| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Comments:
<0> didn't understand well your sentence
<1> well, you certainly aren't just writing random stuff in programs and hoping it turns out to be legal C++... so the question is .... what things are you using to try to learn C++ ... where do you look to find out what the language is?
<0> i look on msdn
<1> that's not the best place to try to learn the language
<1> I recommend the books at http://rudbek.com/books.html
<0> ah thanks
<0> i'll see if i can buy one or two of them
<2> vawjrwrk: what do you do for a living?
<1> tell computers where to go
<3> Personally, I think that recommending books only works for people who are looking for books. It takes way too much time and effort to work through those books, so if people just want to start coding and learn on the way, they will never actually read the books anyway. In that case, a good reference might be what they need/want.
<1> Run I think you're full of **** in this instance
<0> i prefer eBooks
<1> zindo sorry, there aren't any _good_ ones on the topic of C++
<0> but books are always welcome :)
<4> Run: a book can be a good reference
<4> vawjrwrk: the standard is an ebook, no one sensible buys the deadtree version instead of the pdf :P
<5> :)
<4> though there's the new hardbound version that's reasonable
<1> The problem with "just do it", Run, is that people never learn about some stuff they really need to learn, because they never know it exists...so they can't go look for it
<3> I bought a reference from the ANSI/ISO site, but it's too large to really use as everyday reference -- it's more ment for compiler writers etc. For daily use I use a reference that I bought from dinkum.com. Both references are in digital form, the latter is clickable even (it's html).
<4> I've used dinkum's online reference when I didn't want to open the pdf
<3> vawjrwrk: You're totally right in that. I'm just saying that if they are not looking for books, they will not read them anyway.
<4> after a little time though, I find the standard itself readable enough
<1> well, I try to point them in the right direction, and if they claim that they know better, I discuss, then argue it with them
<1> Kniht yes, but hardly the thing one want's to read to learn how to USE the language
<1> or to use the library
<4> my first c++ book was a donated copy of c++ for dummies
<1> mine was Stroustrup I (very poorly written if one wanted a tutorial) followed by Teale's book on IOStreams
<1> then Meyers
<3> The books you recommend are a MUST for C++ professionals... people who code in C++ for their jobs. I'm not saying the books are bad for anyone else, it will just take too much time to read them for most people. I think the average teenager interested in coding can learn some C++ from lesser sources too.
<1> name them
<1> I'll be glad to add them to my recommendations....if they're actually useful and don't teach wrong stuff
<3> Anything that allows them to start using C++
<1> AC++
<1> period!!!
<1> no exceptions
<3> lol
<1> which at least lets them start USING the language instead of slaving through char* and malloc etc just to read someone's name and print it back
<3> It took me three months to make a start with C++... That wasn't with AC++, so I can't judge how I would have felt using that instead. When I later tried to read it, I got too bored after chapter 5 or so and couldn't get myself to finish it :/. Maybe it's better for people for who everything in it is new.
<3> I'm different anyway... I rather read reference books, absorp them, than 'learning books' with examples, yuk.
<1> or those who don't ***ume they know everything... I read it when it came out... about 10 years after I started doing C++
<3> Oh, I'm sure there's lots in it that I don't know; and IF someone has the time and energy to work through them then will be know C++ better than me (not the same as being a better coder ;). But I doubt many people will.
<3> s/be//
<1> I find it helpful to see what "connections" authors make between the elements.... a "what" book can certainly be useful, but without some explanation of WHY some stuff "is" can make using it effectively difficult
<3> I do almost nothing else than coding C++, 7 days per weeks, and even I can't find the time to actually READ those books (though I bought them all)
<1> it's a very small book
<3> I'm making a paradigm shift by the way... a personal one.
<1> gonna take a couple days a week off?
<3> Since I grew up with ***embly and 8080 cpu's... it's in my blood to put time into making programs fast. Now, I think, I finally start to realize that it might be better to forget about the program being fast, and instead concentrate on writing a working program in a short time instead.
<1> substitute correct for working (unless you consider them synonomous) and I'm with you 100%
<1> let the compiler / code-generator geeks make them go fast
<2> vawjrwrk: that's why I do most work in Java now
<3> At the moment I'm writing a program that is HEAVILY using std::string ... it's rather slow I think (takes 10 seconds for something that could be done easily in 1 or even 0.1 seconds), but to write it such that it WOULD do the same in 0.1 seconds would take me a year instead of the two weeks I needed to write it now. ANd it wouldn't be maintainable.
<2> BTW, hey, Run... long time
<1> jottinger it's why I write in C++
<2> different strokes, I guess. I knew C++ back before ISO took it over; it's now a completely different (and foreign, to me) animal.
<1> I kept tripping over the casts one needed to use with containers in Java... C++ didn't have that problem
<2> vawjrwrk: those casts are now gone.
<1> yeah, but they weren't 5 years ago which was the last time I looked at switching languages
<6> Java has generics now, but if you mention using them in #java you will get bitched out :) (well at least I did)
<2> *nod* not without custom compilers
<2> pizza was around back then, but it wasn't in common use
<2> cn28h: heh
<3> hmm, RAPID Chess Tournament in 40 minutes.
<3> I suppose I should eat breakfast... its 19:20
<3> My current problem is this:
<3> I have a container with std::string filenames, with path.
<3> For example "/usr/local/gcc-graph/include/c++/3.4.4/i686-redhat-linux/bits/ctype_inline.h"
<3> A whole list of files like that.
<3> Now I want to ***ign to each a 'shortname', being the shortest path (right-hand side) that is uniq.
<3> In this case, "ctype_inline.h", but if there is any ctype_inline.h anywhere else too, then BOTH need to get one subdirectory added, thus "bits/ctype_inline.h" (and if that is still equal, another etc).
<3> I'm having a hard time thinking of a reasonable algorithm :/
<3> Obviously I can and will put all shortnames into a set<>, so I can easily detect duplicates.
<7> You'll have to relate those short names to a table of full pagts.
<7> paths
<3> But when i detect a duplicate, then I need to go back and find the previous original filename... hmm. I guess I'll just add an iterator to the set that points back to the FileName it came from.
<3> yeah
<1> Run have you considered finding the smallest result from mismatch between all pairs?
<3> For this program I've now like six cl***es that are all almost the same. Lemme put a general case of them on a paste site ...
<1> althought iirc, if you sort them you can do it in less than n^2
<1> or did I misunderstand...i.e if there's only widget.txt that will suffice for the short name
<4> why not use a trie? it could store the whole paths and give you the shortest unique string
<7> Cool... I just spotted the first robin of the year.
<7> That's *early*
<7> He probably got here early to have first dibs on the women.
<3> Pseudo code:
<3> http://www.rafb.net/paste/results/5YPUwH62.html
<3> That is, whenever I create an instance, it is automatically inserted in a set<> and keeps an iterator to itself that I can retrieve from the created instance.
<8> damn it.. I accidentally hit the "Mark as unread" button while all my messages were selected, how do you "Mark as read" in hotmail?
<7> By switching to gmail.
<8> I do have gmail as well :)
<3> hm, that should be TheCl***(container_type::iterator iter) ...
<3> bbl
<4> blitz chess over the internet doesn't do it for me
<4> although some of the +N seconds per move variants seem promising
<1> anyone know if Peter's going to be online again today
<1> Run what does making operator< friend do for you?
<2> vawjrwrk: he said around 6 his time
<9> Kia Ora
<1> jottinger thanks
<1> well, I'll just ask the question maybe someone else knows... are there any tools readily available that will locate (collect) the dlls needed to send to a customer so they can install ?
<1> we're just switching to vc8 and they're in very weird places
<10> hah, I wrote such a tool. However, it's for Linux.
<11> dependancy walker
<11> depends.exe
<9> bealtine: Did you hear, Tana Umanga has retired from International Rugby
<1> bealtine does that tell us where the file is located? or just it's filename?
<11> ja
<12> grrr, I think the gloves I ordered are too big
<12> it'd be nice if they'd standardize on the sizes of gloves
<12> these larges are HUGE compared to the other brands I tried on.
<11> filenames iirc
<13> :))))))))))
<10> Germans ****.
<4> haha
<11> you can switch on "full path" and see where the files are
<11> heh
<11> you can switch on "full path" and see where the files are
<11> vawjrwrk^
<12> hmmm
<12> stringstream
<12> you can't test it as a bool?
<12> like you can with file streams?
<10> stringstream has the same operator void*() that other streams have.
<12> weird
<12> it's telling me I can't do that with stringstream
<4> code example?
<14> NoideaWRK, hey what was the name of the make again lfs make who makes it again ?
<15> Hm, I've got a design question. Are there any good pointers to find out when a message architecture is best implemented with special message structures rather than a set of functions?
<15> IE, right now I'm making a rather simple card game with a client/server type relationship between the Game object and the Player(Which will either be a distant player, local player, or an AI). I use virtual fucntions to get responses from the Player hierarchy, and a general "receive message" one to notify them of the various status changes of the game. And I jsut got to wondering why I don't make that a set of functions, too.
<15> Both would work. Using the message struct seems like it'd make it easier to implement the distant Player object, and would improve compatibility with external clients... On the other hand, functions seem closer tot he design of C++.
<15> Anyway. Both'll work, I jsut can't seem to make up my mind about it. o_O
<16> i've been using signals & slots to notify any listeners of the object about changes
<16> (and it's working out alright)
<16> following the 'observer' design pattern
<15> Ah, ok. I'll look into that.
<15> I remember using them with Qt, but implementing them must be soemthign else. =P
<16> boost::signals
<15> I <3 boost.
<17> I don't even remember how it looks like, but who cares
<15> Eh, what the hell. I'm jsut messing around anyway, so I might jsut as well try implementing both ways.
<9> Yeah jsut do that
<18> Scripts off, please
<19> interesting question ... how can i find the physical dimensions of my monitor? i'm interested in calculating the dpi (or if i could get the dpi directly, that would be great too)
<17> You can get the dpi from the specs
<17> Because its size won't help you
<1> I suspect Apocal` wants the pixels per inch
Return to
#c++
or
Go to some related
logs:
proftp root acces
dragan_ionut
#MissKitten
#skype
#MissKitten
#mirc
#linux
#MissKitten
#linux
evii, aol
|
|