>> There is no realistic chance that Ada will
>>> ever be changed enough to make it amenable to GC.

    [snip]

> If one uses the canonical linked-list
> implementation of finalization, objects with controlled parts remain
> reachable until the appropriate master goes away -- so nothing gets
> collected. (And a similar effect is necessary regardless of how finalization
> is implemented.) There still can be an underlying GC and it can be used just
> fine -- no one said anything about it actually be able to collect much.

Can one summarize the effect of the current Ada rules as follows:

- for dynamically allocated objects that do not need finalization, a GC 
implementation could release them when they become unreachable, just as 
in other languages,

- for objects that do need finalization, a GC implementation would have 
to delay releasing the objects until their types go out of scope, which 
in most cases would not happen before the program is about to terminate?

If so, the conclusion is that GC is becoming ineffective for Ada, 
because Ada programming style is changing to make more use of controlled 
types and finalization.

-- 
Niklas Holsti
Tidorum Ltd
niklas holsti tidorum fi
       .      @       .