| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8
Comments:
<0> threat, iterate through args, ***igning each to a vector of std::string?
<1> std::vector<std::string> args(argv, argv+argc);
<1> but why do you think they'll be more useful in a vector
<2> vawjrwrk, I just like to work with stl instead of char*
<1> likewise, but those are args to the program
<2> vawjrwrk, programming preference
<1> you shouldn't be doing much with them
<1> and, btw, they're really char const*
<3> DrkMatter, i was asking myself that earlier today too.. it's getting very popular
<0> i beg to differ :) they can be modified, and modifications can have impacts in many cases
<1> tiocsti that's not guaranteed they can be modified
<0> it's undefined
<4> damn, ive coded all this to know that i can't replace a string by another in tab
<3> i wonder if they've considered allowing another main, int main(std::array<std::string const> const& args) in the next standard
<4> WTF !!!!
<1> YUY0x7 no need really
<0> doesnt windows do something similar with WinMain?
<0> conceptually, anyways
<3> it's nicer
<0> i wouldnt hold your breath on it being standardized, though
<1> toast huh??
<0> I think you'll see main replaced by a object named Main first :)
<1> tiocsti using something that's undefined is a pretty good way to cause problems
<0> vaw, indeed, but that doesnt = its const, you dont know what it is one way or another
<0> to say 'its really const' is just not true
<1> tiocsti I said it's really char const*
<1> which IS true... it's only that so many C weenies have NO ****ing clue about const correctness that the standard allows char*
<1> for the 2nd arg to main
<0> allows? it specifies that the 2nd argument is not const
<0> the 2nd argument is char *argv[], it's as simple as that. c is looselt typed though
<1> tiocsti it's a special case
<1> it SHOULD have been (and would have been in the committee had enough guts to tell the C weenies to **** off) char const* []
<0> possibly
<1> no possibly
<0> that's not the same as 'its really const' thats 'i wish it was really const'
<1> I said it's really char const*
<1> that doesn't mean that it's const
<1> it means that the compiler won't let you change it through the pointer
<1> you can trivially make a char const* point to something that is KNOWN to be non-const
<0> yes, i know what it means, and it isnt really that, the standard says its char *argv[]
<1> NO ****
<1> THEY CAVED TO THE ****ING C WEENES
<1> but IF You try to treat it that way, you're in undefined territory
<0> sure, but const would imply that the caller should rely on its contents
<0> which is not true
<5> vawjrwrk: ay up mate, how goes it, just to let you know, if you still wanted to use that page i made all that time ago, i reposted it at http://planetearthworm.com/cppbooks.php - its hosted on a paid for server rather than my cable modem so it ought to be more permanent now
<1> tiocsti huh? "rely on it's contents" ??
<0> after main() exits
<1> earthworm ok, cool
<1> tiocsti I have NO idea what you're saying
<0> const is an agreement with the caller, that this or that you can depend on the contents being unchanged or not
<1> except for people who lie to the compiler
<0> well sure, const is difficult to enforce
<1> especially with cast being available
<1> but then, how many "callers" of main(int, char*[]) do you exped there to be?
<0> yes, but if it wasnt, then they wouldnt be compatible with c at all
<1> and I'd care??
<0> theres only 1
<1> and it's NOT a user written thing either
<1> and that 1 is free to put the arguments into read-only memory
<0> yes, if they wantd to they could map it as read only
<1> equivalent
<0> although that'd be a lot more work than what they tend to do now
<0> which is just stick it on the stack
<1> depends on the system, I would guess
<0> ive never personally seen a compiler that didnt stick the args on the stack
<6> sleep time
<0> but there might be some
<6> oyasumi
<0> or more precisely, the os sticks it in the stack, and calls the ep
<0> which p***es it to main() after some preprocessing
<0> but if c++ is going to abandon c compatibility, there's bigger fish to fry than this
<1> such as?
<0> i wanna see the init code create an instance of the cl*** Main and p*** to the constructor argc and argv
<1> why?
<1> what will that buy you?
<0> and make basic types objects
<1> so you can inherit from them?
<0> it lets you do away with global functions
<0> either that or make main an exception, and dont allow any other global functions
<1> that seems like an exceptionally bad idea
<7> What's wrong with global functions?
<0> what do you need functions for?
<1> operator + would do poorly as a member function
<3> no std::find ?
<0> you could pull that into the Main object
<1> seems kinda pointless to me
<1> more specifically I don't see any advantage to it
<0> perhaps they need to do simple obvious things first
<0> like allow for static int i=0; in cl***es
<1> they do
<0> its not standardized, is it?
<1> yes it is
<0> ive used compilers where i needed to do cl***name::i=0; outside of the cl***
<0> to pull the static into scope
<1> I've used non-standard compilers also
<3> you need to do that if you need to use the address of the static member
<3> the initialization can still go inside the cl*** def
<7> Well, it doesn't work with VC++'s compiler, anyway.
<0> foo.cc:6: error: ISO C++ forbids in-cl*** initialization of non-const static
<0> member `i'
<0> im calling bs on your claims
<1> which vc++
<7> Express 2005.
<7> I ***ume it's teh same compiler as VS's.
<0> g++ claims iso forbids it
<1> sorry, I didn't read yoru thing as non-const
<8> http://www.sun-sentinel.com/business/local/chi-0602080095feb08,0,1099044.story?coll=sfla-business-front
<8> what happened to freedom of ***ociation?
<1> Noidea they stepped all over that ages ago
<1> you're free to ***ociate as long as it's not some white males doing it
<8> but, you're not
<8> you can't even discriminate against who you take as a roommate
<8> under the fair housing act
<8> you can't say, "Gay men only."
<1> IMO the "fair housing act" the word "fair" means something between poor and good
<8> bah, it's poor
<8> it's below poor
<8> it's stupid.
<8> and should be scrapped.
<1> it should be declared unconstitutional
<1> 1st amendment violation
<8> exactly
<8> but, it wasn't, was it?
<8> this has been to court I imagine?
<1> the gov't _claims_ authority because when you're a business you need a license and they can then dictate how you do business
<8> you're not in business when you're finding a roommate.
<8> you're not earning any money.
<1> I wouldn't want to test that in court
<8> MAYBE if you own the house, MAYBE
<8> but if you're living in an apartment
<8> and are looking for someone to live with
<8> to split the rent and utilities with
<8> it's like saying you can't discriminate against who you car pool with.
<8> Where's the fair car pooling act?
<9> Noidea : I have No Idea...
<9> :)
<1> you don't get it!!! the liberals are willing to put up w/ possible discrimination just to get carpools going
<8> well! I demand fair access to other people's cars!
<1> nobody ever said they were consistant
<1> NuliX turn off the auto messages
<9> oops
<1> Noidea I thought you might like this: http://www.townhall.com/opinion/columns/walterwilliams/2006/02/08/185444.html
<8> http://www.marginalrevolution.com/marginalrevolution/2006/02/what_about_the_.html
<10> Greetings,
<10> what's wrong w/ the following alias ? :
<10> alias rm mv\!*~/tomb
<10> i want to mv to ~/tomb instead of erasing data.
<8> nice
<8> that's getting e-mailed around.
Return to
#c++
or
Go to some related
logs:
pam_mount rhel
#mirc
#teens
#linux
hosting sitook
#MissKitten
#MissKitten
#linuxhelp
#linuxhelp
#java
|
|