| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11
Comments:
<0> do you have to do something special when freeing memory with delete in a templated cl***
<0> if i leave the destructor using delete [] ptr it crashes on ints
<0> if i leave it as delete ptr it crashes on strings
<1> well, just match your news with your deletes
<1> new[]? Use deletep[
<1> er, delete[]
<1> otherwise, idernet.org) has joined #c++
<1> [22:00] <0> do you have to do
<1> sigh. My typing is all gone to hell, sorry
<0> hehe
<1> new[] pairs delete[]. new pairs delete
<1> don't mix and match
<0> how do I test a template type then
<0> to see which to use
<1> why would you have to?
<1> if you allocated it with new[], you use delete[]
<0> what's the diff between new and new[]
<0> & delete & delete[]
<1> basically deals with allocation/destruction of more than one object at once (in the array returned)
<1> don't forget, new and delete do more than just malloc ram - they deal with the constructors, destructors and housekeeping
<1> if you new foo[10], you get back 10 nicely prepared and constructed foos
<1> if you then "delete (new foo[10])", the behaviour is not really defined
<0> elem = new T;
<0> *elem = x;
<0> is that perhaps whats doing it?
<1> yes
<1> wel
<0> Node (const T x) : next(NULL), prev(NULL) {
<0> elem = new T;
<0> *elem = x;
<1> depends on x
<1> elem = new T; // Construct a T and ***ign a pointer of it to elem
<0> right
<1> *elem = x; // ***uming x is of type T, then copy x into elem
<0> right
<1> however
<1> well, no, that should be OK. Where does it crash?
<1> afk for moment - proof reading
<1> someone else will hopefully pick up in my absense :)
<0> on the delete
<0> if it's a string, it'll crash with delete ptr
<0> if it's an int, it'll crash with delete[] ptr
<1> (still afk, but post your code on the site in the topic)
<1> (also, unless this is homework, you're probably better off just using std::list or std::vector)
<0> it's a project
<0> no one has any ideas?
<2> eci around here string means "std::string" nothing else
<0> ?
<0> i know what a string is
<0> :)
<1> post your code to the site in the topic
<0> i may be on to something.. sec
<2> then this is bull****
<2> <0> if it's a string, it'll crash with delete ptr
<2> <0> if it's an int, it'll crash with delete[] ptr
<1> templated code should be type agnostic - you shouldn't need to deal with delete[], or delete whatever, nor new[] or new
<1> just new and delete, in terms of the type you were templated as
<0> vaw?
<2> those comments you made are crap
<0> they're crap why
<2> becasuse std::string* blah = new string
<2> does NOT crash with delete blah
<3> bye all G`night all
<0> what does that have to do with knowing that "string" in #C++ refers to std::string
<2> because IF you meant char* by mistake, THEN it might have been a correct statement
<2> so either you misspoke or misknew
<2> your call
<0> i was explaining the behavior i was seeing
<0> i didn't mis-speak
<2> I don't believe it crashes on "strings"
<2> btw, the simplest way to keep such a destructor from crashing, ***uming you really NEED to do a new[] for at least one case, is to do new type[1]; for the other cases
<4> new and delete are evil. you shouldn't really have them spread all over your code.
<2> Achamenes interesting bit of wisdom
<2> and they are hardly evil...... malloc() and free() are evil though
<4> OK, I agree that they're not evil in their essence... but they become evil when people don't know how and when to use them, and most don't.
<4> It's very annoying to see people use new everywhere in the code, almost randomly, and then go on whining about how C++ does not have a garbage collector.
<4> vawjrwrk: do you have much experience dealing with Unicode?
<2> Achamenes not very much, no
<4> it's such a confusing topic...
<5> Achamenes: Why is that anyway?
<4> davsn: Why is Unicode confusing (to me)?
<5> yeah
<4> because there are a LOT of little details here and there, and it's not clear whether, say, the .NET implementation of it implements a given feature or not.
<5> nope'
<4> what do you mean?!
<5> .Net has the regionalinfo cl***es to distinguish easily such stuff.
<5> also the resource files can hold the different regional settings for the application.
<4> OK, let's forget .NET for the moment. To C++ and MFC.
<6> vawjr, I've got no ops here, can you unban EwIck, before I forget?
<6> oh geeze
<6> how'd you do that?
<7> really quick typer
<4> davsn: In MFC applications, you can set it to use "Unicode character set." But what that does is to change TCHAR to unsigned short instead of just char.
<6> that's amazing
<7> off again, later
<6> you should have a tv show
<5> yo TechRaven
<7> yo davsn, i'm out of here ;> ttyl
<2> Achamenes what compiler are you using?
<6> you could have all kinds of adventures that revolve around your typing abilities.
<6> like, in that one James Bond
<6> where he keeps guessing p***words until he magically gets it.
<4> vawjrwrk: VC++7.1
<5> heh
<2> you MIGHT want to tell it that wchar_t is it's own type (which is the default in vc++2005)
<4> vawjrwrk: What difference would that make?
<6> well
<6> typedefs really don't create a new type
<5> Achamenes: it's for unicode.
<6> they just give you a new name for a type
<6> but the type itself is exactly what you typedefed it to.
<6> there should really be something like a strict typedef or something
<2> I concur
<6> where you actually get a new type.
<5> 6.0 even has it, vawjrwrk knows better then me of course.
<2> and neither should you
<4> davsn: See, that's exactly what confuses me. What do you mean "it's for Unicode"?
<5> vawjrwrk: you are definetely right.
<5> 2005 is great.
<4> davsn: You can work with Unicode and use char.
<2> Achamenes what is it you're wanting to do?
<5> Achamenes: http://www.cppreference.com/keywords/wchar_t.html
<4> vawjrwrk: I just want to make sense how Windows, MFC, and .NET deal with Unicode. (Even though, .NET is a lot better documented about its use of Unicode.)
<4> davsn: I know what wchar_t is, but that doesn't make anything any different. The point is that you can use chars and Unicode together just fine.
<4> am I making any sense here?
<5> i don't see the point, why you feel so problematic with unicode.
<4> the problem is not Unicode, it's how Microsoft uses the term and has implemented it that confuses me.
<4> davsn: Unicode itself is not a text encoding, right?
<2> Achamenes nope, you're not making a lot of sense
<4> OK let me try again :)
<8> hmm, filezilla is issuing some odd commands
<4> Unicode provides us with a set of tables, a set of rules, and many different algorithms to store, interpret, and render text right?
<8> like REST 0 constantly
<8> is that it's way of keeping alive the connection? that's what NOOP's for
<2> Achamenes I haven't seen the Unicode spec since around 1.1
<4> oh... it has changed a lot i suppose. the current version is 4.1, and they just released 5.0 beta 2,
<8> pfft, everything should be ASCII representable!
<4> in a nutshell, at least in the current version, Unicode itself is not an encoding scheme, it's just an abstract set of rules and descriptions of characters.
<8> want some hiragana? ASCII art baby!
<8> want some complex scientific symbols? ASCII art baby!
<9> S THERE any function which evaluates the evalution time?
<4> to encode text in unicode, that is to get a binary array representing Unicode text, you should use encoding schemes like UTF8 or UTF16.
<8> use a high resolution timer, OS dependent
<6> http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2006/04/12/BAG3KI7INT1.DTL
<6> "We're saying it's not OK to recruit on high school campuses, it's not OK to recruit on university campuses,'' Marla Zubel, a UC Santa Cruz senior and member of Students Against War, said. "In order to stop the war, you have to make it more difficult to wage war."
<2> evet
<4> not politics again :(
<2> Noidea, c'mon, it's the Republik of Kalifornia
<2> Santa Kruz for god's sake
<6> it's all politics here.
Return to
#c++
or
Go to some related
logs:
#chatzone
malta isaac tattoo
noorx
dead rat alpha Boot floppy
scott aki
irene dumalan
#chatzone
#linux
#java
#linux
|
|