| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Comments:
<0> Nothing gets my morning started like a discussion of implicit statics.
<1> :)
<2> what`s an array do?
<0> Nothing, on its own.
<0> But it's typically used to keep an easily indexed collection of items.
<2> ok
<3> _m_: still here? declaring "l" as "static const int l" and initializing via "const mycl***::l = 1024" doenst work either. just gives me the same error
<4> 'morning
<1> suriel: 12:52 <@_m_> (that needs to be in the cl*** definition)
<5> why bother?
<1> I.e. like: struct foo { static const int l = 1024; };
<5> just to declare a dynamic array?
<1> It's probably not that dynamic.
<5> well no :)
<5> boost::array springs to mind
<3> _m_: "static const l = 100" gives me compiler error: error C2258: Ungueltige Syntax fr rein virtuelle Methode; '= 0' erforderlich
<4> Try using a type
<0> What compiler are you using?
<3> _m_: that means: invalid syntax for a pure virtual method. "=0" needed
<4> learn to read
<0> If it's VC6, then you're screwed, and you need to upgrade.
<3> Solamente: yup. vc6
<0> Upgrade.
<0> VC6 ****s as a C++ compiler.
<4> lemme paste what _m_ said first and then what you tried:
<5> vc6
<5> ick
<0> That's legal code, but VC6 doesn't realize it.
<4> static const int l = 1024;
<3> Run: i meant "static const int l=1000", ok?
<4> static const l = 100
<1> VC6 does not support static member constants.
<5> calc vc2005e
<6> vc2005e = Visual C++ 2005 Express (aka VC++ 8). Available for download at - http://msdn.microsoft.com/vstudio/express/visualc/ . Get the Tools Refresh at http://www.microsoft.com/downloads/details.aspx?FamilyID=afd04ff1-9d16-439a-9a5e-e13eb0341923&displaylang=en
<1> I suggest you use a C++ compiler instead.
<3> Run: just forgot to write the "int" here. everybody else managed to understand it
<4> suriel: what ok? you tried that or not?
<0> Run: He did, and VC6 barfed on it.
<4> suriel: I'd like you to appologise of that tone...
<4> "everybody else managed to understand it" is not what I like to hear when I try to help someone.
<0> Run, I know you're trying to make a point, but that's really not helping.
<4> feel free to remove the ban
<3> Run: "learn to read" is not what i like to hear either when i just forgot a word
<0> Just drop it.
<5> ok children...back off
<4> learn to paste then
<5> vc6 is not a c++ compiler
<4> How are we supposed to know whether you're too stupid to type the 'int' in your program or that you're too stupid to type the 'int' on irc? I fail to see the difference.
<3> so: thanks to everyone who helped
<0> Dammit, Run, we've all mis-typed, and I had no trouble figuring out what he had done just by reading the error. Now let it go.
<3> Run: so your hint for me was "learn to read" ok, thanks then
<0> suriel: The VC6 work-around for this limitation is to use an enum instead.
<0> enum { foobar = 1024 };
<3> Solamente: hm... ok, ill try that too
<0> But, really, you should ditch VC6.
<7> suriel: It's good advice, if you don't know how to read, but then again - IRC is a written medium, so chances are you do sorta know already
<5> considering vc8 is free
<0> Especially if this is a personal project, because Visual C++ Express 2005 is free.
<0> If it's a professional project, we'd love to chat with your boss.
<3> Solamente: i'd like to switch to vc7 or something but its not that easy to switch the project to it
<0> I'
<5> ummm why not?
<0> I'm talking about VC8, acutally.
<0> But it's really quite simple to switch.
<0> You import the project, and build.
<0> 7 or 8
<5> and vc8 finds 100s more "bugs" than vc6
<4> Another advise I could give is to use your mouse to copy&paste things between IRC and the window where you're compiling stuff. It is seldom useful to retype a compiler error here and not showing _exactly_ what the compiler says.
<3> Solamente: just tried to switch to vc7 (vs2003 i think) and the converted project was not compilable
<5> compatible?
<5> what does that mean?
<1> Note, VC7.0 isn't much better than VC6.
<0> Eh, pfft.
<5> 7.1 was :)
<0> 2002, yeah. 2003, miles ahead.
<1> Yes, 7.1. is a lot better.
<3> _m_: hm... maybe i'll give vc8 another chance but since im not the only one working on this project it may get hard anyway...
<5> i'm still curious as to what "compatible" means here
<1> I doubt it will get more "compilable" with a more recent version of VC++.
<3> _m_: by vc8 you mean the vc included in vs2005?
<0> yes
<1> yes
<5> if the compiler issues lots of warnings...then you really should fix them
<1> If it doesn't compile then it's likely because you wrote code that depends on VC6's brokenness.
<3> _m_: that may be true but i cant change most of it because it simply isnt my code
<7> whose is it?
<3> jottinger: what do you want to hear? his name?
<5> tell him to move foreard into the 21st century
<5> forward
<3> bealtine: he quittet...
<7> suriel: dunno - a general description is probably enough. "My professor's." or "My homework-doin friend's." or "My company's code."
<5> if he quit...take ownership
<3> bealtine: and the responsibility of the millions of lines code he produced?
<5> sigh...stay in your 1997 box then
<1> millions of VC6 dependent lines?
<1> Get yourself a gun.
<3> bealtine: try to explain my boss what i did the last weeks, when all still works as ever but i build nothing new
<5> you will have to do it sometime soon
<3> what else than static membervariables has vc8 to offer? if i go and tell this would be a great thing, they laugh at me
<5> PTS...more compliant code...
<5> template that work
<5> support for boost and loki stuff
<3> my boss saays "if it runs, its compliant enough"
<5> he's an idiot
<3> bealtine: but hes the idiot who gives me money
<7> find a different idiot?
<3> i just did
<7> hahaha
<5> vc6 issues are rapidly being forgotten
<5> and fixes
<3> well.. at least i can try to install vc8 and comvert the project. if it works _simply_, maybe i have a chance
<5> i remember wrestling with that enum issue but only after Sola mentioned it
<0> suriel: Basically, with a standards-conforming compiler you can download a C++ library or type in code from a current book and it'll actually *work*
<0> Library writers used to write hacks specifically for VC6, but that's becoming rare now.
<0> I've quit, myself.
<3> install is running...
<4> Under what circumstances is it required to type a 'template' keyword in the middle of nested name? Like: Foo::template Bar<int>::x
<4> hmm, I suppose only when that template dependent (and thus inside another template).
<4> Hmm, Foo has to be dependent
<4> But in error messages, g++ doesn't print the 'template':
<4> F {typename D::A<j>::d_t B<j, D>::f(int) [with int j = 5, D = Foo]} {troep.cc:10}
<4> That's wrong thus... A is inside D, which is dependent. This should be printed as D::template A<j>::etc, right?
<8> if my function accepts a short* .. how do i correctly cast an unsigned char** to that?
<8> (short*) doesn't seem to work
<8> i was thinking (short*)*uchardptr
<8> bu that doesn't work either..
<4> It should work (though I wouldn't call it correct)
<4> What is the type of *uchardptr ?
<9> Apocal` changing the "level" of dereference is somewhat problematic
<8> vawjr: you mean from a logistics point of view?
<10> also by "casting"
<4> vawjr: Hi, I need my compiler to generate a few complex types, for testing purposes. In particular, I'd like to see a type that contains an expressions (would be between square brackets), but I can't think of a way to generate that :/
<4> I'm pretty sure that it's possible to have types that look like... say... <int i, int Foo[i + 5]>
<0> Only if i is const.
<0> Or was that a template arg list?
<4> yes
<0> I'm unsure about arrays as template parameters. It might be possible, though.
<4> I found an example in the demangler testsuite :)
<4> template<int I> void foo (int (&)[I + 1]) { };
<4> template void foo<2> (int (&)[3]);
<4> That generates the test case (which I must be able to parse):
<4> F {void foo(int (&)[(I + 1)]) [with int I = 2]} {troep.cc:1}
<4> I think that I'll just parse whatever is between the [ ] by reading up till the next ], without caring what it means :/
<10> Apocal` what makes you beleive that you _can_ change a char** to a short* and what do you expect to happen when you do?
<10> one presumes that if one [] is legal, then multiple would be, including [....[..] .]
<4> yes it is
<4> I make something up that works for that too.
<11> vawjr, can i join #programming?
<12> Off to office
Return to
#c++
or
Go to some related
logs:
#AllNiteCafe
cangorous in Australia
#linux
spog boo
#MissKitten
#AllNiteCafe
could not convert javascript argument serialize
#mirc
wget flood
#java
|
|