... Yes, the Borland compilers (BCB4, BCB5 and BCC5.5) all support them. Mark...
Mark Rodgers
mark.rodgers@...
Sep 1, 2000 7:43 am
4942
... class ... These compilers are broken. The standard is quite clear that the only private members forbidden in an aggregate are private non-static data...
Mark Rodgers
mark.rodgers@...
Sep 1, 2000 7:43 am
4943
... This could just be unfamiliar terminology... which part of "The C++ Programming Language" (3e) describes: * template template parameters * partial template...
Ian Bruntlett
Ian@...
Sep 1, 2000 7:56 am
4944
In message <200009010021.e810LkC27140@...>, jsiek@... writes ... Borland C++ 5.5 ...
Kevlin Henney
kevlin@...
Sep 1, 2000 9:13 am
4945
In message <024b01c013ac$27171260$37781990@...>, Greg Colvin <gcolvin@...> writes ... And the only things they do "promise" are related to...
Kevlin Henney
kevlin@...
Sep 1, 2000 9:14 am
4946
In message <v03110700b5d46b34744f@[153.32.222.4]>, Jesse Jones <jejones@...> writes ... Most of the time I'm probably right, and in this case I'm...
Kevlin Henney
kevlin@...
Sep 1, 2000 9:14 am
4947
In message <003d01c01367$7064fe40$1d0524d4@pdimov>, Peter Dimov <pdimov@...> writes ... For this I have tended to use a cloning smart pointer, so that...
Kevlin Henney
kevlin@...
Sep 1, 2000 9:51 am
4948
In message <006601c013ea$c9ee8160$0801a8c0@AESL>, Ian Bruntlett <Ian@...> writes ... This is under "C.13.3 Templates as Template...
Kevlin Henney
kevlin@...
Sep 1, 2000 9:51 am
4949
... a ... But 4.2.44 isn't shipping; for Win 2K the beta is due in a few days, for example. I think we better stay away from any feature so poorly support as ...
Beman Dawes
beman@...
Sep 1, 2000 11:30 am
4950
... I, for one, haven't forgotten copy and/or cloning pointers. I'd like to include them in my smart pointer generator experiments before Toronto. --Beman...
... CodeWarrior 5.3 does not. 6.0 will. -Howard...
Howard Hinnant
hinnant@...
Sep 1, 2000 12:40 pm
4953
... But ref<T> is a cloning smart pointer. So is any, for that matter. :) In C++ we are limited to op-> and smart references are impossible. The difference is...
Peter Dimov
pdimov@...
Sep 1, 2000 12:46 pm
4954
... I've been thinking about this, trying to 'unify' ref<void> and any, and the two methods to get the value (any_cast is not a primitive) seemed a bit ...
Peter Dimov
pdimov@...
Sep 1, 2000 12:46 pm
4955
... Back to the list I think we're talking about: typedef ... atomic_t; long increment(atomic_t& dest); long decrement(atomic_t& dest); long add(atomic_t&...
Bill Wade
bill.wade@...
Sep 1, 2000 1:27 pm
4956
... From: "Peter Dimov" <pdimov@...> ... code. ... How do you make a vector of polymorphic objects such that insertion doesn't cause lots of unneccessary...
David Abrahams
abrahams@...
Sep 1, 2000 1:34 pm
4957
... of ... doesn't ... I was thinking of making ref<T> reference counted. The non-const methods would clone on demand. -- Peter Dimov Multi Media Ltd....
Peter Dimov
pdimov@...
Sep 1, 2000 1:42 pm
4958
In message <006c01c01411$b4b78720$480524d4@pdimov>, Peter Dimov <pdimov@...> writes ... Although source-level compatible, I think the naming may be a...
Kevlin Henney
kevlin@...
Sep 1, 2000 2:01 pm
4959
In message <006d01c01411$b72308e0$480524d4@pdimov>, Peter Dimov <pdimov@...> writes ... Well, any uses cloning internally, but it's not a smart pointer. ...
Kevlin Henney
kevlin@...
Sep 1, 2000 2:02 pm
4960
In message <016701c0141a$940da000$480524d4@pdimov>, Peter Dimov <pdimov@...> writes ... <sound of can of worms opening> ... ...
Kevlin Henney
kevlin@...
Sep 1, 2000 2:16 pm
4961
Hi, Yet one more code comment: I was reading C&E's book "Generative Programming" last night and came across this gem: if<condition, class_name1::type,...
Gary Powell
Gary.Powell@...
Sep 1, 2000 3:43 pm
4962
Yes, that is a great suggestion! I'll try to apply this to adjacency_list.cct....
jsiek@...
Sep 1, 2000 3:46 pm
4963
... Hum... The standard says [8.5.1 Aggregates]: "An aggregate is an array or a class (clause 9) with no user-declared constructors (12.1), no private or...
Beman Dawes
beman@...
Sep 1, 2000 3:48 pm
4964
... It is. to_ptr<U> is better in the first case, while is<U> is better in the second. ... [versus] ... Yes, I see. So copy_to() can be implemented in terms of...
Peter Dimov
pdimov@...
Sep 1, 2000 4:19 pm
4965
... A question of terminology, I'm afraid. What's the definition of a smart pointer? 'any' doesn't overload op-> and doesn't look like a pointer, but it acts...
Peter Dimov
pdimov@...
Sep 1, 2000 4:41 pm
4966
In message <001901c01430$7e973b80$130524d4@pdimov>, Peter Dimov <pdimov@...> writes ... Not really. Free functions are not that free: they are still part...
Kevlin Henney
kevlin@...
Sep 1, 2000 5:04 pm
4967
In message <001f01c01433$7a1bbe20$130524d4@pdimov>, Peter Dimov <pdimov@...> writes ... Well, what a smart pointer is is fairly well understood, and any ...
Kevlin Henney
kevlin@...
Sep 1, 2000 5:15 pm
4968
... interface ... Having copy_to as a free function makes it easier for somebody to make it work with, say, ref<void> instead of any. ... Don't do this. :)...
Peter Dimov
pdimov@...
Sep 1, 2000 5:25 pm
4969
... And since ref<T> is a cloning smart pointer, it somehow offers its facilities for cloning to the public and is not encapsulated; yet ref<void> == any. A...
Peter Dimov
pdimov@...
Sep 1, 2000 5:47 pm
4970
For those interested, I wrote an article of a similar iterator type (limited to just words in a string) for CUJ in Aug '98 -- It's online at...