| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10
Comments:
<0> another of the many ways they serve the cl*** of wealth and greed
<1> MishMan and what if i can't do better? becouse i can asure that it will use less "jumps" using the goto statement... and the code should be smaller too, is it still bad to use it?
<2> [E]dvinas: all c can be written without goto, if you can't do better than goto, you're a bad programmer
<3> [E]dvinas: I,personally, n3ever di backward goto, but I do forward goto albeit sparingly
<3> s/di/do/
<1> i need to do forward...
<3> you always can add flags to avoid goto, the question is ...
<1> i need to do a variable jump after the given parameter...
<4> Ok, so what would you propose for GUI programming in linux ***uming that I more or less know C, and have engine programmed in C ?
<3> care to post code ?
<4> Tcl/Tk ?
<3> i'll tell you if goto can be avoided or not
<5> Tcl rocks !
<1> MishMan i'm only planing to write it...
<3> but then the discussion doesn't make sense
<3> i rewrite code 20 times
<3> some versions are with goto, some without
<3> until i reach perfection
<6> What's the best way to p*** "bool foobar[ m ][ n ]" to a function, where m and n are variable? I thought I would just "bool *foo", but inside the function, I cannot access foo[ 1 ][ 3 ] or the likes. I tried "bool **foo", but I don't believe that is what I want either. Any suggestions?
<7> MishMan, i like using goto instead of flags.
<1> but what's the bad side of goto/label? in executing/compiling such code, besides that i may be a bad programmer becouse i do so?
<8> float: They're really variable? As in you're using C99's VLAs?
<3> OT yes, this is often smaller code
<7> you're a bad programmer if you refuse to use goto.
<6> Xirtam: funny :P
<1> so goto it is :P
<1> :)
<9> [E]dvinas: Wrong use makes it unreadable and unmaintainable. Just as with every other construct, but it's particulary easy with goto.
<1> i mean will be...
<8> Float: It's not supposed to be funny. It was supposed to elicit a real answer.
<7> if you use goto sparingly for the purpose of the behavior of the code clearer then you've done a good thing.
<10> I need to p*** a pointer to a linked list node into one function, which will in turn call a 2nd function, p*** it into that and have it modify the value of list *node->next. so in main if I have list *node; and I call function1(&node); and then function1 calls function2() how do I p*** the address of node into function2 so function2() can modify the value of node->next ?
<1> not that my method will make the code clearlier but also will make it smaller and i think that the real compiled and ***embled code will be smaller
<3> Float_: depends on how you allocate the 2d array
<3> [E]dvinas yuo can never say like this until you actually write and rewrite the code
<7> the "goto considered harmful" article is something that people never apparently read. it basically says don't use goto for doing loops when constructs like while and for exist.
<3> i don;t believe you write finally perfect code from the 1st time w/o rewriting and polishing
<6> Xirtam: Sorry, no I am not using C99's VLAs
<1> i do rewrite my code...
<3> so how you allocate this array
<8> Float: Then you'll have to explain how you declared your array.
<1> today i've done it 6 times already
<1> ok, i've got my answer... thanks :)
<3> i wish language had goto-with-parameters
<3> ilke goto(x,y,"abc");
<7> MishMan, tail-recursion optimization can do that.. :)
<3> and label had formal arguments like: mylabel(x,y,str)
<3> just as syntax sugar
<3> there are often soe ***ginments before goto
<11> MishMan: those are called "functions"
<11> hehe
<3> no no
<3> functions, you call then
<11> I know.
<3> labels, you goto them
<11> <-- was being a smart***
<11> see? ***, smart.
<12> MishMan: how is that different from functions?
<3> this is smart ***: http://www.stanwelltrailers.co.uk/images/smart2.jpg
<3> mauke: because they are not callable, just labels inside functions
<12> they're "callable" by goto
<3> they are goto-able
<7> does myrddin ever read his privmsgs?
<11> whoops.
<7> ehhe
<11> hehe
<12> MishMan: great, so you have a new kind of functions that is incompatible with normal functions
<7> it took me a while to get used to irssi's privmsgs
<3> the syntax is goto label(args);
<11> lemme employ my privmsg log reader (ie. grep)
<3> you don't understand
<12> no, you don't understand
<7> i wish C have native coroutines.
<3> int foo() { int err=0; ...; goto fail("message"); ... fail(str): return -1; }
<3> i just sometimes do it with macros
<3> so that ***ignment + goto takes 1 line
<12> replace goto fail("message") by return fail("message")
<12> surprise, it's a function
<3> the problem if you have common cleanup before all those returns .... then you prefer the goto
<11> aw, crap. I forgot I restarted this session. no log for the last day and a half. :(
<7> MishMan, i think it should be goto label(expression); where an expression can contain a lambda
<12> MishMan: that's also incompatible
<10> I want to modify a linked list node p***ed two functions deep, but I'm p***ing it wrong and not sure where, any ideas? - http://www.rafb.net/paste/results/g0mYhf72.html
<7> oh! and you should be allowed to change the namespace as part of a goto.
<11> I should have my privmsgs go to a different logfile anyway
<7> myrddin, aww.
<12> MishMan: labels with arguments shouldn't be reachable by normal control flow
<12> that's pretty much the opposite problem of goto foo; { int i = 42; foo: i; }
<12> only here it's { foo(i); i; }
<12> s/;/:/
<3> mauke: why ?
<12> because i doesn't exist there
<12> it's created by a goto foo(42);
<3> with label parameters, i'd not declare the type but only mention the var name, the var itself is declared elsewhere like:
<3> int foo() { char *msg=""; goto fail("error");... fail(msg): print(msg); return -1; }
<12> that's even more evil
<3> maybe
<3> C is evil
<12> implicit ***ignment ...
<3> function call is implicit asgnment, too, for every arg
<12> yeah, it should be implicit binding, though
<12> (call by value is evil)
<12> MishMan: it's initialization, not ***ignment
<3> maybe int foo() { char *msg=""; int rc; goto fail(msg="error",rc=-1); ... fail: print(msg); return rc; }
<3> this is pretty easy to do with macro
<3> #define GOTO_AND_***IGN(label,expr) do { expr; goto label; }while(0)
<3> np
<3> #define ***IGN_AND_GOTO(label,expr) do { expr; goto label; }while(0)
<12> who submitted OrngeTide to qdb.us?
<3> which one
<3> normally, it's me
<12> <7> if you use goto sparingly for the purpose of the behavior of the
<12> +code clearer then you've done a good thing.
<3> me
<12> why?
<3> i like it
<3> what's your problem
<12> it doesn't even make sense (there's a missing "making")
<12> and it's not funny
<7> yea. my english ****s many.
<3> it's words of wisdom
<7> soldier boy! me love you long time.
<3> OT: i bought the parts for that rs232-controlled relay
<7> MishMan, cool
<3> i will be soldering some day
<3> i'm not sure the relay i got is right but i'll see
<7> mauke, there are tons of qdb.us quotes of me. i think all submitted by Mish
<3> wat's max323s
<3> mauke: i don't edit them before posting. that's why it's called "quote"
<7> MishMan, it's TTL to RS232 signal convertor.
<10> what does this mean from gcc: warning: void format, pointer arg (arg 2) The code for the line is: printf("stuff %p\n", ptr); ptr is of type char **
<12> it means *ptr should have the type void, not char *
<10> why should it be void, when I am p***ing it the address of a char *
<3> strange i dont get such warning
<12> because %p expects a void *
<3> even with -Wall
<3> something is fishy here
<10> well perhaps more of the code is relevant than I posted
<12> no, just fix the type
<12> (void *)ptr
<10> just cast it in the printf statements
<3> 'void format' ? %p is not 'void format'
<12> it is, if you dereference one level
<3> ncaller_: let's see if mauke's cast helps
<10> they do in fact get rid of the warning
<3> mauke: where do you see the latest quotes in qd.us ?
<12> qdb.us/queue
Return to
#c
or
Go to some related
logs:
#beginner
#cisco
windows xp memory limit
#windows
#beginner
ddrvtt overclock
nerif.pk
dkestop
#goal
#computers
|
|
