Nabble - CINT

Re: CINT 5.24 build query

2009-12-16T05:43:14Z

Hi,

install cygwin (cygwin.com). Run the Visual Studio Command Prompt from
the Visual Studio Tools in the start / programs menu. In that terminal,
start C:\cygwin\cygwin.bat. Now cd to /cygdrive/c/CINT (i.e. C:\CINT) or
wherever you installed CINT. Now run ./configure && make. Let me know
how this worked for you.

An alternative approach is a project file provided by Louis, an external
contributer, see <http://root.cern.ch/phpBB2/viewtopic.php?t=8935> I
haven't tried it and we only provide support for the first method.

Cheers, Axel.

Niharika01 wrote on 12/16/2009 01:21 PM:

>
>
> Hi All
>
>
>
> We have to build CINT 5.24 for the highest version of Visual Studio with
> which CINT 5.24 is compatible.
>
>
>
> We have downloaded the source code of CINT 5.24 from the SVN repository
> link http://root.cern.ch/svn/root/tags/v5-24-00b .
>
>
>
> The Release Notes of CINT 5.24 mention about ‘adding proper support for
> Microsoft Visual C++ 9.0’.
>
>
>
> Does it mean that CINT 5.24 is compatible with Microsoft Visual Studio
> 2008 (having internal version number as 9.0)?
>
>
>
> Also, we are not able to build CINT.
>
>
>
> Earlier, we had built 5.16.1 by following the steps mentioned below.
>
>
>
> 1. Copy source code of CINT (to be built) to "C:\CINT"
>
> 2. Added environment variables "CINTSYSDIR=C:\CINT" and "PATH=%
> CINTSYSDIR%;%PATH%".
>
> 3. Executed "C:\Program Files\Microsoft Visual
> Studio\VC98\bin\vcvars32.bat"
>
> 4. Executed setup.bat or setupXP.bat from
> "C:\cint\platform\visualCpp6"
>
>
>
> Since the directory structure and source code of CINT have undergone
> major changes, we believe the procedure to build CINT has also changed.
>
>
>
> Please guide us about building CINT version 5.24.
>
> Also please let us know the latest version of Visual Studio with which
> CINT 5.24 is compatible.
>
>
>
> Regards
>
> Niharika
>
>
>

Re: Re: Re: [ROOTDEV] I have questions about cint, could you help me?

2009-11-26T06:53:53Z

Hi Kelvin,

Sorry, I didn't see your email immediately; my spam filter didn't like
it :-(

wangzju wrote on 01/-10/-28163 08:59 PM:
> Thank you very much for your quick reply. I downloaded a too old
> version, that's why I can't make it successfully. When I downloaded the
> latest version by SVN, all the probelms disappeared. You did a very good
> job. CINT is so great! Thanks again.

Thanks!

> How can I port CINT to platforms that have no operating system? Is it
> possible? I don't need muti-thread mechanism of CINIT, so whether is
> it enough to provide a c stdandard lib? That means I provide c standard
> lib of a platform, then I make CINT run on it. Is that possible?

A tool chain would also help :-) I.e. somehow you need to build it. And
CINT is C++, so you'll need a C++ std lib, too.

Good luck!

Cheers, Axel.

>
>
> 2009-11-24
> wangzju
> *发件人：* Axel Naumann
> *发送时间：* 2009-11-22 20:33:35
> *收件人：* 'wangzju'
> *抄送：* Masaharu Goto; 'cint'; 'rootdev'
> *主题：* Re: [ROOTDEV] I have questions about cint, could you help me?
> Hi Kelvin,
> we of course also support stand-alone CINT! We run nightly tests on it,
> making sure it works as expected.
> If you take a recent version (ideally one from subversion, because we
> have neglected to release binaries for almost three years now - sorry
> about that) you can simply run ./configure; make. If your system comes
> with libreadline-dev then CINT can use it.
> See <http://root.cern.ch/cint> on how to get CINT from subversion and
> how to build it; it's really easy:
> $ svn co http://root.cern.ch/svn/root/trunk/cint
> $ cd cint
> $ ./configure
> $ make -j2
> I will upload a new source and binary release in January.
> Cheers, Axel.
> On 2009-11-22 10:16, Masaharu Goto wrote:
> >
> >
> > Hello Kelvin,
> >
> >
> >
> > I guess you downloaded CINT from CERN. Now , there are 2 versions of
> > CINT, standalone and built-in component in ROOT. I think CERN only
> > supports ROOT built-in version. Platform directory and the README
> > information is described about standalone version. You have a choice
> > between standalone version and ROOT/CINT version. I am keeping a bit old
> > but still working standalone version which I can send you in 11Mbyte
> > tar.gz archive if you want.
> >
> >
> >
> > Thank you
> >
> > Masaharu Goto
> >
> >
> >
> >
> >
> > * From: * owner-rootdev@... [mailto:owner-rootdev@...]
> > *On Behalf Of *wangzju
> > *Sent:* Saturday, November 21, 2009 9:35 PM
> > *To:* cint; rootdev
> > *Subject:* [ROOTDEV] I have questions about cint, could you help me?
> >
> >
> >
> > hi,
> >
> > I am a undergradute in Zhejiang Univ. I want to use cint for academic
> > purpose. According to readme file, there should be a folder called
> > platform. but I can't find it in the latest version of cint. Could you
> > tell me how I build cint and makecint? I try to run ./configure and then
> > make on Linux. But error occures:
> > make: *** No rule to make target `readline/*.c', needed by `readline/libreadline.a'. Stop.
> >
> >
> >
> >
> > I need your help. Thanks a lot.
> >
> >
> >
> > Kelvin Wang.
> >
> >
> >
> >
> >
> > 2009-11-21
> >
> > wangzju
> >

Re: Cint7 Status

2009-11-25T08:24:50Z

Hi,

the correct link is
<http://indico.cern.ch/materialDisplay.py?contribId=3&materialId=slides&confId=71078>

Cheers, Axel.

Philippe Canal wrote on 11/24/2009 10:17 PM:

> Dear friends, supporters and early adopters of Cint[7],
>
> First let us thank you for your support, your patience and your help
> while we were developing and testing our tentative merge of Cint and
> Reflex (code named Cint 7).
>
> After analyzing the results of our integration tests, we came to the
> conclusion that Cint 7 would not provide any immediate and visible
> gain in performance or memory use while also not helping us achieve any
> the benefits we were aiming for to improve Cint for the LHC experiments;
> Cint 7 was also not yet to the point of being able to improve on any of
> the Cint 5 deficiencies or on the lack of support for the upcoming
> C++ standard. Hence we decided to no longer pursue this direction.
>
> As an alternative we started investigating LLVM (<http://llvm.org>), a
> cutting edge popular open source compiler project that is already used
> in commercial environments. Apple is currently supporting the main
> developers and relying on LLVM for Mac OS 10.6 and above. Its
> strengths include binary platform compatibility, a C++ API, large number
> of optimizer, and a just-in-time compiler, amongst others; these
> strengths makes it an ideal candidate for building upon to develop a new
> interpreter. The LLVM team is currently implementing a promising C/C++
> and Objective C/C++ front end (Clang) which can already parse more than
> 95% of the ROOT header files.
>
> After working with the LLVM and Clang team for a while, we are confident
> that they will deliver a very good C++ compiler. We believe that we can
> re-use large parts of it to produce a production-grade interpreter with
> C++ API, leveraging the large LLVM developer community. Thanks to LLVM,
> this future interpreter would offer better and faster parsing and
> execution (for both the dictionary generation and the interpretation).
> We are currently exploring the implementation of the interpreter and
> reflection database based on LLVM with an early prototype, code named
> Cling. Amongst other things, it demonstrate the usability of LLVM's
> just-in-time compiler.
>
>
> You can see more details on the Cint7 conclusion as well as LLVM and Cling
> in Axel's presentation:
> <http://indico.cern.ch/materialDisplay.py?contribId=3&materialId=slides&confId=7107>
>
>
> We hope that you can understand our decision, and that you will continue
> to support us, be it with CINT or Cling.
>
> Cheers,
> The CINT Team.
>
>
>
>
>
>

Cint7 Status

2009-11-24T13:17:55Z

Dear friends, supporters and early adopters of Cint[7],

First let us thank you for your support, your patience and your help
while we were developing and testing our tentative merge of Cint and
Reflex (code named Cint 7).

After analyzing the results of our integration tests, we came to the
conclusion that Cint 7 would not provide any immediate and visible
gain in performance or memory use while also not helping us achieve any
the benefits we were aiming for to improve Cint for the LHC experiments;
Cint 7 was also not yet to the point of being able to improve on any of
the Cint 5 deficiencies or on the lack of support for the upcoming
C++ standard. Hence we decided to no longer pursue this direction.

As an alternative we started investigating LLVM (<http://llvm.org>), a
cutting edge popular open source compiler project that is already used
in commercial environments. Apple is currently supporting the main
developers and relying on LLVM for Mac OS 10.6 and above. Its
strengths include binary platform compatibility, a C++ API, large number
of optimizer, and a just-in-time compiler, amongst others; these
strengths makes it an ideal candidate for building upon to develop a new
interpreter. The LLVM team is currently implementing a promising C/C++
and Objective C/C++ front end (Clang) which can already parse more than
95% of the ROOT header files.

After working with the LLVM and Clang team for a while, we are confident
that they will deliver a very good C++ compiler. We believe that we can
re-use large parts of it to produce a production-grade interpreter with
C++ API, leveraging the large LLVM developer community. Thanks to LLVM,
this future interpreter would offer better and faster parsing and
execution (for both the dictionary generation and the interpretation).
We are currently exploring the implementation of the interpreter and
reflection database based on LLVM with an early prototype, code named
Cling. Amongst other things, it demonstrate the usability of LLVM's
just-in-time compiler.

You can see more details on the Cint7 conclusion as well as LLVM and Cling
in Axel's presentation:
<http://indico.cern.ch/materialDisplay.py?contribId=3&materialId=slides&confId=7107>

We hope that you can understand our decision, and that you will continue
to support us, be it with CINT or Cling.

Cheers,
The CINT Team.

Re: [ROOTDEV] I have questions about cint, could you help me?

2009-11-22T04:32:41Z

Hi Kelvin,

we of course also support stand-alone CINT! We run nightly tests on it,
making sure it works as expected.

If you take a recent version (ideally one from subversion, because we
have neglected to release binaries for almost three years now - sorry
about that) you can simply run ./configure; make. If your system comes
with libreadline-dev then CINT can use it.

See <http://root.cern.ch/cint> on how to get CINT from subversion and
how to build it; it's really easy:

$ svn co http://root.cern.ch/svn/root/trunk/cint
$ cd cint
$ ./configure
$ make -j2

I will upload a new source and binary release in January.

Cheers, Axel.

On 2009-11-22 10:16, Masaharu Goto wrote:

>
>
> Hello Kelvin,
>
>
>
> I guess you downloaded CINT from CERN. Now , there are 2 versions of
> CINT, standalone and built-in component in ROOT. I think CERN only
> supports ROOT built-in version. Platform directory and the README
> information is described about standalone version. You have a choice
> between standalone version and ROOT/CINT version. I am keeping a bit old
> but still working standalone version which I can send you in 11Mbyte
> tar.gz archive if you want.
>
>
>
> Thank you
>
> Masaharu Goto
>
>
>
>
>
> * From: * owner-rootdev@... [mailto:owner-rootdev@...]
> *On Behalf Of *wangzju
> *Sent:* Saturday, November 21, 2009 9:35 PM
> *To:* cint; rootdev
> *Subject:* [ROOTDEV] I have questions about cint, could you help me?
>
>
>
> hi,
>
> I am a undergradute in Zhejiang Univ. I want to use cint for academic
> purpose. According to readme file, there should be a folder called
> platform. but I can't find it in the latest version of cint. Could you
> tell me how I build cint and makecint? I try to run ./configure and then
> make on Linux. But error occures:
> make: *** No rule to make target `readline/*.c', needed by `readline/libreadline.a'. Stop.
>
>
>
>
> I need your help. Thanks a lot.
>
>
>
> Kelvin Wang.
>
>
>
>
>
> 2009-11-21
>
> wangzju
>

RE: [ROOTDEV] I have questions about cint, could you help me?

2009-11-22T01:16:49Z

Hello Kelvin,

I guess you downloaded CINT from CERN. Now , there are 2 versions of CINT, standalone and built-in component in ROOT. I think CERN only supports ROOT built-in version. Platform directory and the README information is described about standalone version. You have a choice between standalone version and ROOT/CINT version. I am keeping a bit old but still working standalone version which I can send you in 11Mbyte tar.gz archive if you want.

Thank you

Masaharu Goto

From: owner-rootdev@... [mailto:owner-rootdev@...] On Behalf Of wangzju
Sent: Saturday, November 21, 2009 9:35 PM
To: cint; rootdev
Subject: [ROOTDEV] I have questions about cint, could you help me?

hi,

I am a undergradute in Zhejiang Univ. I want to use cint for academic purpose. According to readme file, there should be a folder called platform. but I can't find it in the latest version of cint. Could you tell me how I build cint and makecint? I try to run ./configure and then make on Linux. But error occures: make: *** No rule to make target `readline/*.c', needed by `readline/libreadline.a'. Stop.

I need your help. Thanks a lot.

Kelvin Wang.

2009-11-21

wangzju

Re: Error: Return type mismatch. basic_ostream > not a public base of string

2009-11-19T07:53:14Z

Hi Joost,

good point! And I'm back to investigating...

Cheers, Axel.

Joost Kraaijeveld wrote on 11/17/2009 03:08 PM:

> Hi Axel,
>
> On Tue, 2009-11-17 at 14:54 +0100, Axel Naumann wrote:
>> Hi Joost,
>>
>> as you probably already guessed, CINT is not able to interpret
>>
>> "std::copy(v1.begin(), v1.end(), std::ostream_iterator<
>> std::pair<unsigned long, std::string > >(std::cout, \"\\n\"));"
>>
>> You'll have to use a less STLly version of it...
>
> OK, that's a pity and a bit strange as it works in an ordinary Cint
> script (see the program source).
>
> Is there any additional documentation available about limitations on STL
> usage (or other parts), other than the rather generic comments in
> http://root.cern.ch/viewvc/trunk/cint/doc/limitati.txt ?
>
> TIA
>

Re: Error: Return type mismatch. basic_ostream > not a public base of string

2009-11-17T06:08:48Z

Hi Axel,

On Tue, 2009-11-17 at 14:54 +0100, Axel Naumann wrote:
> Hi Joost,
>
> as you probably already guessed, CINT is not able to interpret
>
> "std::copy(v1.begin(), v1.end(), std::ostream_iterator<
> std::pair<unsigned long, std::string > >(std::cout, \"\\n\"));"
>
> You'll have to use a less STLly version of it...

OK, that's a pity and a bit strange as it works in an ordinary Cint
script (see the program source).

Is there any additional documentation available about limitations on STL
usage (or other parts), other than the rather generic comments in
http://root.cern.ch/viewvc/trunk/cint/doc/limitati.txt ?

TIA

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

Re: Error: Return type mismatch. basic_ostream > not a public base of string

2009-11-17T05:54:00Z

Hi Joost,

as you probably already guessed, CINT is not able to interpret

"std::copy(v1.begin(), v1.end(), std::ostream_iterator<
std::pair<unsigned long, std::string > >(std::cout, \"\\n\"));"

You'll have to use a less STLly version of it...

Cheers, Axel.

Joost Kraaijeveld wrote on 11/17/2009 02:40 PM:

> Hi,
>
>
> I am trying to run a C++ script by creating a string in memory with the
> source (so not in a file/script).
>
> Running the program from the link results in an error when calling the
> function callMemoryString:
>
> + callMemoryString()
> tempFileName: (tmp0)
> + test()
> 471
> texttexttext
> Error: G__getvariable: expression FILE:(tmp0) LINE:9
> Error: Return type mismatch. basic_ostream<char,char_traits<char> > not a public base of string FILE:(tmp0) LINE:9
> !!!Dictionary position rewound... !!!Error recovered!!!
> 0result G__exec_text( test()): 11357936
> + callMemoryString()
>
> Anyone any idea?
>
> TIA
>
>
> The code and Eclipse CDT project are available here:
>
> http://trac.askesis.nl/downloads/CintTest.tar.bz2
>
> Unzip and run the setup script file in the root after adapting the
> "export CINTSYSDIR=/home/jkr/Programming/C++/cint" into something
> suitable for your machine.
>

Error: Return type mismatch. basic_ostream > not a public base of string

2009-11-17T05:40:46Z

Hi,

I am trying to run a C++ script by creating a string in memory with the
source (so not in a file/script).

Running the program from the link results in an error when calling the
function callMemoryString:

+ callMemoryString()
tempFileName: (tmp0)
+ test()
471
texttexttext
Error: G__getvariable: expression FILE:(tmp0) LINE:9
Error: Return type mismatch. basic_ostream<char,char_traits<char> > not a public base of string FILE:(tmp0) LINE:9
!!!Dictionary position rewound... !!!Error recovered!!!
0result G__exec_text( test()): 11357936
+ callMemoryString()

Anyone any idea?

TIA

The code and Eclipse CDT project are available here:

http://trac.askesis.nl/downloads/CintTest.tar.bz2

Unzip and run the setup script file in the root after adapting the
"export CINTSYSDIR=/home/jkr/Programming/C++/cint" into something
suitable for your machine.

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

Re: G__load_text problems

2009-11-17T05:34:33Z

Hi Axel,

On Mon, 2009-11-16 at 16:56 +0100, Axel Naumann wrote:
> you're missing the newlines in the text source, at least after the
> #include line. This works for me:
>
> "#include <iostream> \n"
> "void test() \n"
> "{\n"
> "std::cout << \"test()\" << std::endl;\n"
> "}\n";

OK, this works. Thanks.

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

Re: G__load_text problems

2009-11-16T07:55:23Z

Hi Joost,

you're missing the newlines in the text source, at least after the
#include line. This works for me:

"#include <iostream> \n"
"void test() \n"
"{\n"
"std::cout << \"test()\" << std::endl;\n"
"}\n";

Cheers, Axel.

Joost Kraaijeveld wrote on 10/31/2009 08:23 AM:

> Hi,
>
> I am trying to run a null terminated string a source code (for the full
> source of my program see below). It fails however with the following
> output:
>
> + callMemoryString()
> tempFileName: (tmp0)
> result G__exec_text( test()): 0
> Error: Function test() is not defined in current scope + callMemoryString()
> FILE:(tmpfile) LINE:2
> !!!Dictionary position rewound... !!!Error recovered!!!
>
> Using this as actual source code works:
>
> std::string code =
> "void test() "
> "{"
> "printf(\"Kees Kroket\\n\");"
> "}
>
> What is the correct way of doing this?
>
> TIA
>
>
> /////////////////////////////// The code
>
> #include <iostream>
> #include "G__ci.h"
> #include <string>
>
> void callMemoryString()
> {
> std::cout << "+ callMemoryString()" << std::endl;
>
> int state = G__init_cint( "cint");
> switch (state)
> {
> case G__INIT_CINT_SUCCESS:
> {
> std::string code =
> "#include <iostream> "
> "void test() "
> "{"
> "std::cout << \"test()\" << std::endl;"
> "}";
>
> const char* tempFileName = G__load_text(code.c_str());
> if(tempFileName)
> {
> std::cout << "tempFileName: " << tempFileName << std::endl;
>
> long result = 0;
> G__value resultValue;
>
> resultValue = G__exec_text( "test()");
> result= G__int(resultValue);
> std::cout << "result G__exec_text( test()): " << result << std::endl;
>
> G__unloadfile(tempFileName);
> }else
> {
> std::cout << "Could not create a temp file" << std::endl;
> }
>
> break;
> }
> case G__INIT_CINT_SUCCESS_MAIN:
> {
> std::cout << "G__INIT_CINT_SUCCESS_MAIN" << std::endl;
> break;
> }
> case G__INIT_CINT_FAILURE:
> {
> std::cout << "G__INIT_CINT_FAILURE" << std::endl;
> break;
> }
> default:
> {
> std::cout << "G__init_cint default" << std::endl;
> }
> }
> G__scratch_all();
>
> std::cout << "+ callMemoryString()" << std::endl;
> }
> int main(int argc, char** argv)
> {
> callMemoryString();
> return 0;
> }
>

Re: G__load_text problems

2009-11-04T07:19:31Z

Hi Joost,

we got your email but sadly none of us has had the time yet to look at
the issue you reported :-( I expect that we will react by the end of the
week. Apologies for the delay.

Cheers, Axel.

Joost Kraaijeveld wrote on 10/31/2009 08:23 AM:

G__load_text problems

2009-10-31T00:23:26Z

Hi,

I am trying to run a null terminated string a source code (for the full
source of my program see below). It fails however with the following
output:

+ callMemoryString()
tempFileName: (tmp0)
result G__exec_text( test()): 0
Error: Function test() is not defined in current scope + callMemoryString()
FILE:(tmpfile) LINE:2
!!!Dictionary position rewound... !!!Error recovered!!!

Using this as actual source code works:

std::string code =
"void test() "
"{"
"printf(\"Kees Kroket\\n\");"
"}

What is the correct way of doing this?

TIA

/////////////////////////////// The code

#include <iostream>
#include "G__ci.h"
#include <string>

void callMemoryString()
{
std::cout << "+ callMemoryString()" << std::endl;

int state = G__init_cint( "cint");
switch (state)
{
case G__INIT_CINT_SUCCESS:
{
std::string code =
"#include <iostream> "
"void test() "
"{"
"std::cout << \"test()\" << std::endl;"
"}";

const char* tempFileName = G__load_text(code.c_str());
if(tempFileName)
{
std::cout << "tempFileName: " << tempFileName << std::endl;

long result = 0;
G__value resultValue;

resultValue = G__exec_text( "test()");
result= G__int(resultValue);
std::cout << "result G__exec_text( test()): " << result << std::endl;

G__unloadfile(tempFileName);
}else
{
std::cout << "Could not create a temp file" << std::endl;
}

break;
}
case G__INIT_CINT_SUCCESS_MAIN:
{
std::cout << "G__INIT_CINT_SUCCESS_MAIN" << std::endl;
break;
}
case G__INIT_CINT_FAILURE:
{
std::cout << "G__INIT_CINT_FAILURE" << std::endl;
break;
}
default:
{
std::cout << "G__init_cint default" << std::endl;
}
}
G__scratch_all();

std::cout << "+ callMemoryString()" << std::endl;
}
int main(int argc, char** argv)
{
callMemoryString();
return 0;
}

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

Re: Return vaue from library problem with embedded cint

2009-10-30T10:06:11Z

Hi Joost!

Joost Kraaijeveld wrote on 10/30/2009 05:15 PM:

> On Fri, 2009-10-30 at 10:47 +0100, Axel Naumann wrote:
>> Hi Joost!
>>
>> Thanks for your example - it forced me to fix two bugs in CINT just to
>> build it :-)
>>
>> Now that I have it I can even tell you why that happens: you're missing
>> a dictionary for the vector<pair<...> >. I got it to work by simply
>> adding in Library.h:
>>
>> #include <vector>
>> #include <map>
>> #include <string>
>>
>> using namespace std;
>> #pragma link C++ class vector< pair< unsigned long, string > >;
>
> Cool, thanks. Do I understand it correctly that this means that I should
> always link an instantiated STL class if I use one?

CINT is pretty limited in its abilities to instantiate a
library-compatible STL class. So: yes :-(

Cheesr, Axel.

Re: Return vaue from library problem with embedded cint

2009-10-30T09:15:34Z

On Fri, 2009-10-30 at 10:47 +0100, Axel Naumann wrote:

> Hi Joost!
>
> Thanks for your example - it forced me to fix two bugs in CINT just to
> build it :-)
>
> Now that I have it I can even tell you why that happens: you're missing
> a dictionary for the vector<pair<...> >. I got it to work by simply
> adding in Library.h:
>
> #include <vector>
> #include <map>
> #include <string>
>
> using namespace std;
> #pragma link C++ class vector< pair< unsigned long, string > >;

Cool, thanks. Do I understand it correctly that this means that I should
always link an instantiated STL class if I use one?

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

Re: Return vaue from library problem with embedded cint

2009-10-30T02:47:38Z

Hi Joost!

Thanks for your example - it forced me to fix two bugs in CINT just to
build it :-)

Now that I have it I can even tell you why that happens: you're missing
a dictionary for the vector<pair<...> >. I got it to work by simply
adding in Library.h:

#include <vector>
#include <map>
#include <string>

using namespace std;
#pragma link C++ class vector< pair< unsigned long, string > >;

Cheers, Axel.

Joost Kraaijeveld wrote on 10/29/2009 12:09 PM:

> Hi,
>
> I have made a little program that crashes. It consists of an application
> that calls functions in a user library. It looks as if values returned
> from the library get lost somewhere. Is it actually possible to return
> values from a user library?
>
>
> The code and Eclipse CDT project are available here:
>
> http://trac.askesis.nl/downloads/CintTest.tar.bz2
>
> Unzip and run the setup script file in the root after adapting the
> "export CINTSYSDIR=/home/jkr/Programming/C++/cint" into something
> suitable for your machine.
>
> TIA
>

Return vaue from library problem with embedded cint

2009-10-29T04:09:32Z

Hi,

I have made a little program that crashes. It consists of an application
that calls functions in a user library. It looks as if values returned
from the library get lost somewhere. Is it actually possible to return
values from a user library?

The code and Eclipse CDT project are available here:

http://trac.askesis.nl/downloads/CintTest.tar.bz2

Unzip and run the setup script file in the root after adapting the
"export CINTSYSDIR=/home/jkr/Programming/C++/cint" into something
suitable for your machine.

TIA

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

Re: Can I use std::string in an external library interface?

2009-10-29T02:00:56Z

Hi,

CINT removed the namespace std by default. You have two options: build
CINT with -DG__STD_NAMESPACE (which is not exercised regularly and might
thus fail) or create a new header that's only used for the dictionaries
and which contains nothing but "namespace std{} using namespace std;"
You pass it to cint together with the actual headers when creating the
dictionary. That way your program will not use it, only the CINT
dictionaries will #include it.

Cheers, Axel.

Joost Kraaijeveld wrote on 10/29/2009 09:36 AM:

> Hi,
>
> I am experimenting with embedding Cint in my own application and giving
> it access to my own libraries. The interface of my libraries uses
> "std::string", e.g:
>
> class Library
> {
> public:
> void fn(const std::string& a);
> }
>
>
>
> If I compile this library with makecint I get the following error:
>
> G__cpp_myLibrary.cxx: In function ‘int G__myLibrary_46_0_2(G__value*,
> const char*, G__param*, int)’:
> G__cpp_myLibrary.cxx:83: error: ‘string’ was not declared in this scope
> G__cpp_myLibrary.cxx:83: error: expected primary-expression before ‘)’
> token
> G__cpp_myLibrary.cxx: In function ‘void G__setup_memfuncLibrary()’:
> G__cpp_myLibrary.cxx:243: error: expected primary-expression before
> ‘void’
> G__cpp_myLibrary.cxx:243: error: expected `)' before ‘void’
> make: *** [G__cpp_myLibrary.o] Error 1
>
>
> If I change the above to
>
> using std::string;
> class Library
> {
> public:
> void fn(const string& a);
> }
>
> it works OK.
>
> Is this how it is supposed to be? Or can I somehow use std::string in my
> interface headers, without resorting to dummy wrapper headers or magic
> macros?
>
> TIA
>

Can I use std::string in an external library interface?

2009-10-29T01:36:15Z

Hi,

I am experimenting with embedding Cint in my own application and giving
it access to my own libraries. The interface of my libraries uses
"std::string", e.g:

class Library
{
public:
void fn(const std::string& a);
}

If I compile this library with makecint I get the following error:

G__cpp_myLibrary.cxx: In function ‘int G__myLibrary_46_0_2(G__value*,
const char*, G__param*, int)’:
G__cpp_myLibrary.cxx:83: error: ‘string’ was not declared in this scope
G__cpp_myLibrary.cxx:83: error: expected primary-expression before ‘)’
token
G__cpp_myLibrary.cxx: In function ‘void G__setup_memfuncLibrary()’:
G__cpp_myLibrary.cxx:243: error: expected primary-expression before
‘void’
G__cpp_myLibrary.cxx:243: error: expected `)' before ‘void’
make: *** [G__cpp_myLibrary.o] Error 1

If I change the above to

using std::string;
class Library
{
public:
void fn(const string& a);
}

it works OK.

Is this how it is supposed to be? Or can I somehow use std::string in my
interface headers, without resorting to dummy wrapper headers or magic
macros?

TIA

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

Re: Debian package standalone cint remarks

2009-09-26T09:43:15Z

If you try

http://pivot.cs.unb.ca/debian/pool/main/c/cint/cint_7.3.00~r30149-2.dsc

it should at least build cint7 and libReflex.so now. You do have to use
dpkg-buildpackage (or debuild) to build it.

d

Re: Thread safeness of cint?

2009-09-26T06:39:22Z

Hi Joost,

cint5 is easier to use in a MT environment than cint7, because it uses
less STL containers. I believe Cint5 was once sort of thread safe. Sort
of, because the trick was to load a separate libCint.so per thread,
having separate functions and most importantly separate globals. You can
see in the lower half of G__ci.h how this is done. We have not tried
this for a long time, so it will most probably not work. But it should
be the easiest way for you, if the "one libCint.so per thread" approach
is acceptable for you.

Cheers, Axel.

On 2009-09-25 10:23, Joost Kraaijeveld wrote:
> Hi,
>
> The source of cint explicitly say that cint is not MT safe. Is that
> still the case for cint7? If so, does anyone know if it is a lot of work
> (going to impossible without a complete rewrite)?
>
> TIA
>

Re: Debian package standalone cint remarks

2009-09-25T05:23:43Z

Joost Kraaijeveld wrote:

>Thanks for the link. Your package has the same problems as the one I
>build. Did you ever try to compile the socket library with the installed
>version and did it work for you?

No, I didn't try, sorry. Did you build my packages via "make deb" or
using the usual Debian way (e.g. using debuild?). "make deb" will
overwrite the debian packaging I provide.

I know something about debian packaging, but less about cint. These
are my rough draft packages for upload to Debian, so if something is
really not right, it would nice to fix it before uploading.

It would be great if you could follow up with whatever you figure out
to

mailto:533858@...

or

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=533858

Re: Debian package standalone cint remarks

2009-09-25T05:08:27Z

Hi David,

Thanks for the link. Your package has the same problems as the one I
build. Did you ever try to compile the socket library with the installed
version and did it work for you?

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

Re: Debian package standalone cint remarks

2009-09-25T04:19:41Z

Joost Kraaijeveld wrote:

>Hi,

>In an attempt to create a Debian .deb package I noticed several
>problems. Before I try to "improve" anything, I hope that this post will
>get me some advise on how to actually improve it, as I have no
>experience in creating Debian packages at all.

You might find it helpful to look at the debian package I am using at

http://pivot.cs.unb.ca/debian/pool/main/c/cint/cint_7.3.00~r30149-1.dsc

Thread safeness of cint?

2009-09-25T01:23:50Z

Hi,

The source of cint explicitly say that cint is not MT safe. Is that
still the case for cint7? If so, does anyone know if it is a lot of work
(going to impossible without a complete rewrite)?

TIA

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

Debian package standalone cint remarks

2009-09-25T01:21:42Z

Hi,

In an attempt to create a Debian .deb package I noticed several
problems. Before I try to "improve" anything, I hope that this post will
get me some advise on how to actually improve it, as I have no
experience in creating Debian packages at all.

1. "./configure --coreversion=new ; make deb" creates a cint_7.3.00-1
package that actually contains cint 5.17.00 ( I used a checkout of svn
trunk) and it does not contain the reflex library

2. the included cint-config contains (striped some things):

--incdir) $pathconvert "/usr/include" ;;
--cflags) $pathconvert "-I/usr/include -I/home/jkr/src/cint/reflex/inc" ;;
--libs) $pathconvert "-L/usr/lib -lCint -lReflex" ;;

ad --incdir: the examples that depend on finding the cint includes there
will not work as the actual directory in /usr/include/cint, which seems
to be reasonable to me
ad -cflags: reflex is not included in the deb-package and the path for
the reflex include is just available on my local machine, which renders
the deb unusable for distribution
ad --libs: reflex is not included in the deb-package
ad --version: is correct for the actual inclused version but is not

3. it seems to be impossible to compile any of the libraries
in /usr/share/cin/cint/lib. All compilation attempts lead tot the
following error (commandline of cint included):
/usr/bin/cint -K -w2 -zcintsock -nG__c_cintsock.c -D__MAKECINT__
-DG__MAKECINT -DG__SHAREDLIB -DG__OSFDLL -DG__ANSI -DG__ERRORCALLBACK
-DG__SIGNEDCHAR -DG__NEWSTDHEADER -DG__CINT_VER6 -DG__NATIVELONGLONG
-DG__P2FCAST -DG__STD_EXCEPTION -DG__HAVE_CONFIG -DG__NOMAKEINFO -c-2
-I/usr/include -DG__SHAREDLIB -DG__OSFDLL -DG__ANSI -DG__ERRORCALLBACK
-DG__SIGNEDCHAR -DG__NEWSTDHEADER -DG__CINT_VER6 -DG__NATIVELONGLONG
-DG__P2FCAST -DG__STD_EXCEPTION -DG__HAVE_CONFIG -DG__NOMAKEINFO -Z0
cintsock.h
#error "You need a ISO C conforming compiler to use the glibc headers"

Can anyone give me some advise/directions of how to get the package
properly compiled and installed for Debian (both cint and cint7),
additional libraries included? (All OOTB solutions are also very
welcome ;-))

TIA

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

Re: Problem running demo/tcpip / compiling dlls

2009-09-25T00:28:10Z

Hi Philippe,

[Added a CC to the mailing list which got lost with my previous ]

On Thu, 2009-09-24 at 10:18 -0500, Philippe Canal wrote:

> Hi Joost,
>
> Do you see the same problem with the 'old'/'current core.
>
> Philippe.
>
> Joost Kraaijeveld wrote:
> > Hi Phillipe,
> >
> > Thanks for your response.
> >
> > I succeeded in compiling the socket dll. What I did was:
> >
> > svn co http://root.cern.ch/svn/root/trunk/cint cint
> > cd cint
> > ./configure --coreversion=new
> > make all
> > export PATH=$PWD/bin:$PATH
> > export LD_LIBRARY_PATH=$PWD/lib:$LD_LIBRARY_PATH
> > export CINTSYSDIR=$PWD
> > cd cint7/lib/socket/
> > sh ./setup
> >
> > copy the dll to $CINTSYSDIR/cint7/include
> >
> > cd ../../../demo/tcpip/
> > cint server.cxx
> >

I did not need to copy the dll manually (because of your path to the
setup script I believe). otherwise the procedure was the same and leaded
to a working standalone cint and a compilation of the socket dll.

I also investigated the Debian package problems a bit further and will
start a new Debian thread for that.

thanks,

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

Re: bugs in parsing ++/-- operators?

2009-09-22T07:51:40Z

Hi David,

David Bremner wrote on 09/22/2009 04:39 PM:

>> we are aware of these parsing limitations. We don't plan to fix them -
>> it would require a substantial rewrite of the operator parser. I hope
>> that the workaround that you already found yourself is sufficient for
>> you... If it's not then your patches are always welcome!
>
> It is a bit unfortunate for teaching purposes. I have been using CINT
> to quickly run snippets of C code during lectures, and although these
> expressions are ugly, they are exactly the kind of corner cases one
> wants to test. I understand however that exact C standard compliance
> is not really a design goal for CINT.

It wasn't, and now it's too late for some of the corners... Sorry for
the bad news :-(

Cheers, Axel.

Re: bugs in parsing ++/-- operators?

2009-09-22T07:39:46Z

Err, sorry about the bad threading. I think anything I do now will make things worse...

At Tue, 22 Sep 2009 15:43:31 +0200,
Axel Naumann wrote:

> we are aware of these parsing limitations. We don't plan to fix them -
> it would require a substantial rewrite of the operator parser. I hope
> that the workaround that you already found yourself is sufficient for
> you... If it's not then your patches are always welcome!

It is a bit unfortunate for teaching purposes. I have been using CINT
to quickly run snippets of C code during lectures, and although these
expressions are ugly, they are exactly the kind of corner cases one
wants to test. I understand however that exact C standard compliance
is not really a design goal for CINT.

d

Re: bugs in parsing ++/-- operators?

2009-09-22T06:43:31Z

Hi David,

we are aware of these parsing limitations. We don't plan to fix them -
it would require a substantial rewrite of the operator parser. I hope
that the workaround that you already found yourself is sufficient for
you... If it's not then your patches are always welcome!

Cheers, Axel.

David Bremner wrote on 09/22/2009 03:24 PM:

> It seems like the cint parser has trouble with
>
> - --j
>
> and
>
> + ++j
>
> here are some test cases
>
> ,----
> | cint> {int i=1; int j=1; printf("%d", --i - --j); }
> | -1(const int)2
> | cint>
> | cint> {int i=1; int j=1; printf("%d\n", --i - --j); }
> | -1
> | (const int)3
> | cint> {int i=1; int j=1; printf("%d\n", ++i - --j); }
> | -1
> | (const int)3
> | cint> {int i=1; int j=1; printf("%d\n", ++i - ++j); }
> | 0
> | (const int)2
> | cint> {int i=1; int j=1; printf("%d\n", ++i + ++j); }
> | 1
> | (const int)2
> | cint>
> `----
>
> Adding enough parentheses fixes the failing ones.
>
> This is svn revision 30149, banner says
> revision : 5.17.00, Dec 21, 2008 by M.Goto
>
>
>

bugs in parsing ++/-- operators?

2009-09-22T06:24:03Z

It seems like the cint parser has trouble with

- --j

and

+ ++j

here are some test cases

,----
| cint> {int i=1; int j=1; printf("%d", --i - --j); }
| -1(const int)2
| cint>
| cint> {int i=1; int j=1; printf("%d\n", --i - --j); }
| -1
| (const int)3
| cint> {int i=1; int j=1; printf("%d\n", ++i - --j); }
| -1
| (const int)3
| cint> {int i=1; int j=1; printf("%d\n", ++i - ++j); }
| 0
| (const int)2
| cint> {int i=1; int j=1; printf("%d\n", ++i + ++j); }
| 1
| (const int)2
| cint>
`----

Adding enough parentheses fixes the failing ones.

This is svn revision 30149, banner says
revision : 5.17.00, Dec 21, 2008 by M.Goto

Re: Problem running demo/tcpip / compiling dlls

2009-09-18T04:52:58Z

Hi Joost,

I just updated (revision 30247) the setup script (it nows works for me :)).
Can you give it a try?

Thanks,
Philippe.

Joost Kraaijeveld wrote:

> Hi,
>
> I am trying to get the tcp/ip demo running on my Debian (testing)
> machine. I am using a freshly checked out version of cint 5.17.00 (svn
> co http://root.cern.ch/svn/root/trunk/cint)
>
> After compiling and installing the Debian package using "make deb" I can
> run several demo's e.g. the stl demo.
>
> Running the tcp/ip demo however results in:
>
> jkr@panoramix:~/src/cint-5.17.00/demo/tcpip$ cint server.cxx
> #error cintsock.dll is not ready. Run setup script in
> $CINTSYSDIR/lib/socket directory
>
>
> Doing so results in the message below, which basically says that I
> should have a decent compiler (which I obviously do or I could not have
> compiled cint in the first place). I have tried "make dlls" but that
> does not build it either.
>
> The dlls that get build (and also installed) are:
> ./debian/cint/usr/share/cint/cint/include/sys/ipc.dll
> ./debian/cint/usr/share/cint/cint/include/long.dll
> ./debian/cint/usr/share/cint/cint/include/stdfunc.dll
> ./debian/cint/usr/share/cint/cint/include/posix.dll
> ./debian/cint/usr/share/cint/cint/include/stdcxxfunc.dll
> ./debian/cint/usr/share/cint/cint/stl/vector.dll
> ./debian/cint/usr/share/cint/cint/stl/list.dll
> ./debian/cint/usr/share/cint/cint/stl/deque.dll
> ./debian/cint/usr/share/cint/cint/stl/map.dll
> ./debian/cint/usr/share/cint/cint/stl/map2.dll
> ./debian/cint/usr/share/cint/cint/stl/set.dll
> ./debian/cint/usr/share/cint/cint/stl/multimap.dll
> ./debian/cint/usr/share/cint/cint/stl/multimap2.dll
> ./debian/cint/usr/share/cint/cint/stl/multiset.dll
> ./debian/cint/usr/share/cint/cint/stl/stack.dll
> ./debian/cint/usr/share/cint/cint/stl/queue.dll
> ./debian/cint/usr/share/cint/cint/stl/valarray.dll
> ./debian/cint/usr/share/cint/cint/stl/exception.dll
> ./debian/cint/usr/share/cint/cint/stl/stdexcept.dll
> ./debian/cint/usr/share/cint/cint/stl/climits.dll
> ./debian/cint/usr/share/cint/cint/stl/complex.dll
> ./debian/cint/usr/share/cint/cint/stl/string.dll
> ./debian/cint/usr/share/cint/cint/stl/vectorbool.dll
>
>
> Is there a solution for this problem? Is there anything I could try or
> look at?
>
>
> panoramix:/usr/share/cint/cint/lib/socket# sh ./setup
> cp: cannot stat `/MAKEINFO': No such file or directory
> cc -o mksockh mksockh.c
> rm: cannot remove `mksockh.exe': No such file or directory
> ##########################################################################
> # makecint : interpreter-compiler for cint (UNIX version)
> # Copyright(c) 1995~2007 Masaharu Goto. Mailing list:
> cint@...
> ##########################################################################
> Run 'make -f Makefile' to compile the object
> gcc -DG__SHAREDLIB -DG__OSFDLL -DG__ANSI -DG__ERRORCALLBACK
> -DG__SIGNEDCHAR -DG__NEWSTDHEADER -DG__CINT_VER6 -DG__NATIVELONGLONG
> -DG__P2FCAST -DG__STD_EXCEPTION -DG__HAVE_CONFIG -DG__NOMAKEINFO -O2
> -m64 -fPIC -o cintsock.o -c cintsock.c
> /usr/bin/cint -K -w2 -zcintsock -nG__c_cintsock.c -D__MAKECINT__
> -DG__MAKECINT -DG__SHAREDLIB -DG__OSFDLL -DG__ANSI -DG__ERRORCALLBACK
> -DG__SIGNEDCHAR -DG__NEWSTDHEADER -DG__CINT_VER6 -DG__NATIVELONGLONG
> -DG__P2FCAST -DG__STD_EXCEPTION -DG__HAVE_CONFIG -DG__NOMAKEINFO -c-2
> -I/usr/include -DG__SHAREDLIB -DG__OSFDLL -DG__ANSI -DG__ERRORCALLBACK
> -DG__SIGNEDCHAR -DG__NEWSTDHEADER -DG__CINT_VER6 -DG__NATIVELONGLONG
> -DG__P2FCAST -DG__STD_EXCEPTION -DG__HAVE_CONFIG -DG__NOMAKEINFO -Z0
> cintsock.h
> #error "You need a ISO C conforming compiler to use the glibc headers"
> !!!Removing G__c_cintsock.c G__c_cintsock.h !!!
> make: *** [G__c_cintsock.c] Error 1
> rm: cannot remove `G__*': No such file or directory
>
>
> TIA
>
>

Re: AW: Problem running demo/tcpip / compiling dlls

2009-09-15T13:40:35Z

Hi Louis,

Thanks for answering.

On Tue, 2009-09-15 at 20:46 +0200, Louis wrote:
> Hi J.
> You problem is exactly at this line:
> cc -o mksockh mksockh.c
> You compiled cint with gcc, but you try to compile the lib with cc.
> Change the compiler executable within the makefile from cc to gcc and it
> should compile.

The makefile is generated by the setup script (which also automatically
remove both the makefile and o file after compilation).

Fiddling with some "export CC=" statements changed the compiler but
there definitely seems to be something wrong:

panoramix:/usr/share/cint/cint/lib/socket# sh ./setup
cp: cannot stat `/MAKEINFO': No such file or directory
gcc -o mksockh mksockh.c
rm: cannot remove `mksockh.exe': No such file or directory
##########################################################################
# makecint : interpreter-compiler for cint (UNIX version)
# Copyright(c) 1995~2007 Masaharu Goto. Mailing list: cint@...
##########################################################################
Run 'make -f Makefile' to compile the object
gcc -DG__SHAREDLIB -DG__OSFDLL -DG__ANSI -DG__ERRORCALLBACK -DG__SIGNEDCHAR -DG__NEWSTDHEADER -DG__CINT_VER6 -DG__NATIVELONGLONG -DG__P2FCAST -DG__STD_EXCEPTION -DG__HAVE_CONFIG -DG__NOMAKEINFO -O2 -m64 -fPIC -o cintsock.o -c cintsock.c
/usr/bin/cint -K -w2 -zcintsock -nG__c_cintsock.c -D__MAKECINT__ -DG__MAKECINT -DG__SHAREDLIB -DG__OSFDLL -DG__ANSI -DG__ERRORCALLBACK -DG__SIGNEDCHAR -DG__NEWSTDHEADER -DG__CINT_VER6 -DG__NATIVELONGLONG -DG__P2FCAST -DG__STD_EXCEPTION -DG__HAVE_CONFIG -DG__NOMAKEINFO -c-2 -I/usr/include -DG__SHAREDLIB -DG__OSFDLL -DG__ANSI -DG__ERRORCALLBACK -DG__SIGNEDCHAR -DG__NEWSTDHEADER -DG__CINT_VER6 -DG__NATIVELONGLONG -DG__P2FCAST -DG__STD_EXCEPTION -DG__HAVE_CONFIG -DG__NOMAKEINFO -Z0 cintsock.h
#error "You need a ISO C conforming compiler to use the glibc headers"
!!!Removing G__c_cintsock.c G__c_cintsock.h !!!
make: *** [G__c_cintsock.c] Error 1
rm: cannot remove `G__*': No such file or directory
panoramix:/usr/share/cint/cint/lib/socket#

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl

AW: Problem running demo/tcpip / compiling dlls

2009-09-15T11:46:01Z

Hi J.
You problem is exactly at this line:
cc -o mksockh mksockh.c
You compiled cint with gcc, but you try to compile the lib with cc.
Change the compiler executable within the makefile from cc to gcc and it
should compile.

Greets Louis

-----Ursprüngliche Nachricht-----
Von: owner-cint@... [mailto:owner-cint@...] Im Auftrag von
Joost Kraaijeveld
Gesendet: Dienstag, 15. September 2009 10:50
An: cint@...
Betreff: [CINT] Problem running demo/tcpip / compiling dlls

Hi,

I am trying to get the tcp/ip demo running on my Debian (testing)
machine. I am using a freshly checked out version of cint 5.17.00 (svn
co http://root.cern.ch/svn/root/trunk/cint)

After compiling and installing the Debian package using "make deb" I can
run several demo's e.g. the stl demo.

Running the tcp/ip demo however results in:

jkr@panoramix:~/src/cint-5.17.00/demo/tcpip$ cint server.cxx
#error cintsock.dll is not ready. Run setup script in
$CINTSYSDIR/lib/socket directory

Doing so results in the message below, which basically says that I
should have a decent compiler (which I obviously do or I could not have
compiled cint in the first place). I have tried "make dlls" but that
does not build it either.

The dlls that get build (and also installed) are:
./debian/cint/usr/share/cint/cint/include/sys/ipc.dll
./debian/cint/usr/share/cint/cint/include/long.dll
./debian/cint/usr/share/cint/cint/include/stdfunc.dll
./debian/cint/usr/share/cint/cint/include/posix.dll
./debian/cint/usr/share/cint/cint/include/stdcxxfunc.dll
./debian/cint/usr/share/cint/cint/stl/vector.dll
./debian/cint/usr/share/cint/cint/stl/list.dll
./debian/cint/usr/share/cint/cint/stl/deque.dll
./debian/cint/usr/share/cint/cint/stl/map.dll
./debian/cint/usr/share/cint/cint/stl/map2.dll
./debian/cint/usr/share/cint/cint/stl/set.dll
./debian/cint/usr/share/cint/cint/stl/multimap.dll
./debian/cint/usr/share/cint/cint/stl/multimap2.dll
./debian/cint/usr/share/cint/cint/stl/multiset.dll
./debian/cint/usr/share/cint/cint/stl/stack.dll
./debian/cint/usr/share/cint/cint/stl/queue.dll
./debian/cint/usr/share/cint/cint/stl/valarray.dll
./debian/cint/usr/share/cint/cint/stl/exception.dll
./debian/cint/usr/share/cint/cint/stl/stdexcept.dll
./debian/cint/usr/share/cint/cint/stl/climits.dll
./debian/cint/usr/share/cint/cint/stl/complex.dll
./debian/cint/usr/share/cint/cint/stl/string.dll
./debian/cint/usr/share/cint/cint/stl/vectorbool.dll

Is there a solution for this problem? Is there anything I could try or
look at?

panoramix:/usr/share/cint/cint/lib/socket# sh ./setup
cp: cannot stat `/MAKEINFO': No such file or directory
cc -o mksockh mksockh.c
rm: cannot remove `mksockh.exe': No such file or directory
##########################################################################
# makecint : interpreter-compiler for cint (UNIX version)
# Copyright(c) 1995~2007 Masaharu Goto. Mailing list:
cint@...
##########################################################################
Run 'make -f Makefile' to compile the object
gcc -DG__SHAREDLIB -DG__OSFDLL -DG__ANSI -DG__ERRORCALLBACK
-DG__SIGNEDCHAR -DG__NEWSTDHEADER -DG__CINT_VER6 -DG__NATIVELONGLONG
-DG__P2FCAST -DG__STD_EXCEPTION -DG__HAVE_CONFIG -DG__NOMAKEINFO -O2
-m64 -fPIC -o cintsock.o -c cintsock.c
/usr/bin/cint -K -w2 -zcintsock -nG__c_cintsock.c -D__MAKECINT__
-DG__MAKECINT -DG__SHAREDLIB -DG__OSFDLL -DG__ANSI -DG__ERRORCALLBACK
-DG__SIGNEDCHAR -DG__NEWSTDHEADER -DG__CINT_VER6 -DG__NATIVELONGLONG
-DG__P2FCAST -DG__STD_EXCEPTION -DG__HAVE_CONFIG -DG__NOMAKEINFO -c-2
-I/usr/include -DG__SHAREDLIB -DG__OSFDLL -DG__ANSI -DG__ERRORCALLBACK
-DG__SIGNEDCHAR -DG__NEWSTDHEADER -DG__CINT_VER6 -DG__NATIVELONGLONG
-DG__P2FCAST -DG__STD_EXCEPTION -DG__HAVE_CONFIG -DG__NOMAKEINFO -Z0
cintsock.h
#error "You need a ISO C conforming compiler to use the glibc headers"
!!!Removing G__c_cintsock.c G__c_cintsock.h !!!
make: *** [G__c_cintsock.c] Error 1
rm: cannot remove `G__*': No such file or directory

TIA

--
Groeten,

Joost Kraaijeveld
Askesis B.V.
Molukkenstraat 14
6524NB Nijmegen
tel: 024-3888063 / 06-51855277
fax: 024-3608416
web: www.askesis.nl