| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
<igbeserk> i need to enable my debugger i don't know how and why it got disabled
<igbeserk> it happened before and i reinstalled it and it worked again
<igbeserk> i tried reinstalling again and it was like 2 hours and it still wasn't finishing
<igbeserk> so i cancelled it
<moralcode> i use a lot of boost::shared_ptr<std::string>, is that good or bad? should i start using char* instead?
<moralcode> nevermind shared_ptr makes me feel safe :)
<james-_> moralcode: unless you have measured performance issues, I see no advantage in going to raw pointers.
<james-_> And I don't imagine that your shared_ptr<string> is going to be the performance bottleneck in most cases.
<moralcode> i see, read that boost::shared_ptr is slow on performance though
<moralcode> ah ok
<Genia4> slow? never
<Genia4> overhead, yes
<Genia4> but how many pointers are you really gonna use?
<james-_> Slow? Um... probably you're reading people ranting.
<Genia4> people will rant about anything
<Genia4> including other people's rants
<james-_> shared_ptr does use more dynamic allocation, but in most cases it just doesn't matter.
<james-_> And updating ref counts takes a non-zero number of cycles... but still not a lot.
<moralcode> good info, ill continue using shared_ptr :)
<james-_> Or if you don't need shared semantics, just p*** strings around directly.
<Genia4> there's scoped_ptr
<Genia4> which is probably marginally more efficient, although I can't say for sure
<james-_> scoped_ptr is more efficient; should be pretty much as quick as a raw pointer.
<james-_> And no space overhead unless you need a custom deleter, I'd hope (hmmmn, does scoped_ptr even have support for custom deleters? I don't use it often.)
<Genia4> scoped_ptr is good for cl***-level pointer member vars
<james-_> Genia4: True, though I rarely need dynamically allocated things as members. (Not never, just rarely.)
<Genia4> I do a lot (Qt crap :( )
<Genia4> anything that uses signals/slots is better off being dynamically allocated
<Jostein-_> hi, dunno if this is the right channel to ask in...
<Jostein-_> i've got a bunch of files which contains chinese text. all files are ansi-format, not unicode. opening them shows weird characters even when I have chinese installed on my winxp. is there a way to read the chinese text?
<Genia4> because when that kind of stuff goes out of scope, dtor will be called
<Genia4> and deleting stuff while there are signals pending is bad
<james-_> capital B, "Bad".
<james-_> Jostein: you need to know what encoding the files are in, for one. I don't think ANSI covers any Chinese character sets, though I may be wrong.
<igbeserk> so does anyone know how to enable the debugger in microsoft visual studios .net?
<moralcode> if there is an exception thrown that is unknown, is there a way to capture it? (it does not inherit from std::exception)
<Genia4> yes
<KimmoA> What's wrong with cout << "I'm alive!" << endl;?
<Genia4> catch(...)
<xdank--> set a breakpoint and run with debugger
<james-_> igbeserk: google and Microsoft's website don't help? Try #winprog.
<KimmoA> That's all the debug you need.
<igbeserk> well i'm looking in the search help for the program
<igbeserk> and it's giving me nothing
<igbeserk> i'll try google then
<Genia4> I need to start using exceptions and boost in my cod
<Genia4> e
<james-_> moralcode: you can catch(...), but you can't do much with it then except report it or rethrow it.
<KimmoA> Exceptions = code-bloaters.
<james-_> igbeserk: note: it might well be that it's Windows that you need to tell to use the debugger.
<Genia4> james-_: how can you rethrow it?
<james-_> KimmoA: Myths = bad
<james-_> Genia4: "throw;"
<Genia4> and what good will that do?
<Genia4> it will travel up anyway
<james-_> KimmoA: exceptions can make critical pieces of code smaller.
<james-_> Genia4: it's useful (rarely) if you have a reason not to use RAII for some local cleanup.
<moralcode> james-_, i already do catch(...) but i want to know what the exception is...trying to solve the problem
<james-_> Genia4: or, it can be used to bring exception handling into a function.
<KimmoA> james-_: Huh? How is that possible?
<igbeserk> james: it's worked before like when i first installed it and previously and all this is the second time it gets to the point where the debugger doesn't want to work
<james-_> KimmoA: by reducing the need for explicit tests of conditions.
<james-_> igbeserk: I heard you the first time.
<james-_> igbeserk: Windows has a framework for debugging. Visual Studio is one thing that can plugin into that.
<moralcode> im using a xerces-method and it throws an exception...i already catch all the exceptions in the library but its not one of them, and it does not inherit from std::exception
<moralcode> xdank--, im not very good with the debugger :/
<moralcode> might be time to practice now then :)
<Genia4> why was kennedy killed?
<james-_> moralcode: Windows? Sure it's not a structured exception reporting something like a bad pointer problem?
<moralcode> james-_, it might be
<Genia4> I mean, what did the killer say?
<KimmoA> Bah.
<moralcode> james-_, yes windows
<KimmoA> Java made me hate exceptions.
<Genia4> java overuses them
<james-_> Checked exceptions seem to be a failed experiment.
<Genia4> what's a checked exception?
<xdank--> forcing exception handling at compile time
<xdank--> ?
<james-_> Genia4: ones where the compiler checks your functions' throw specifications.
<Genia4> oh
<Genia4> void func() throws whatever?
<Genia4> yeah, gay
<KimmoA> Okay... It seems to ignore my #ifndef and still runs the stuff inside it and #endif... What do you put in place of "foo" in "#ifndef foo"? "FOO", "Foo", "foo", "Foo.h" or something else? I have seriously tried to find an answer online.
<Genia4> huh?
<james-_> KimmoA: what the heck are you talking about?
<james-_> KimmoA: you use #ifndef to check if something is not defined... so you have a reason for checking... so you know what to check.
<KimmoA> Hmm...
<KimmoA> I wanna check if something is included...
<james-_> What do you mean, "check if something is included"?
<moralcode> if someone has any idea: http://www.rafb.net/paste/results/O5TOU940.html
<KimmoA> If the stupid header has been included, I don't wanna include it in this file...
<james-_> KimmoA: why not? Shouldn't it be harmless?
<KimmoA> Apparently not...
<YUY0x7> #if !defined(UNIQUE_IDENTIFIER_FOR_THIS_HEADER) #define UNIQUE_IDENTIFIER_FOR_THIS_HEADER .... #endif
<james-_> KimmoA: if not, you should probably fix the header if it's yours.
<KimmoA> I get redefinition errors.
<james-_> You need include guards in the header file, as YUY0x7 showed.
<james-_> ***uming those are compile-time redefinition errors, as I guess they are, not link errors with multiply defined symbols.
<YUY0x7> moralcode, where is the documentation of XMLString::substring ?
<KimmoA> Compile-time.
<KimmoA> Right. Now I remember how they work :)
<moralcode> YUY0x7: http://xml.apache.org/xerces-c/apiDocs/cl***XMLString.html#z1009_0
<KimmoA> I ***ume that there's a good reason why this can't be done, but wouldn't it be possible for a compiler to know this somehow?
<james-_> KimmoA: to know what?
<KimmoA> To know what is safe to include.
<james-_> KimmoA: note: sometimes including a header twice is deliberate.
<YUY0x7> moralcode, the first parameter is expected to be allocated with enough space to receive the substring
<moralcode> and the memory is released later by myself, so i ***ume that the memorymanager allocates the memory for *day
<moralcode> YUY0x7, doesnt the memory manager sort that out?
<YUY0x7> moralcode, what memory manager?
<moralcode> YUY0x7, XMLPlatformUtils::fgMemoryManager
<YUY0x7> oh, nope
<moralcode> oh ok
<YUY0x7> you're p***ing a char * const
<moralcode> ah i see, i cant figure why i need to p*** the memorymanager in the first place
<moralcode> stupid library
<moralcode> cant take care of its own memory *grunt*
<YUY0x7> i'm puzzled as to why they use char* strings instead of std::string
<moralcode> me too, never going to use xerces again
<james-_> 1. Xerces was around before the standard C++ library was decently portable, and 2. XML requires decent Unicode support, that standard C++ doesn't yet have.
<james-_> Also Xerces is somewhat based on the Java version, I think.
<moralcode> i see, good explanation
<moralcode> doh
<moralcode> expat or libxml2 it is then
<james-_> It's still fugly interfacing to Xerces...
<YUY0x7> using char* strings makes it more portable to unicode?
<james-_> moralcode: if SAX2 is all you need, expat is simple and quick.
<james-_> YUYU0x7: they use their own string cl***.
<moralcode> james-_, yup thats all i need...sweet
<moralcode> i thought standards were standards because people should follow them :)
<james-_> moralcode: sometimes takes a while for compiler/library vendors to catch up. And C++98 was a little too inventive.
<KimmoA> Bah. The stupid **** ain't workin'.
<KimmoA> I do a #define FOO_H in the "main" file including the file, and then do #ifndef FOO_H\n#include "foo.h"\n#endif.
<KimmoA> I guess it's wrong too.
<moralcode> xml, std::string and unicode has been around for a while...they arent very fast
<YUY0x7> <YUY0x7> #if !defined(UNIQUE_IDENTIFIER_FOR_THIS_HEADER) #define UNIQUE_IDENTIFIER_FOR_THIS_HEADER .... #endif
<igbeserk> yea i'm finding no luck
<igbeserk> i'll just reinstall it
<james-_> KimmoA: header guards are used in the header file only; they are not normally mentioned anywhere else.
<james-_> KimmoA: in particular, do *not* define the macro anywhere else.
<KimmoA> But... but...
<KimmoA> Of course not, but I must check it somewhere...
<KimmoA> Sigh. BRB.
<james-_> moralcode: Unicode has been quite a moving target. Windows and Java have it ugly, as they use 16-bit datatypes for Unicode, but Unicode hasn't fitted into 16 bits for a number of years.
<james-_> KimmoA: you check it in the header: that's what the #ifndef line does.
<vvortex3_> are you people still here?
<moralcode> james-_, i see...wonder why developers are so slow :)
<moralcode> unicode slowing down now?
<vvortex3_> because we are payed by the hour
<moralcode> kicking out xerces as of now
Return to c++
or
Go to some related
logs:
"time of your life wav.""
politics
linux-2.6.13-mppe-mppc-1.3.patch.gz
|
|