[EMAIL PROTECTED] (Lars Gullik Bjønnes) writes:
| | IMO there is a draw between 'old fashioned code' and lambda. If one
| | needs to decide for one or the other, other criteria are needed.
| | 'simple standard constructs' would be one of them.
>
| so bind(&Foo::name, _1, name)
This was not the ex
Andre Poenitz <[EMAIL PROTECTED]> writes:
>> | IMO there is a draw between 'old fashioned code' and lambda and the tie
>> | breaker 'simple standard constructs' prefers 'old fashioned code'.
>>
>> can you add some commas to make me understand the sentence?
>
| IMO there is a draw between 'old fas
On Tue, Jan 06, 2004 at 07:44:47PM +0100, Lars Gullik Bjønnes wrote:
> struct SameName : public std::unary_function {
>SameName(std::string const & name) : name_(name) {}
>bool operator()(Foo const & foo) const { return foo.name() == name_; }
>std::string name_;
> };
>
> a
Lars Gullik BjÃnnes wrote:
> I thought it as a nice start as it made me allocate some time, rather
> than none, on lyx code.
Cheers for that.
Alfredo
Andre Poenitz <[EMAIL PROTECTED]> writes:
| On Tue, Jan 06, 2004 at 07:14:48PM +0100, Lars Gullik Bjønnes wrote:
>> Andre Poenitz <[EMAIL PROTECTED]> writes:
>>
>> | case 1 seems to be uniformly better. So why not use it?
>>
>> because it is some ~10 lines longer.
>
| I see about 5 which goes do
On Tue, Jan 06, 2004 at 07:14:48PM +0100, Lars Gullik Bjønnes wrote:
> Andre Poenitz <[EMAIL PROTECTED]> writes:
>
> | case 1 seems to be uniformly better. So why not use it?
>
> because it is some ~10 lines longer.
I see about 5 which goes down to 4 for the unnecessary 'public' when
changing th
Angus Leeming <[EMAIL PROTECTED]> writes:
| Here I get run-time ratios (for 1000 loops) of
>
| boost::bind / Named Functor == 1.89
| boost::function / Named Functor == 1.89
| boost::lambda / Named Functor == 2.45
>
| Draw your own conclusions.
Gcc 3.4 is better than gcc 3.3.
--
Lgb
Angus Leeming wrote:
> For completeness, here is the same code with an additional
> boost::function test:
>
> $ size bind_lambda?
> dec hex filename
> 8799225f bind_lambda1 // Named Functor
> 10508290c bind_lambda2 // boost::bind
> 158943e16 bind_lambda3 // boost::function
> 105
Andre Poenitz <[EMAIL PROTECTED]> writes:
| case 1 seems to be uniformly better. So why not use it?
because it is some ~10 lines longer.
And if the functor is only being used once... (the tables are turned
if the functor is used in several places)
Also it is not unlikely that case 2 and 3 will c
John Levon <[EMAIL PROTECTED]> writes:
| On Tue, Jan 06, 2004 at 06:44:21PM +0100, Andre Poenitz wrote:
>
>> case 1 seems to be uniformly better. So why not use it?
>
| It's a bit long winded, that's why. But since we have serious bloat
| problems already ...
And the best way to combat bloat is t
Lars Gullik Bjønnes wrote:
> Angus Leeming <[EMAIL PROTECTED]> writes:
>
> | $ size trial_case?
> |textdata bss dec hex filename
> |7850 816 8867421e2 trial_case1
> |9510 840 16 10366287e trial_case2
> |9530 872 8 10410
On Tue, Jan 06, 2004 at 06:44:21PM +0100, Andre Poenitz wrote:
> case 1 seems to be uniformly better. So why not use it?
It's a bit long winded, that's why. But since we have serious bloat
problems already ...
john
--
Khendon's Law:
If the same point is made twice by the same person, the thread
On Tue, Jan 06, 2004 at 06:40:06PM +0100, Lars Gullik Bjønnes wrote:
> Angus Leeming <[EMAIL PROTECTED]> writes:
>
> | $ size trial_case?
> |textdata bss dec hex filename
> |7850 816 8867421e2 trial_case1
> |9510 840 16 10366287e trial_c
Angus Leeming <[EMAIL PROTECTED]> writes:
| $ size trial_case?
|textdata bss dec hex filename
|7850 816 8867421e2 trial_case1
|9510 840 16 10366287e trial_case2
|9530 872 8 1041028aa trial_case3
With gcc 3.4:
size t
Angus Leeming <[EMAIL PROTECTED]> writes:
| Lars Gullik Bjønnes wrote:
| Whould have been nice if you could actually send the trial.C file.
>>
>> Hard to duplicate the test here without it...
>
| ???
| It's attached to the original mail. See either of
I am stupid ignore me.
--
Lgb
Lars Gullik Bjønnes wrote:
Whould have been nice if you could actually send the trial.C file.
>
> Hard to duplicate the test here without it...
???
It's attached to the original mail. See either of
http://articles.gmane.org/gmane.editors.lyx.devel:32848
http://marc.theaimsgroup.com/?l=lyx-devel&
Angus Leeming <[EMAIL PROTECTED]> writes:
| Attached is a sample code that implements the three approaches that Lars has
| been playing with.
>
| std::string const search_name = "william";
>
| FooVec::const_iterator const begin = foovec.begin();
| FooVec::const_iterator const en
On Tue, Jan 06, 2004 at 04:02:51PM +, Angus Leeming wrote:
> I have little idea how to make sense of the assembler code but, at least for
> gcc 3.3, it appears that the named functor approach produces the
> smallest-sized executable.
>
> $ size trial_case?
>textdata bss dec
Attached is a sample code that implements the three approaches that Lars has
been playing with.
std::string const search_name = "william";
FooVec::const_iterator const begin = foovec.begin();
FooVec::const_iterator const end = foovec.end();
FooVec::const_iterator
19 matches
Mail list logo
