www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Can GC be beneficial - Thread.

reply ZZ <ZZ zz.com> writes:
Here is a comment by Andrei Alexandrescu.
in a very long thread onm the benefits of using a GC.
from comp.lang.c++.moderated.

I think the one language that can be used to assess a combo of GC,
efficiency goals, and deterministic finalization is D.

The link:
http://groups.google.com/group/comp.lang.c++.moderated/browse_thread/thread/84253d37f970dd2b/d63fafb85011e823#d63fafb85011e823

Zz
Mar 10 2006
parent reply Kris <Kris_member pathlink.com> writes:
In article <dus4uo$1b0f$1 digitaldaemon.com>, ZZ says...
Here is a comment by Andrei Alexandrescu.
in a very long thread onm the benefits of using a GC.
from comp.lang.c++.moderated.

I think the one language that can be used to assess a combo of GC,
efficiency goals, and deterministic finalization is D.

I'd agree, except that D does not have deterministic finalization. Far from it.
Mar 10 2006
parent reply Mike Capp <mike.capp gmail.com> writes:
In article <dusjqh$24jq$1 digitaldaemon.com>, Kris says...
I think the one language that can be used to assess a combo of GC,
efficiency goals, and deterministic finalization is D.

I'd agree, except that D does not have deterministic finalization. Far from it.

Sure it does. Old-style "auto". It's very limited compared to C++'s deterministic finalization, but in some ways that's the point. It forces users to find out the hard way whether or not that level of support is enough. As a guinea-pig language, I think the bigger worry is people drawing premature conclusions from the not-especially-optimized state of D's current GC. cheers, Mike
Mar 10 2006
next sibling parent Kris <Kris_member pathlink.com> writes:
In article <dusl55$26jl$1 digitaldaemon.com>, Mike Capp says...
Sure it does. Old-style "auto". It's very limited compared to C++'s
deterministic finalization, but in some ways that's the point.

That's true. I was grimacing about finalization at global scope, which inevitably makes me wince. Local finalization support via scope(x) and/or auto is certainly handy ~ mea culpa. - Kris
Mar 10 2006
prev sibling parent reply Bruno Medeiros <daiphoenixNO SPAMlycos.com> writes:
Mike Capp wrote:
 In article <dusjqh$24jq$1 digitaldaemon.com>, Kris says...
 I think the one language that can be used to assess a combo of GC,
 efficiency goals, and deterministic finalization is D.


Sure it does. Old-style "auto". It's very limited compared to C++'s deterministic finalization, but in some ways that's the point. It forces users to find out the hard way whether or not that level of support is enough.

And there is also 'delete', which is pretty deterministic, no? :P -- Bruno Medeiros - CS/E student http://www.prowiki.org/wiki4d/wiki.cgi?BrunoMedeiros#D
Mar 11 2006
parent reply kris <foo bar.com> writes:
Bruno Medeiros wrote:
 Mike Capp wrote:
 
 In article <dusjqh$24jq$1 digitaldaemon.com>, Kris says...

 I think the one language that can be used to assess a combo of GC,
 efficiency goals, and deterministic finalization is D.

I'd agree, except that D does not have deterministic finalization. Far from it.

Sure it does. Old-style "auto". It's very limited compared to C++'s deterministic finalization, but in some ways that's the point. It forces users to find out the hard way whether or not that level of support is enough.

And there is also 'delete', which is pretty deterministic, no? :P

Aye; yet, I vaguely recall 'delete' does not invoke the dtor ... perhaps the wrong kind of finalization :)
Mar 11 2006
parent Bruno Medeiros <daiphoenixNO SPAMlycos.com> writes:
kris wrote:
 Bruno Medeiros wrote:
 And there is also 'delete', which is pretty deterministic, no? :P

Aye; yet, I vaguely recall 'delete' does not invoke the dtor ... perhaps the wrong kind of finalization :)

:o ... You should know better than that. :] -- Bruno Medeiros - CS/E student http://www.prowiki.org/wiki4d/wiki.cgi?BrunoMedeiros#D
Mar 11 2006