| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9
Comments:
<0> http://en.wikipedia.org/wiki/IronPython
<1> oh neat
<1> he's talking abotu 'active objects' and 'futures' (for concurrency)
<1> ...awesome
<2> Whats the .NET equivalent of CreateProcess/exec?
<3> Process.Start
<2> thank you
<4> hmm I'm having this problem again and I don't want to use a for loop for it
<4> vector<pair<int, int> >::iterator x = find_if(a.begin(), a.end(), _1.first == code);
<4> how do I do that _1.first
<4> ?
<1> _1.first - is that valid?
<4> and bind the whole expression
<4> nope it's not
<1> oh
<1> boost::bind( &Type::first, _1 ) == code
<1> (I think)
<4> hmm
<5> sk8ing you're trying to write the functor for your find_if ?
<4> vawjrwrk: yes
<5> and you want it inline rather than just writing it, right?
<4> yup
<5> rdragon does bind work on variables?
<1> yes
<4> I thought it works only on functions
<1> well, member variables
<1> no, if you bind to a member variable, it'll evaluate to that
<4> nice
<1> you can also compare the results of bind
<5> someyear we'll get the language to recongnize _1.blah
<1> like bind( &A::x, _1 ) == bind( &A::x, _2 )
<1> etc
<4> error C2039: 'first' : is not a member of 'std::vector<_Ty>'
<1> http://boost.org/libs/bind/bind.html#with_member_pointers
<1> yes, that error message is correct
<1> vawjrwrk have you seen anything regarding 'active' objects/code and 'futures' ?
<4> the example I'm seeing there is for member functions, used to call the binded function from an external object
<4> oh, nevermind, it seems it should work on data members too, but it doesn't
<1> it does, maybe you've got something wront
<4> x = find_if(a.begin(), a.end(), boost::bind( &XType::first, _1 ) == code);
<1> what's XType ?
<4> vector<pair<int, int> >
<1> and what is a?
<4> code is int too
<4> a is an XType isntantioation
<4> x is an iterator
<4> uh
<4> sorry
<4> damn
<1> :)
<1> hit your head with the desk
<4> :P
<6> http://www.joannejacobs.com/mtarchives/016408.html
<3> are they even proposing these concurrency changes for 0x?
<1> im not sure peterhu
<1> I think it's more of a side project ("Concur")
<3> makes it sound like it's either something they want to do for the next cli, or an extension
<3> the language should could use something like this
<1> yeah
<3> should = sure
<3> did they propose concepts too?
<3> because those *definitely* need to go in
<1> not sure if it's proposed for 0x
<1> it might be
<3> i haven't kept up with the proposals beyond tr1
<1> but yeah he covered concepts... awesome
<4> thanks, it works
<1> sk8ing - cool
<4> boosts docs could be improved
<4> I always have a hard time with them
<1> hehe
<0> "multiplayer spreadsheets!"
<0> Sheesh
<2> Is there something like a getCurrentDirectory in .NET?
<1> my guess is System.CurrentDirectory
<1> but it's probably wrong ;)
<2> Whood yoo laike tou treii agaain?
<3> Directory.GetCurrentDirectory
<2> peterhu is da man!
<2> thanks
<1> sweet: vector< future<int> > results;
<5> rdragon what's that?
<3> something from concur
<1> yeah
<3> sutter's pet project right now it seems =)
<1> a 'future' is basically data that 'will get filled in at some point', and you can call .wait() on it to actually block and guarantee the result is there
<5> ah
<1> this presentation might be a year old
<1> 'pdc05'
<5> a "mailbox" that will hold exactly one T (does it "go away" after its filled and read?)
<1> no I don't think so vawAFKhome
<5> what happens if someone ***igns to it twice before you wait() on it
<3> likely doesn't allow copying
<3> hmm
<3> good question though
<1> I think the intention is to combine it with the 'abstract' concept, so that it holds the result of only one operation
<1> future<int> x = active{ return someFunction(); }
<1> I don't think the intention is to push the future object to something that can fill it in, but rather pull it from the result
<5> ah, thought maybe it was some concurrency thing
<1> well, it is
<1> basically so you can do x = f(); and not block on the call to f()
<1> seems to be good stuff brewing
<3> yes, quite cool
<3> i'll be happy just with auto and concepts for 0x
<1> yeah
<3> i prefer the for each (foo f in bar) C++/CLI syntax to for (foo f : bar), though
<1> yeah, it's a bit more friendly
<3> i know they want to avoid new keywords, but :? blech
<3> quite an effective presentation however
<1> yup
<1> he has 3 on his site http://www.gotw.ca/
<5> typeof ??
<1> ?
<5> wouldn't you like to see typeof added
<1> not sure what it is
<1> do you mean 'auto' ?
<3> typeof gives the result of an expression
<3> per my understanding, at least
<3> (type of the result)
<5> no I mean typeof typeof(v)::iterator
<3> like this: template <typename A, typename B> X foo(A a, B b) { typeof(A * B) c = a * b; ... }?
<3> oh
<5> for example with std::vector <int> or std::list v<widge>;
<7> is the 'for each' syntax going in the ISO standard?
<3> Buoy: the commitee is considering some syntax, but i don't think they've decided on it yet
<1> vawjrwrk - ah, gotcha
<3> vawjrwrk: sounds like the same use, where the type of expression (v) is the type of the variable
<5> yes
<3> oh, you were repsonding to rdragon-wrk
<3> sorry
<7> is '>>' in templates also being considered?
<3> Buoy: i believe it will be corrected, yes
<3> Visual C++ 2005 already fixes it
<1> yeah I think that's old news
<1> oh, does it?
<1> neat
<3> yep
<1> too bad I don't get to use it anymore
<3> if you want to use a operator, you'll have to use parens
<3> though anyone that does will be shot on sight
<3> time to head home
<1> that's slacker talk if i've ever heard it
<5> we've got a new VP and we're heading to the brew pub in a few minutes
<1> man, you too
<1> what a bunch of slackers around here
<5> it will be 5pm before we leave
<5> you're the one hung out on the left coast
<1> hehe
<5> well, time to try the 2006 (export) stout award winner
<6> a new VP?
Return to
#c++
or
Go to some related
logs:
chelseahair
#networking
install-mbr fedora
Liba Maltija
#linuxhelp
#linux
install rtl8185 fedora core 5
ubuntu 6 how to start wine
ubuntu aspire 3623
#c++
|
|