Re: Cygwin installation problem (Igor Gierymski)

[Javier Revilla]

> You hit the nail the on the head. I had to edit the Makefile in the
> test-suite directory. I changed
> LDFLAGS = -L /usr/local/lib
> to
> LDFLAGs = -L/usr/local/lib
>
> And now it builds. Thanks alot!

Hello,

I have followed this thread with great interest because I have the same problem
Igor had.

I am compiling QuantLib-0.3.7 with gcc-3.3.3 on CygWin.
Configure is ok.
Make fails at the test-suite stage.

Before compilation I have done :
export CPPFLAGS="-I/usr/local/include/boost-1_31"
export LDFLAGS="-L/usr/local/lib"

My QuantLib test-suite Makefile has: LDFLAGs = -L/usr/local/lib

I am unable to compile Boost's test-suite.

You can find below, the error I receive after doing make on QuantLib.
Thank you very much for your help.

Javier



(Error message: begin)
Making all in test-suite

make[1]: Entering directory `/home/Administrator/jrr/c++/quantlib_project/ql/Qua
ntLib-0.3.7/test-suite'

/bin/bash ../libtool --mode=link g++  -g -O2  -L/usr/local/lib -o quantlib-test-
suite.exe -lboost_unit_test_framework-gcc quantlibtestsuite.o americanoption.o a
sianoptions.o barrieroption.o basketoption.o calendars.o capfloor.o cliquetoptio
n.o compoundforward.o covariance.o dates.o daycounters.o digitaloption.o distrib
utions.o dividendeuropeanoption.o europeanoption.o factorial.o forwardoption.o i
nstruments.o integrals.o interpolations.o jumpdiffusion.o lowdiscrepancysequence
s.o matrices.o mersennetwister.o operators.o piecewiseflatforward.o quantooption
.o quotes.o riskstats.o rounding.o solvers.o stats.o swap.o swaption.o termstruc
tures.o old_pricers.o utilities.o ../ql/libQuantLib.la ../functions/ql/Functions
/libQuantLibFunctions.la

g++ -g -O2 -o quantlib-test-suite.exe quantlibtestsuite.o americanoption.o asian
options.o barrieroption.o basketoption.o calendars.o capfloor.o cliquetoption.o
compoundforward.o covariance.o dates.o daycounters.o digitaloption.o distributio
ns.o dividendeuropeanoption.o europeanoption.o factorial.o forwardoption.o instr
uments.o integrals.o interpolations.o jumpdiffusion.o lowdiscrepancysequences.o
matrices.o mersennetwister.o operators.o piecewiseflatforward.o quantooption.o q
uotes.o riskstats.o rounding.o solvers.o stats.o swap.o swaption.o termstructure
s.o old_pricers.o utilities.o  -L/usr/local/lib -lboost_unit_test_framework-gcc
../ql/.libs/libQuantLib.a ../functions/ql/Functions/.libs/libQuantLibFunctions.a

quantlibtestsuite.o(.text+0xb99): In function `_ZNSs12_S_constructIPcEES0_T_S1_R
KSaIcESt20forward_iterator_tag':
/usr/include/c++/3.3.3/bits/basic_string.h:325: undefined reference to `RiskStat
isticsTest::suite()'

collect2: ld returned 1 exit status

make[1]: *** [quantlib-test-suite.exe] Error 1
make[1]: Leaving directory `/home/Administrator/jrr/c++/quantlib_project/ql/Quan
tLib-0.3.7/test-suite'
make: *** [all-recursive] Error 1

(Error message: end)

[cuchulainn]

Life is easier if you don't have to use Makefiles!!

 

Why do people need makefiles? Am I missing something?

 

best

 

Daniel

 

---

From: [hidden email] on behalf of Javier Revilla
Sent: Sat 21/08/2004 12:39
To: [hidden email]
Subject: [Quantlib-users] Re: Cygwin installation problem (Igor Gierymski)

> You hit the nail the on the head. I had to edit the Makefile in the
> test-suite directory. I changed
> LDFLAGS = -L /usr/local/lib
> to
> LDFLAGs = -L/usr/local/lib
>
> And now it builds. Thanks alot!

Hello,

I have followed this thread with great interest because I have the same problem
Igor had.

I am compiling QuantLib-0.3.7 with gcc-3.3.3 on CygWin.
Configure is ok.
Make fails at the test-suite stage.

Before compilation I have done :
export CPPFLAGS="-I/usr/local/include/boost-1_31"
export LDFLAGS="-L/usr/local/lib"

My QuantLib test-suite Makefile has: LDFLAGs = -L/usr/local/lib

I am unable to compile Boost's test-suite.

You can find below, the error I receive after doing make on QuantLib.
Thank you very much for your help.

Javier



(Error message: begin)
Making all in test-suite

make[1]: Entering directory `/home/Administrator/jrr/c++/quantlib_project/ql/Qua
ntLib-0.3.7/test-suite'

/bin/bash ../libtool --mode=link g++  -g -O2  -L/usr/local/lib -o quantlib-test-
suite.exe -lboost_unit_test_framework-gcc quantlibtestsuite.o americanoption.o a
sianoptions.o barrieroption.o basketoption.o calendars.o capfloor.o cliquetoptio
n.o compoundforward.o covariance.o dates.o daycounters.o digitaloption.o distrib
utions.o dividendeuropeanoption.o europeanoption.o factorial.o forwardoption.o i
nstruments.o integrals.o interpolations.o jumpdiffusion.o lowdiscrepancysequence
s.o matrices.o mersennetwister.o operators.o piecewiseflatforward.o quantooption
.o quotes.o riskstats.o rounding.o solvers.o stats.o swap.o swaption.o termstruc
tures.o old_pricers.o utilities.o ../ql/libQuantLib.la ../functions/ql/Functions
/libQuantLibFunctions.la

g++ -g -O2 -o quantlib-test-suite.exe quantlibtestsuite.o americanoption.o asian
options.o barrieroption.o basketoption.o calendars.o capfloor.o cliquetoption.o
compoundforward.o covariance.o dates.o daycounters.o digitaloption.o distributio
ns.o dividendeuropeanoption.o europeanoption.o factorial.o forwardoption.o instr
uments.o integrals.o interpolations.o jumpdiffusion.o lowdiscrepancysequences.o
matrices.o mersennetwister.o operators.o piecewiseflatforward.o quantooption.o q
uotes.o riskstats.o rounding.o solvers.o stats.o swap.o swaption.o termstructure
s.o old_pricers.o utilities.o  -L/usr/local/lib -lboost_unit_test_framework-gcc
../ql/.libs/libQuantLib.a ../functions/ql/Functions/.libs/libQuantLibFunctions.a

quantlibtestsuite.o(.text+0xb99): In function `_ZNSs12_S_constructIPcEES0_T_S1_R
KSaIcESt20forward_iterator_tag':
/usr/include/c++/3.3.3/bits/basic_string.h:325: undefined reference to `RiskStat
isticsTest::suite()'

collect2: ld returned 1 exit status

make[1]: *** [quantlib-test-suite.exe] Error 1
make[1]: Leaving directory `/home/Administrator/jrr/c++/quantlib_project/ql/Quan
tLib-0.3.7/test-suite'
make: *** [all-recursive] Error 1

(Error message: end)



-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
Quantlib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users

[Zhou Wu-2]

Daniel,

What I wanted to say that this was an open world.
Frequently we wanted to use tools and libraries from
others. But others will have their own ways to create
these things. For example, boost (quantlib depends on)
uses jam to build. I cannot imagine makefiles will go
away any time soon, just as you claimed the C++ would
stay for another twenty years in your book. This does
not mean I like it, but we have to live with it.

I use both command line tools and ide for my work: IDE
for writing and debugging the code; Command lines for
integration, nightly build, unit testing, etc. I
mostly lived in Java world for my job in the past
several years. Building environment for java is much
easier (It may be because I'm familiar with it).

Zhou


--- "Daniel J. Duffy" <[hidden email]> wrote:

> Zhou,
>  
> Sorry, I don't understand your answer. I cannot
> interpolate to the current discussion
>  
> Getting back to my point, VC++ takes care of all
> makefile for you. I think I can remember that
> something similar is possible with gnu and
> KDEDevelop.
>  
> Are you working with a command line compiler?
>  
> regards
>  
> Daniel
>
> ________________________________
>
> From: Zhou Wu [mailto:[hidden email]]
> Sent: Sat 21/08/2004 20:50
> To: Daniel J. Duffy
> Subject: RE: [Quantlib-users] Re: Cygwin
> installation problem (Igor Gierymski)
>
>
>
>
>   This is not an ideal world. Just as you live in
> one
> country you are happy with, but you have to abide by
> the laws or rules when you visit other countries.
> You
> may have pleasures too if ...
>
>   Zhou
>
>
>
> --- "Daniel J. Duffy" <[hidden email]> wrote:
>
> > Life is easier if you don't have to use
> Makefiles!!
> >
> > Why do people need makefiles? Am I missing
> > something?
> >
> > best
> >
> > Daniel
> >
> >
> > ________________________________
> >
> > From: [hidden email]
> on
> > behalf of Javier Revilla
> > Sent: Sat 21/08/2004 12:39
> > To: [hidden email]
> > Subject: [Quantlib-users] Re: Cygwin installation
> > problem (Igor Gierymski)
> >
> >
> >
> > > You hit the nail the on the head. I had to edit
> > the Makefile in the
> > > test-suite directory. I changed
> > > LDFLAGS = -L /usr/local/lib
> > > to
> > > LDFLAGs = -L/usr/local/lib
> > >
> > > And now it builds. Thanks alot!
> >
> > Hello,
> >
> > I have followed this thread with great interest
> > because I have the same problem
> > Igor had.
> >
> > I am compiling QuantLib-0.3.7 with gcc-3.3.3 on
> > CygWin.
> > Configure is ok.
> > Make fails at the test-suite stage.
> >
> > Before compilation I have done :
> > export CPPFLAGS="-I/usr/local/include/boost-1_31"
> > export LDFLAGS="-L/usr/local/lib"
> >
> > My QuantLib test-suite Makefile has: LDFLAGs =
> > -L/usr/local/lib
> >
> > I am unable to compile Boost's test-suite.
> >
> > You can find below, the error I receive after
> doing
> > make on QuantLib.
> > Thank you very much for your help.
> >
> > Javier
> >
> >
> >
> > (Error message: begin)
> > Making all in test-suite
> >
> > make[1]: Entering directory
> >
> `/home/Administrator/jrr/c++/quantlib_project/ql/Qua
> > ntLib-0.3.7/test-suite'
> >
> > /bin/bash ../libtool --mode=link g++  -g -O2
> > -L/usr/local/lib -o quantlib-test-
> > suite.exe -lboost_unit_test_framework-gcc
> > quantlibtestsuite.o americanoption.o a
> > sianoptions.o barrieroption.o basketoption.o
> > calendars.o capfloor.o cliquetoptio
> > n.o compoundforward.o covariance.o dates.o
> > daycounters.o digitaloption.o distrib
> > utions.o dividendeuropeanoption.o europeanoption.o
> > factorial.o forwardoption.o i
> > nstruments.o integrals.o interpolations.o
> > jumpdiffusion.o lowdiscrepancysequence
> > s.o matrices.o mersennetwister.o operators.o
> > piecewiseflatforward.o quantooption
> > .o quotes.o riskstats.o rounding.o solvers.o
> stats.o
> > swap.o swaption.o termstruc
> > tures.o old_pricers.o utilities.o
> > ../ql/libQuantLib.la ../functions/ql/Functions
> > /libQuantLibFunctions.la
> >
> > g++ -g -O2 -o quantlib-test-suite.exe
> > quantlibtestsuite.o americanoption.o asian
> > options.o barrieroption.o basketoption.o
> calendars.o
> > capfloor.o cliquetoption.o
> > compoundforward.o covariance.o dates.o
> daycounters.o
> > digitaloption.o distributio
> > ns.o dividendeuropeanoption.o europeanoption.o
> > factorial.o forwardoption.o instr
> > uments.o integrals.o interpolations.o
> > jumpdiffusion.o lowdiscrepancysequences.o
> > matrices.o mersennetwister.o operators.o
> > piecewiseflatforward.o quantooption.o q
> > uotes.o riskstats.o rounding.o solvers.o stats.o
> > swap.o swaption.o termstructure
> > s.o old_pricers.o utilities.o  -L/usr/local/lib
> > -lboost_unit_test_framework-gcc
> > ../ql/.libs/libQuantLib.a
> >
>

../functions/ql/Functions/.libs/libQuantLibFunctions.a

> >
> > quantlibtestsuite.o(.text+0xb99): In function
> > `_ZNSs12_S_constructIPcEES0_T_S1_R
> > KSaIcESt20forward_iterator_tag':
> > /usr/include/c++/3.3.3/bits/basic_string.h:325:
> > undefined reference to `RiskStat
> > isticsTest::suite()'
> >
> > collect2: ld returned 1 exit status
> >
> > make[1]: *** [quantlib-test-suite.exe] Error 1
> > make[1]: Leaving directory
> >
>

`/home/Administrator/jrr/c++/quantlib_project/ql/Quan
> > tLib-0.3.7/test-suite'
> > make: *** [all-recursive] Error 1
> >
> > (Error message: end)
> >
> >
> >
> >
>
-------------------------------------------------------
> > SF.Net email is sponsored by Shop4tech.com-Lowest
> > price on Blank Media
> > 100pk Sonic DVD-R 4x for only $29 -100pk Sonic
> DVD+R
> > for only $33
> > Save 50% off Retail on Ink & Toner - Free Shipping
> > and Free Gift.
> >
>
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
> > _______________________________________________
> > Quantlib-users mailing list
> > [hidden email]
> >
>
https://lists.sourceforge.net/lists/listinfo/quantlib-users

> >
> >
> >
>
>
>
>                
> _______________________________
> Do you Yahoo!?
> Win 1 of 4,000 free domain names from Yahoo! Enter
> now.
> http://promotions.yahoo.com/goldrush
>
>

=== message truncated ===



               
__________________________________
Do you Yahoo!?
Yahoo! Mail - You care about security. So do we.
http://promotions.yahoo.com/new_mail

[Luigi Ballabio-2]

On 2004.08.21 12:39, Javier Revilla wrote:
> I am compiling QuantLib-0.3.7 with gcc-3.3.3 on CygWin.
> Configure is ok.
> Make fails at the test-suite stage.
>
> /usr/include/c++/3.3.3/bits/basic_string.h:325: undefined reference  
> to `RiskStatisticsTest::suite()'

Javier,
        I'm kind of puzzled. RiskStatisticsTest::suite() is defined in  
riskstats.cpp, and your log shows that you did compile it and that the  
resulting riskstats.o is being linked with the rest.  Maybe the object  
files got messed up during a previous try?  It's just a lame  
suggestion, but what if you try a 'make clean' and rebuild the thing  
again?

Later,
        Luigi

[Luigi Ballabio-2]

On 2004.08.21 12:47, Daniel J. Duffy wrote:
> Life is easier if you don't have to use Makefiles!!

Daniel,
        life would be easier if there were a single platform :)
On Unix-like platforms, we're using the so-called GNU autotools  
(autoconf, automake and libtool) as they are, to date, the most  
complete answer to the problem of providing portability across  
different systems---proper unixes such as Solaris, Mac OS X, Linux on a  
number of processors besides x86...  The user can just uncompress the  
source tarball and type

./configure ; make

The configure step detects the current platform, checks for needed  
libraries and tools and automatically generates makefiles whose rules  
are aware of most (if not all) particular quirks of the platform. The  
configure script is itself automatically generated from the autotools,  
based on some info given by the developer (not much more than the list  
of source files.)

So it's not like we're maintaining Makefiles--as you suggest, that  
would be a _major_ headache...  we just have to keep the source list up  
to date, but that's no more difficult that adding new files to a VC++  
project.

Later,
        Luigi

[Javier Revilla]

Quoting Luigi Ballabio <[hidden email]>:

> On 2004.08.21 12:39, Javier Revilla wrote:
> > I am compiling QuantLib-0.3.7 with gcc-3.3.3 on CygWin.
> > Configure is ok.
> > Make fails at the test-suite stage.
> >
> > /usr/include/c++/3.3.3/bits/basic_string.h:325: undefined reference
> > to `RiskStatisticsTest::suite()'
>
> Javier,
> I'm kind of puzzled. RiskStatisticsTest::suite() is defined in
> riskstats.cpp, and your log shows that you did compile it and that the
> resulting riskstats.o is being linked with the rest.  Maybe the object
> files got messed up during a previous try?  It's just a lame
> suggestion, but what if you try a 'make clean' and rebuild the thing
> again?
>
> Later,
> Luigi
>

Hello Luigi,
You were right. I made a clean and then rebuilt it and it works.
I only got two warnings but that should be ok.
Curious how the files got messed up.

Thanks.
Have a nice day,
Javier

[Luigi Ballabio-2]

On 2004.08.23 15:53, Javier Revilla wrote:
> Hello Luigi,
> You were right. I made a clean and then rebuilt it and it works.
> I only got two warnings but that should be ok.

Javier,
        what were the warnings? We try to remove them whenever it's  
possible...

Thanks,
        Luigi

[Javier Revilla]

Quoting Luigi Ballabio <[hidden email]>:

> On 2004.08.23 15:53, Javier Revilla wrote:
> > Hello Luigi,
> > You were right. I made a clean and then rebuilt it and it works.
> > I only got two warnings but that should be ok.
>
> Javier,
> what were the warnings? We try to remove them whenever it's
> possible...
>
> Thanks,
> Luigi
>
>

I get the following message :

make[2]: Entering directory
`/home/Administrator/jrr/c++/quantlib_project/ql/QuantLib-0.3.7'

WARNING: Warnings can be ignored. :-)
if test "no" != no; then \
  set x; \
  list='quantlib.el'; for p in $list; do \
    if test -f "$p"; then d=; else d="./"; fi; \
    set x "$@" "$d$p"; shift; \
  done; \
  shift; \
  EMACS="no" /bin/bash ./config/elisp-comp "$@" || exit 1; \
else : ; fi

make[2]: Leaving directory
`/home/Administrator/jrr/c++/quantlib_project/ql/QuantLib-0.3.7'


Thanks,
Javier

[Luigi Ballabio-2]

On 2004.08.23 16:29, Javier Revilla wrote:

>
> I get the following message :
>
> make[2]: Entering directory
> `/home/Administrator/jrr/c++/quantlib_project/ql/QuantLib-0.3.7'
>
> WARNING: Warnings can be ignored. :-)
> if test "no" != no; then \
>   set x; \
>   list='quantlib.el'; for p in $list; do \
>     if test -f "$p"; then d=; else d="./"; fi; \
>     set x "$@" "$d$p"; shift; \
>   done; \
>   shift; \
>   EMACS="no" /bin/bash ./config/elisp-comp "$@" || exit 1; \
> else : ; fi
>
> make[2]: Leaving directory
> `/home/Administrator/jrr/c++/quantlib_project/ql/QuantLib-0.3.7'

Oh, THAT warning. It can indeed be ignored, as there are no further  
warnings to ignore...

Thanks,
        Luigi

[Javier Revilla]

Quoting Luigi Ballabio <[hidden email]>:

> On 2004.08.23 16:29, Javier Revilla wrote:
> >
> > I get the following message :
> >
> > make[2]: Entering directory
> > `/home/Administrator/jrr/c++/quantlib_project/ql/QuantLib-0.3.7'
> >
> > WARNING: Warnings can be ignored. :-)
> > if test "no" != no; then \
> >   set x; \
> >   list='quantlib.el'; for p in $list; do \
> >     if test -f "$p"; then d=; else d="./"; fi; \
> >     set x "$@" "$d$p"; shift; \
> >   done; \
> >   shift; \
> >   EMACS="no" /bin/bash ./config/elisp-comp "$@" || exit 1; \
> > else : ; fi
> >
> > make[2]: Leaving directory
> > `/home/Administrator/jrr/c++/quantlib_project/ql/QuantLib-0.3.7'
>
> Oh, THAT warning. It can indeed be ignored, as there are no further
> warnings to ignore...
>
> Thanks,
> Luigi
>

Thanks to warn me that there are no further warnings (to ignore).
:-)

The road goes on.
Now, I'll compile my 1st test program !

Thanks for your help.
Javier