undefined reference error
vf...@talktalk.net
like it is in a .h file. So
what can be the cause of an undefined reference error, when this .h
file containing the definition has been included and the namespace has
been declared.
I blame engineers they are all ponsey zealots.
Victor Bazarov
> I get an undefined reference error for a definition that clearly looks
> like it is in a .h file.
You cannot put "clearly" and "looks like" in the same sentence. It's
either there or it isn't.
> So
> what can be the cause of an undefined reference error, when this .h
> file containing the definition has been included and the namespace has
> been declared.
The error is on the line 42 of your source code.
> I blame engineers they are all ponsey zealots.
Yes. That's it.
V
--
Please remove capital As from my address when replying by mail
Howard
<vf...@talktalk.net> wrote in message
news:1139239775.9...@f14g2000cwb.googlegroups.com...
>I get an undefined reference error for a definition that clearly looks
> like it is in a .h file. So
> what can be the cause of an undefined reference error, when this .h
> file containing the definition has been included and the namespace has
> been declared.
>
Without seeing the code (or even the specific text of the error), it's
impossible to know. Perhaps, since you use the word "undefined", it's a
link error, not a compile error? You might have declared something but not
defined it. (There is a difference.) Or, you might not have properly
included the source file in your build.
More info, please.
> I blame engineers they are all ponsey zealots.
>
What???
-Howard
vf...@talktalk.net
look like the same reference, I cannot say for sure that it is the same
thing, one is a class member and the other is I don't know, the
definition is...
namespace audiere
{
...
class ParameterList {
public:
ParameterList
and the reference to it is said to be...
audiere::ParameterList::ParameterList(char const*)
by the compiler.
Does that satisfy your curiosity?
vf...@talktalk.net
and the header file included.
vf...@talktalk.net
> Without seeing the code (or even the specific text of the error), it's
> impossible to know.
I think that is the problem with using other peoples code and any
hidden information but then again
not many people have the time to create there own electronics,
operating system, compiler ...
Victor Bazarov
> Why do you bother responding if that is your attitude? It does clearly
> look like the same reference, I cannot say for sure that it is the same
> thing, one is a class member and the other is I don't know, the
> definition is...
>
> namespace audiere
> {
> ...
> class ParameterList {
> public:
> ParameterList
And??? Don't bother. I've lost interest in helping you.
>
> and the reference to it is said to be...
>
> audiere::ParameterList::ParameterList(char const*)
>
> by the compiler.
>
> Does that satisfy your curiosity?
Yes. Now go solve your own problem. I was curious how far your arrogance
stretched, and you've shown that it stretches too far for my taste.
If you want help, learn to supply enough information. And read the FAQ
(http://www.parashift.com/c++-faq-lite/).
vf...@talktalk.net
Daniel T.
"vf...@talktalk.net" <vf...@talktalk.net> wrote:
> I get an undefined reference error for a definition that clearly looks
> like it is in a .h file. So
> what can be the cause of an undefined reference error, when this .h
> file containing the definition has been included and the namespace has
> been declared.
Please understand, there are many of us here that are happy to help you,
however your question above is equivalent to the following one, "The
compiler says that my code is broken but everything is right. What is
the problem?"
Trust me, the compiler is right. If it says you have an "undefined
reference error for a definition" then you in fact have one no matter
what you may think.
Now, can we see the exact text of the error message? No one can help you
without it.
--
Magic depends on tradition and belief. It does not welcome observation,
nor does it profit by experiment. On the other hand, science is based
on experience; it is open to correction by observation and experiment.
Markus Moll
vf...@talktalk.net wrote:
> I am definitely not an arrogant person.
But you try hard to make other people think you are.
What do you think does "that Victor guy" sound like?
But I'm more and more getting the impression that your only interest is in
trolling... so... feel free.
Markus
vf...@talktalk.net
$ make
g++ mytest.cpp -I ../src -L ../lib -o mytest.exe
/cygdrive/c/DOCUME~1/Daniel/LOCALS~1/Temp/ccZxZ4i6.o:mytest.cpp:(.text+0x492):
u
ndefined reference to
`audiere::NullAudioDevice::create(audiere::ParameterList c
onst&)'
/cygdrive/c/DOCUME~1/Daniel/LOCALS~1/Temp/ccZxZ4i6.o:mytest.cpp:(.text+0x540):
u
ndefined reference to `audiere::ParameterList::ParameterList(char
const*)'
/cygdrive/c/DOCUME~1/Daniel/LOCALS~1/Temp/ccZxZ4i6.o:mytest.cpp:(.text$_ZN7audie
re14ThreadedDeviceC1EPNS_11AudioDeviceE[audiere::ThreadedDevice::ThreadedDevice(
audiere::AudioDevice*)]+0xae): undefined reference to
`audiere::AI_CreateThread(
void (*)(void*), void*, int)'
/cygdrive/c/DOCUME~1/Daniel/LOCALS~1/Temp/ccZxZ4i6.o:mytest.cpp:(.text$_ZN7audie
re14ThreadedDeviceD0Ev[audiere::ThreadedDevice::~ThreadedDevice()]+0x5f):
undefi
ned reference to `audiere::AI_Sleep(unsigned int)'
/cygdrive/c/DOCUME~1/Daniel/LOCALS~1/Temp/ccZxZ4i6.o:mytest.cpp:(.text$_ZN7audie
re14ThreadedDeviceD1Ev[audiere::ThreadedDevice::~ThreadedDevice()]+0x5f):
undefi
ned reference to `audiere::AI_Sleep(unsigned int)'
collect2: ld returned 1 exit status
make: *** [mytest] Error 1
vf...@talktalk.net
vf...@talktalk.net
to troll.
Howard
<vf...@talktalk.net> wrote in message
news:1139242570....@g47g2000cwa.googlegroups.com...
> "Why do you bother responding if that is your attitude?"
> was directed at that Victor guy.
>
What??? Who was _this_ statement directed to?
>> Without seeing the code (or even the specific text of the error), it's
>> impossible to know.
>
> I think that is the problem with using other peoples code and any
> hidden information but then again
> not many people have the time to create there own electronics,
> operating system, compiler ...
>
Umm, ok.... still no information here. If you want help, you've got to try
to help us help you.
You could start by quoting relevant portions of the message you're
responding to, and continue by adding any additional information you are
asked for and have access to. For more info on how to properly post to this
newsgroup, see:
http://www.parashift.com/c++-faq-lite/
(esp. section 5.8)
-Howard
vf...@talktalk.net
to my query. Where is that info you are refering to?
My qualm with this group is that you sidestep general questions,
my general question is
"what can be the cause of an undefined reference error"
question then I am not asking in the right place.
> You might have declared something but not
> defined it. (There is a difference.)
Thanks for that Howard. I checked this and it made no difference, I
have included the .h file of the cpp file that contains the definition.
The only info I've got from all this is that there are other people
hunched over there computers who don't seem to know any more than I do
about this particular error message.
Why should I have divulge my knowledge, or lack of it to elicit a
useful answer?
Why should people have to post their exact code / compiler setup to get
an answer?
Daniel T.
"vf...@talktalk.net" <vf...@talktalk.net> wrote:
OK, simple. Your compiler isn't finding the audiere object file. You
will have to ask in a newsgroup that specializes in your particular
compiler to find out how to set it up so that it can find the right file.
Default User
> If I could find some software that worked for me then I would not need
> to troll.
You can't troll people who don't see your posts.
*plonk*
Brian
--
Please quote enough of the previous message for context. To do so from
Google, click "show options" and use the Reply shown in the expanded
header.
Andre Kostur
news:1139247962.3...@g47g2000cwa.googlegroups.com:
> There are alot of sections in that faq, most of them are not relevant
> to my query. Where is that info you are refering to?
Exactly the section he specified. Section 5.8.
> My qualm with this group is that you sidestep general questions,
> my general question is
> "what can be the cause of an undefined reference error"
The general answer is that you have constructed an identifier that does
not exist in the object code. There, now wasn't that a useful answer?
> if you people are not prepared to give a straight forward answer that
> question then I am not asking in the right place.
The straightforward answer is probably useless (see the answer above).
>> You might have declared something but not
>> defined it. (There is a difference.)
>
> Thanks for that Howard. I checked this and it made no difference, I
> have included the .h file of the cpp file that contains the definition.
Then there is probably something you are failing to see. Can't see the
forest for the trees, and that sort of thing. Kindof a form of tunnel
vision where you are reading what you want to read, not what's on the
screen (note that this isn't a criticism or failing of you, I'm sure
quite a few of us have spent a long time (OK, probably the first time you
encounter it) trying to track down a compiler problem which worked out to
be a missing semicolon somewhere).
> The only info I've got from all this is that there are other people
> hunched over there computers who don't seem to know any more than I do
> about this particular error message.
Many of us know quite a bit about the error message (including Victor).
We ask the clarification questions so we can show you what's wrong.
> Why should I have divulge my knowledge, or lack of it to elicit a
> useful answer?
> Why should people have to post their exact code / compiler setup to get
> an answer?
Because "the devil is in the details". You apparently think that your
code is perfectly constructed. Your compiler disagrees. The only way
that we can "arbritrate" on who's right (and unfortunately, it's likely
the compiler...) is to see the code. At which point we can then point
out exactly what's wrong (or determine that the problem actually does lie
with the compiler).
Howard
<vf...@talktalk.net> wrote in message
news:1139247962.3...@g47g2000cwa.googlegroups.com...
> There are alot of sections in that faq, most of them are not relevant
> to my query. Where is that info you are refering to?
As I said, check section 5.8, which describes the best way to post here to
get answers to questions.
> My qualm with this group is that you sidestep general questions,
That's not our intent. We assumed you wanted specific help with your
problem, and there simply wasn't enough info to go on. So we asked for more
info.
> my general question is
> "what can be the cause of an undefined reference error"
I gave you a couple "general reasons". If that's all you wanted, then I
guess we're done here...? If you ned more help, then we need more info,
that's all.
> if you people are not prepared to give a straight forward answer that
> question then I am not asking in the right place.
>
>> You might have declared something but not
>> defined it. (There is a difference.)
>
> Thanks for that Howard. I checked this and it made no difference, I
> have included the .h file of the cpp file that contains the definition.
>
That's good, but just saying the definition is there doesn't solve the
problem. If you still need help, and you don't want to post all your code
(which is quite reasonable on your part) then your best bet is to try to
recreate the problem in a small separate project, and post _that_ code so
that we can try it out and see for ourselves.
> The only info I've got from all this is that there are other people
> hunched over there computers who don't seem to know any more than I do
> about this particular error message.
Until you actually posted the error messages, nobody knew for certain what
the error message really was. And without knowing your code, it's just too
hard to tell why you might have gotten an error.
>
> Why should I have divulge my knowledge, or lack of it to elicit a
> useful answer?
> Why should people have to post their exact code / compiler setup to get
> an answer?
>
You don't have to do anything, but if you want help, then the easier you
make it for us, the easier we can make it for you.
If you just want a "general" answer, then you might make that very explicit
in your question. We're just trying to help you find the _actual_ answer to
your question, not make random guesses. But I gave you two random guesses.
If those help, then I'm glad I could help. If not, then you need to decide
if you want more help or not, and if you do, then some more effort on your
part will be required.
-Howard
Daniel T.
"vf...@talktalk.net" <vf...@talktalk.net> wrote:
> There are alot of sections in that faq, most of them are not relevant
> to my query. Where is that info you are refering to?
> My qualm with this group is that you sidestep general questions,
> my general question is
> "what can be the cause of an undefined reference error"
> if you people are not prepared to give a straight forward answer that
> question then I am not asking in the right place.
The straight forward answer is: Your code makes a reference to something
that isn't defined.
Did you notice how easy it was for me to answer the question when you
actually produced the error message rather than the general question
above (in which the answer is obvious?)
> > You might have declared something but not
> > defined it. (There is a difference.)
>
> Thanks for that Howard. I checked this and it made no difference, I
> have included the .h file of the cpp file that contains the definition.
>
> The only info I've got from all this is that there are other people
> hunched over there computers who don't seem to know any more than I do
> about this particular error message.
We know a lot about that particular error message, unfortunately it was
like pulling teeth to get you to actually *show* us the error message.
> Why should I have divulge my knowledge, or lack of it to elicit a
> useful answer?
> Why should people have to post their exact code / compiler setup to get
> an answer?
"I have a question. What's the answer?" "Why should I have to explain
what the question is or why I am asking? Just answer the question
already!" Bet you can't answer it...
Andre Kostur
news:1139245054.4...@f14g2000cwb.googlegroups.com:
> I am only looking at the top error at this time (5 lines down)
>
> $ make
> g++ mytest.cpp -I ../src -L ../lib -o mytest.exe
> /cygdrive/c/DOCUME~1/Daniel/LOCALS~1/Temp/ccZxZ4i6.o:mytest.cpp:(.text+
> 0x492): u
> ndefined reference to
> `audiere::NullAudioDevice::create(audiere::ParameterList c
> onst&)'
> /cygdrive/c/DOCUME~1/Daniel/LOCALS~1/Temp/ccZxZ4i6.o:mytest.cpp:(.text+
> 0x540): u
> ndefined reference to `audiere::ParameterList::ParameterList(char
> const*)'
> /cygdrive/c/DOCUME~1/Daniel/LOCALS~1/Temp/ccZxZ4i6.o:mytest.cpp:(.text$
> _ZN7audie
> re14ThreadedDeviceC1EPNS_11AudioDeviceE[audiere::ThreadedDevice::Thread
> edDevice( audiere::AudioDevice*)]+0xae): undefined reference to
> `audiere::AI_CreateThread(
> void (*)(void*), void*, int)'
> /cygdrive/c/DOCUME~1/Daniel/LOCALS~1/Temp/ccZxZ4i6.o:mytest.cpp:(.text$
> _ZN7audie
> re14ThreadedDeviceD0Ev[audiere::ThreadedDevice::~ThreadedDevice()]+0x5f
> ): undefi
> ned reference to `audiere::AI_Sleep(unsigned int)'
> /cygdrive/c/DOCUME~1/Daniel/LOCALS~1/Temp/ccZxZ4i6.o:mytest.cpp:(.text$
> _ZN7audie
> re14ThreadedDeviceD1Ev[audiere::ThreadedDevice::~ThreadedDevice()]+0x5f
> ): undefi
> ned reference to `audiere::AI_Sleep(unsigned int)'
> collect2: ld returned 1 exit status
> make: *** [mytest] Error 1
>
OK, where are all of the function bodies for audiere::ParameterList ?
Also, where are the function bodies for audiere::AI_Sleep ? As Daniel C.
has mentioned, this looks like the compiler (well, linker actually) can't
find the object code for these functions/classes.
So far you've told the compiler to compile mytest.cpp, and to make an
executable. You've mentioned where your libraries are, but you haven't
told the compiler which libraries to use.
Daniel T.
Andre Kostur <nntp...@kostur.net> wrote:
> So far you've told the compiler to compile mytest.cpp, and to make an
> executable. You've mentioned where your libraries are, but you haven't
> told the compiler which libraries to use.
My guess is he never compiled the audiere library after he d/l'ed it
from sourceforge.
Daniel T.
Andre Kostur <nntp...@kostur.net> wrote:
> Then there is probably something you are failing to see. Can't see the
> forest for the trees, and that sort of thing. Kindof a form of tunnel
> vision where you are reading what you want to read, not what's on the
> screen (note that this isn't a criticism or failing of you, I'm sure
> quite a few of us have spent a long time (OK, probably the first time you
> encounter it) trying to track down a compiler problem which worked out to
> be a missing semicolon somewhere).
I remember once spending two weeks (at 40 hrs/week) tracking down an
error that resolved to a '2' in the code instead of a '4'. (It was
causing memory overwrites.) My boss was very unhappy, I'm glad I didn't
write the original code. :-/
Markus Moll
vf...@talktalk.net wrote:
> If I could find some software that worked for me then I would not need
> to troll.
Thanks for the confirmation.
Welcome to my kill file.
*plonk*
Andre Kostur
news:postmaster-CE40C...@news.east.earthlink.net:
> In article <Xns9762676555D63...@207.35.177.134>,
> Andre Kostur <nntp...@kostur.net> wrote:
>
>> Then there is probably something you are failing to see. Can't see
>> the forest for the trees, and that sort of thing. Kindof a form of
>> tunnel vision where you are reading what you want to read, not what's
>> on the screen (note that this isn't a criticism or failing of you,
>> I'm sure quite a few of us have spent a long time (OK, probably the
>> first time you encounter it) trying to track down a compiler problem
>> which worked out to be a missing semicolon somewhere).
>
> I remember once spending two weeks (at 40 hrs/week) tracking down an
> error that resolved to a '2' in the code instead of a '4'. (It was
> causing memory overwrites.) My boss was very unhappy, I'm glad I
> didn't write the original code. :-/
My favourite is spending a day and a half trying to figure out why my
program was crashing when I was calling a 3rd party library. After large
amounts of frustration and finally resorting to printf's all over the
place (OK, this was originally in C, but is perfectly applicable to C++),
I found out that one of my pointers was NULL. So, as I was yelling to
myself "That pointer can't POSSIBLY be NULL!!! I check it for NULL right
THERE!" as I pointed to the following short code block:
if (ptr = NULL)
{
return;
}
I then proceeded to profusely swear at myself....... (this was before the
compiler would warn me about those sorts of accidental code
constructs....)
Stephan Brönnimann
> The error is on the line 42 of your source code.
It's now the 2nd time within a few days I've seen it.
Could I use my personal favorite 27 instead (smiling)?
Cheers, Stephan
Howard
"Stephan Brönnimann" <bro...@hotmail.com> wrote in message
news:1139254644.5...@z14g2000cwz.googlegroups.com...
Check (or Google) "The Hitchhiker's Guide to the Galaxy". It's a _very_
important number! :-)
-Howard
Stephan Brönnimann
Thanks Howard, I got it. I've read all the books but I had to google
Stephan :-)
Stephan Brönnimann
> I am only looking at the top error at this time (5 lines down)
>
> $ make
> g++ mytest.cpp -I ../src -L ../lib -o mytest.exe
> /cygdrive/c/DOCUME~1/Daniel/LOCALS~1/Temp/ccZxZ4i6.o:mytest.cpp:(.text+0x492):
> u
> ndefined reference to
> `audiere::NullAudioDevice::create(audiere::ParameterList c
> onst&)'
> collect2: ld returned 1 exit status
> make: *** [mytest] Error 1
I was in this thread for about 10 minutes wondering if ever you'd post
the compiler messages. Toghether with your statement "all the library
stuff is there" the problem is immediately clear after I see the
command line: You never link against the library.
Regards, Stephan
red floyd
> [story redacted]
>
> if (ptr = NULL)
> {
> return;
> }
>
> I then proceeded to profusely swear at myself....... (this was before the
> compiler would warn me about those sorts of accidental code
> constructs....)
Reminds me of the very first piece of code I wrote as a professional (as
opposed to a student. This was back in '84, so it was C, not C++...
int index;
/* ... */
for (index = 0 ; index < NUM_ELEMS; ++index);
{
some_array[index].somefield = INITIAL_VALUE;
}
You could hear the swearing four corridors down when I finally found the
bug... :)
red floyd
Perhaps as an alternative, we could tell people that they have errors on
lines 4, 8, 15, 16, 23, and 42?
vf...@talktalk.net
"audiere.h" and in audiere.cpp there is a reference to device.cpp and
device.cpp is included at the top of audiere.cpp.
vf...@talktalk.net
> So far you've told the compiler to compile mytest.cpp, and to make an
> executable. You've mentioned where your libraries are, but you haven't
> told the compiler which libraries to use.
My guess is he never compiled the audiere library after he d/l'ed it
from sourceforge"
There are different versions of the source.
1 is does not compile withoug java, xpcom and VC6. At least that is if
you want to bind it to java and xpcom, VC6 I don't know. Another file
has the windows source and this already has executables and a lib file,
do I need to recompile this?
vf...@talktalk.net
Feb 6, 6:45 pm show options
Thanks for the confirmation. "
How can your file welcome me? You use too many similies or labels,
talk in English! I don't understand what you are on about.
vf...@talktalk.net
g++ mytest.cpp -I ../src ../lib/audiere.lib -o mytest.exe
or
g++ mytest.cpp -I ../src -L ../lib/audiere.lib -o mytest.exe
neither of these worked there were still unresolved references.
I think the problem is that
there are includes off includes in this project and these are not being
linked.
Andre Kostur
news:1139363799....@g43g2000cwa.googlegroups.com:
> So how do I link a file ? In this case say that I am including
> "audiere.h" and in audiere.cpp there is a reference to device.cpp and
> device.cpp is included at the top of audiere.cpp.
Now you're getting beyond the scope of this newsgroup. Time to head over
to gnu.gcc.help for more detailed (and implementation-specific) answers to
your questions. comp.lang.c++ is for discussing the language itself. How
to use a specific compiler is better served in a newsgroup dedicated to
that compiler.
vf...@talktalk.net
42 is the answer to everything in the story, but if I remember
correctly it originates in from a game of scrabble. I guess Adams got
the idea from 41 which has some unusual mathematical properties. Phhh,
what the hell do I know anyway?
Jim Langston
<vf...@talktalk.net> wrote in message
news:1139366374....@g44g2000cwa.googlegroups.com...
Umm.. no. Arther Dent got the question to the answer 42 from a game of
scrabble. Seems the earth was just a hugh computer built to find out what
the true question to the answer 42 was. Kinda funny actually, I won't say
what the true question was cause it's a spoiler, and people really should
read the Hitchhikers trilogy (with 4 books in it).
vf...@talktalk.net
simulation say like a climate model for a period of time and then look
at the difference between the climate model and the simulation then in
a way... like...
Sjouke Burry
a story about a gigacomputer calculating the answer
to everything.
It tells them : 42.
When people complained about the answer, the computer
told them they should have asked the right question.
vf...@talktalk.net
first?