Fink

Daniel Johnson wrote:
>[]
> We still need to decide what to do about passwd, though.

The passwd package does its work in the PostInstScript. There is no 
problem with --build-as-nobody, AFAICT.

-- 
Martin

Martin Costabel wrote:
> Daniel Johnson wrote:
>> []
>> We still need to decide what to do about passwd, though.
> 
> The passwd package does its work in the PostInstScript. There is no 
> problem with --build-as-nobody, AFAICT.

Sorry, I misunderstood. And I had already forgotten the discussion we 
had about this some 5 months ago. I repeat some of it:

Your question was whether passwd should be made essential, which it is 
de facto already for those who use --build-as-nobody. Making it 
essential wouldn't even be enough, because right now you can install the 
passwd package and say "no" when it asks if it should be allowed to add 
users and groups (In the text where it asks this question, it quaintly 
talks about the NetInfo database which went out of fashion a couple of 
years ago when Leopard arrived.)

This is a serious problem that needs to be addressed. I am not in favor 
of adding unneeded junk to the essential packages; there is already far 
too much of this (all the *zip* packages), and it is difficult to get 
rid of. The passwd package would, however, be much easier to digest if 
Hanspeter's split version were adopted.

Then fink could depend on passwd-fink-bld (or the latter plus 
passwd-core be made essential). Or just passwd-core could be made 
essential, which is inoffensive, because it does nothing, and then 
fink.info could simply execute something like

   %p/etc/update-passwd dontask fink-bld

which adds the fink-bld user without asking. This is for macosx >= 10.5, 
of course.

-- 
Martin

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 8/10/10 4:52 PM, Abe, Sumiko wrote:
> Hi, I am running fink configure with downloading unstable packages options.
> And run fink selfupdate. The following errors happened:
> cp ./match.c _match.S
> gcc -std=gnu99 -E -I/sw/include  _match.S > _match.i
> sed '/^ * < _match.i > match_.s
> /bin/sh: -c: line 0: unexpected EOF while looking for matching `''
> /bin/sh: -c: line 1: syntax error: unexpected end of file
> make[3]: *** [match.o] Error 2
> make[2]: *** [all-recursive] Error 1
> make[1]: *** [all] Error 2
> make: *** [all-recursive] Error 1
> ### execution of make failed, exit code 2
> Removing runtime build-lock...
> Removing build-lock package...
> /sw/bin/dpkg-lockwait -r fink-buildlock-gzip-1.4-2
> (Reading database ... 6073 files and directories currently installed.)
> Removing fink-buildlock-gzip-1.4-2 ...
> Failed: phase compiling: gzip-1.4-2 failed
> 
> Before reporting any errors, please run "fink selfupdate" and
> try again.  If you continue to have issues, please check to see if the
> FAQ on fink's website solves the problem.  If not, ask on the fink-users
> or fink-beginners mailing lists, with a carbon copy to the maintainer:
> 
>         Fink Core Group <fink-core@...>
> 
> Note that this is preferable to emailing the maintainer directly, since
> most fink package maintainers do not have access to all possible
> hardware and software configurations.
> 
> 
> Any advice?
> My Mac is running on 10.6.3.
> Thank you,
> 
> Sumiko
> 

What do you get from "which sed"?
And what does "grep Root /sw/etc/fink.conf" show?

- -- 
Alexander Hansen
Fink User Liaison
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkxh4t0ACgkQB8UpO3rKjQ+0AQCgldXCJrQ9EvM4dXQ9KmROFEvc
LtEAoJzb/A1qGI+g7IHgPW6t7ujpkPQQ
=PQou
-----END PGP SIGNATURE-----

Thanks,


> What do you get from "which sed"?
> And what does "grep Root /sw/etc/fink.conf" show?

The result is here:

sh-3.2# which sed
/usr/bin/sed
sh-3.2# grep Root /sw/etc/fink.conf
RootMethod: su
sh-3.2#

Thank you,

Sumiko



-----Original Message-----
From: Alexander Hansen [mailto:alexanderk.hansen@...]
Sent: Tuesday, August 10, 2010 4:38 PM
To: Abe, Sumiko
Cc: Core Group
Subject: Re: DCMTK and xmedcon unstable package install...

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 8/10/10 4:52 PM, Abe, Sumiko wrote:
> Hi, I am running fink configure with downloading unstable packages options.
> And run fink selfupdate. The following errors happened:
> cp ./match.c _match.S
> gcc -std=gnu99 -E -I/sw/include  _match.S > _match.i
> sed '/^ * < _match.i > match_.s
> /bin/sh: -c: line 0: unexpected EOF while looking for matching `''
> /bin/sh: -c: line 1: syntax error: unexpected end of file
> make[3]: *** [match.o] Error 2
> make[2]: *** [all-recursive] Error 1
> make[1]: *** [all] Error 2
> make: *** [all-recursive] Error 1
> ### execution of make failed, exit code 2
> Removing runtime build-lock...
> Removing build-lock package...
> /sw/bin/dpkg-lockwait -r fink-buildlock-gzip-1.4-2
> (Reading database ... 6073 files and directories currently installed.)
> Removing fink-buildlock-gzip-1.4-2 ...
> Failed: phase compiling: gzip-1.4-2 failed
>
> Before reporting any errors, please run "fink selfupdate" and
> try again.  If you continue to have issues, please check to see if the
> FAQ on fink's website solves the problem.  If not, ask on the fink-users
> or fink-beginners mailing lists, with a carbon copy to the maintainer:
>
>         Fink Core Group <fink-core@...>
>
> Note that this is preferable to emailing the maintainer directly, since
> most fink package maintainers do not have access to all possible
> hardware and software configurations.
>
>
> Any advice?
> My Mac is running on 10.6.3.
> Thank you,
>
> Sumiko
>

What do you get from "which sed"?
And what does "grep Root /sw/etc/fink.conf" show?

- --
Alexander Hansen
Fink User Liaison
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkxh4t0ACgkQB8UpO3rKjQ+0AQCgldXCJrQ9EvM4dXQ9KmROFEvc
LtEAoJzb/A1qGI+g7IHgPW6t7ujpkPQQ
=PQou
-----END PGP SIGNATURE-----

This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately by e-mail if you have received this e-mail by mistake and delete this e-mail from your system. E-mail transmission cannot be guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. The sender therefore does not accept liability for any errors or omissions in the contents of this message, which arise as a result of e-mail transmission.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Please try the following:

1)  Edit /sw/fink/dists/unstable/main/finkinfo/base/gzip.info.
2)  Change the CompileScript block to read
CompileScript: <<
#!/bin/bash -ev
 ./configure %c
 make all
<<
3)  Save the file.
4)  Try "fink update gzip" and let us know if it works for you.

On 8/10/10 8:06 PM, Abe, Sumiko wrote:
> Thanks,
> 
> 
>> What do you get from "which sed"?
>> And what does "grep Root /sw/etc/fink.conf" show?
> 
> The result is here:
> 
> sh-3.2# which sed
> /usr/bin/sed
> sh-3.2# grep Root /sw/etc/fink.conf
> RootMethod: su
> sh-3.2#
> 
> Thank you,
> 
> Sumiko
> 
> 
> 
> -----Original Message-----
> From: Alexander Hansen [mailto:alexanderk.hansen@...]
> Sent: Tuesday, August 10, 2010 4:38 PM
> To: Abe, Sumiko
> Cc: Core Group
> Subject: Re: DCMTK and xmedcon unstable package install...
> 
> On 8/10/10 4:52 PM, Abe, Sumiko wrote:
>> Hi, I am running fink configure with downloading unstable packages options.
>> And run fink selfupdate. The following errors happened:
>> cp ./match.c _match.S
>> gcc -std=gnu99 -E -I/sw/include  _match.S > _match.i
>> sed '/^ * < _match.i > match_.s
>> /bin/sh: -c: line 0: unexpected EOF while looking for matching `''
>> /bin/sh: -c: line 1: syntax error: unexpected end of file
>> make[3]: *** [match.o] Error 2
>> make[2]: *** [all-recursive] Error 1
>> make[1]: *** [all] Error 2
>> make: *** [all-recursive] Error 1
>> ### execution of make failed, exit code 2
>> Removing runtime build-lock...
>> Removing build-lock package...
>> /sw/bin/dpkg-lockwait -r fink-buildlock-gzip-1.4-2
>> (Reading database ... 6073 files and directories currently installed.)
>> Removing fink-buildlock-gzip-1.4-2 ...
>> Failed: phase compiling: gzip-1.4-2 failed
> 
>> Before reporting any errors, please run "fink selfupdate" and
>> try again.  If you continue to have issues, please check to see if the
>> FAQ on fink's website solves the problem.  If not, ask on the fink-users
>> or fink-beginners mailing lists, with a carbon copy to the maintainer:
> 
>>         Fink Core Group <fink-core@...>
> 
>> Note that this is preferable to emailing the maintainer directly, since
>> most fink package maintainers do not have access to all possible
>> hardware and software configurations.
> 
> 
>> Any advice?
>> My Mac is running on 10.6.3.
>> Thank you,
> 
>> Sumiko
> 
> 
> What do you get from "which sed"?
> And what does "grep Root /sw/etc/fink.conf" show?
> 



- -- 
Alexander Hansen
Fink User Liaison
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkxh7YAACgkQB8UpO3rKjQ8uwgCeOrqBs+MND7mAkwofDnErCwho
WR4An1dz98VBk9YTkrK15roIR6ukB0E6
=v+hj
-----END PGP SIGNATURE-----

Hi, Thank you so much for your great help.

It doesn't work with the edition of gzip.info file.

The attached file is the revised gzip.info file and error is following and same as before:

.....
  CC       xmalloc.o
  CC       yesno.o
cp ./match.c _match.S
gcc -std=gnu99 -E -I/sw/include  _match.S > _match.i
sed '/^ * < _match.i > match_.s
/bin/sh: -c: line 0: unexpected EOF while looking for matching `''
/bin/sh: -c: line 1: syntax error: unexpected end of file
make[3]: *** [match.o] Error 2
make[2]: *** [all-recursive] Error 1
make[1]: *** [all] Error 2
make: *** [all-recursive] Error 1
### execution of /var/tmp/tmp.1.lgwctH failed, exit code 2
Removing runtime build-lock...
Removing build-lock package...
/sw/bin/dpkg-lockwait -r fink-buildlock-gzip-1.4-2
(Reading database ... 6073 files and directories currently installed.)
Removing fink-buildlock-gzip-1.4-2 ...
Failed: phase compiling: gzip-1.4-2 failed

Before reporting any errors, please run "fink selfupdate" and
try again.  If you continue to have issues, please check to see if the
FAQ on fink's website solves the problem.  If not, ask on the fink-users
or fink-beginners mailing lists, with a carbon copy to the maintainer:

        Fink Core Group <fink-core@...>

Note that this is preferable to emailing the maintainer directly, since
most fink package maintainers do not have access to all possible
hardware and software configurations.

Thank you again,

Sumiko


-----Original Message-----
From: Alexander Hansen [mailto:alexanderk.hansen@...]
Sent: Tuesday, August 10, 2010 5:23 PM
To: Abe, Sumiko
Cc: Core Group
Subject: Re: DCMTK and xmedcon unstable package install...

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Please try the following:

1)  Edit /sw/fink/dists/unstable/main/finkinfo/base/gzip.info.
2)  Change the CompileScript block to read
CompileScript: <<
#!/bin/bash -ev
 ./configure %c
 make all
<<
3)  Save the file.
4)  Try "fink update gzip" and let us know if it works for you.

On 8/10/10 8:06 PM, Abe, Sumiko wrote:
> Thanks,
>
>
>> What do you get from "which sed"?
>> And what does "grep Root /sw/etc/fink.conf" show?
>
> The result is here:
>
> sh-3.2# which sed
> /usr/bin/sed
> sh-3.2# grep Root /sw/etc/fink.conf
> RootMethod: su
> sh-3.2#
>
> Thank you,
>
> Sumiko
>
>
>
> -----Original Message-----
> From: Alexander Hansen [mailto:alexanderk.hansen@...]
> Sent: Tuesday, August 10, 2010 4:38 PM
> To: Abe, Sumiko
> Cc: Core Group
> Subject: Re: DCMTK and xmedcon unstable package install...
>
> On 8/10/10 4:52 PM, Abe, Sumiko wrote:
>> Hi, I am running fink configure with downloading unstable packages options.
>> And run fink selfupdate. The following errors happened:
>> cp ./match.c _match.S
>> gcc -std=gnu99 -E -I/sw/include  _match.S > _match.i
>> sed '/^ * < _match.i > match_.s
>> /bin/sh: -c: line 0: unexpected EOF while looking for matching `''
>> /bin/sh: -c: line 1: syntax error: unexpected end of file
>> make[3]: *** [match.o] Error 2
>> make[2]: *** [all-recursive] Error 1
>> make[1]: *** [all] Error 2
>> make: *** [all-recursive] Error 1
>> ### execution of make failed, exit code 2
>> Removing runtime build-lock...
>> Removing build-lock package...
>> /sw/bin/dpkg-lockwait -r fink-buildlock-gzip-1.4-2
>> (Reading database ... 6073 files and directories currently installed.)
>> Removing fink-buildlock-gzip-1.4-2 ...
>> Failed: phase compiling: gzip-1.4-2 failed
>
>> Before reporting any errors, please run "fink selfupdate" and
>> try again.  If you continue to have issues, please check to see if the
>> FAQ on fink's website solves the problem.  If not, ask on the fink-users
>> or fink-beginners mailing lists, with a carbon copy to the maintainer:
>
>>         Fink Core Group <fink-core@...>
>
>> Note that this is preferable to emailing the maintainer directly, since
>> most fink package maintainers do not have access to all possible
>> hardware and software configurations.
>
>
>> Any advice?
>> My Mac is running on 10.6.3.
>> Thank you,
>
>> Sumiko
>
>
> What do you get from "which sed"?
> And what does "grep Root /sw/etc/fink.conf" show?
>



- --
Alexander Hansen
Fink User Liaison
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkxh7YAACgkQB8UpO3rKjQ8uwgCeOrqBs+MND7mAkwofDnErCwho
WR4An1dz98VBk9YTkrK15roIR6ukB0E6
=v+hj
-----END PGP SIGNATURE-----

This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately by e-mail if you have received this e-mail by mistake and delete this e-mail from your system. E-mail transmission cannot be guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. The sender therefore does not accept liability for any errors or omissions in the contents of this message, which arise as a result of e-mail transmission.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 8/10/10 10:16 PM, Abe, Sumiko wrote:
> Hi, Thank you so much for your great help.
> 
> It doesn't work with the edition of gzip.info file.
> 
> The attached file is the revised gzip.info file and error is following and same as before:
> 
> .....
>   CC       xmalloc.o
>   CC       yesno.o
> cp ./match.c _match.S
> gcc -std=gnu99 -E -I/sw/include  _match.S > _match.i
> sed '/^ * < _match.i > match_.s
> /bin/sh: -c: line 0: unexpected EOF while looking for matching `''
> /bin/sh: -c: line 1: syntax error: unexpected end of file
> make[3]: *** [match.o] Error 2
> make[2]: *** [all-recursive] Error 1
> make[1]: *** [all] Error 2
> make: *** [all-recursive] Error 1
> ### execution of /var/tmp/tmp.1.lgwctH failed, exit code 2
> Removing runtime build-lock...
> Removing build-lock package...
> /sw/bin/dpkg-lockwait -r fink-buildlock-gzip-1.4-2
> (Reading database ... 6073 files and directories currently installed.)
> Removing fink-buildlock-gzip-1.4-2 ...
> Failed: phase compiling: gzip-1.4-2 failed
> 
> Before reporting any errors, please run "fink selfupdate" and
> try again.  If you continue to have issues, please check to see if the
> FAQ on fink's website solves the problem.  If not, ask on the fink-users
> or fink-beginners mailing lists, with a carbon copy to the maintainer:
> 
>         Fink Core Group <fink-core@...>
> 
> Note that this is preferable to emailing the maintainer directly, since
> most fink package maintainers do not have access to all possible
> hardware and software configurations.
> 
> Thank you again,
> 
> Sumiko
> 
> 

Hmm...

You might try using sudo instead of su and see if that works.  Edit
/sw/etc/fink.conf as root and change "RootMethod: su" to "RootMethod:
sudo"  We were discussing yesterday in the Fink IRC channel about a
different package that failed to build when Fink was run as root rather
than under 'sudo' due to differences in the shell options, and this may
be another case of that.


- -- 
Alexander Hansen
Fink User Liaison
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkximecACgkQB8UpO3rKjQ/BOACgonTUpG7jp44PbzqUstZSSKXy
n4AAnid8s7Lrj1/XFbIWkN6898BrHzGo
=jE1G
-----END PGP SIGNATURE-----

Abe, Sumiko wrote:
> Hi, Thank you so much for your great help.
> 
> It doesn't work with the edition of gzip.info file.
> 
> The attached file is the revised gzip.info file and error is following and same as before:
> 
> .....
>   CC       xmalloc.o
>   CC       yesno.o
> cp ./match.c _match.S
> gcc -std=gnu99 -E -I/sw/include  _match.S > _match.i
> sed '/^ * < _match.i > match_.s
> /bin/sh: -c: line 0: unexpected EOF while looking for matching `''
> /bin/sh: -c: line 1: syntax error: unexpected end of file
> make[3]: *** [match.o] Error 2

Here is a different type of guess as to what could be the problem:

What do you get from

grep "ASCPPPOST " /sw/src/fink.build/gzip-1.4-2/gzip-1.4/lib/Makefile

It should be

ASCPPPOST = sed '/^ *#/d; s,//.*,,; s/% /%/g; s/\. /./g'

If this line is truncated before the # sign, then there is indeed a 
problem with the configure script as Alex assumes. If it is the same as 
mine, then the problem would be with your "make" command.

But there is something else that I don't understand anyway: If I look at 
my build of gzip, I don't see any sign of "match.o" being created. That 
is, the lines that lead to the error in your case are never executed in 
my case at all.

Can you post your build log and the config.log file (off-list)?

-- 
Martin

Hi Martin,

Thank you so much for your great information. I run the grep command and it show same result with yours.

sh-3.2#
sh-3.2# grep "ASCPPPOST " /sw/src/fink.build/gzip-1.4-2/gzip-1.4/lib/Makefile
ASCPPPOST = sed '/^ *#/d; s,//.*,,; s/% /%/g; s/\. /./g'
sh-3.2#

Also, Please take a look at the attached config.log file of gzip package.

Any advice would be great appreciated!

Sumiko

-----Original Message-----
From: Martin Costabel [mailto:costabel@...]
Sent: Wednesday, August 11, 2010 6:54 AM
To: Abe, Sumiko
Cc: Core Group
Subject: Re: [fink-core] DCMTK and xmedcon unstable package install...

Abe, Sumiko wrote:
> Hi, Thank you so much for your great help.
>
> It doesn't work with the edition of gzip.info file.
>
> The attached file is the revised gzip.info file and error is following and same as before:
>
> .....
>   CC       xmalloc.o
>   CC       yesno.o
> cp ./match.c _match.S
> gcc -std=gnu99 -E -I/sw/include  _match.S > _match.i
> sed '/^ * < _match.i > match_.s
> /bin/sh: -c: line 0: unexpected EOF while looking for matching `''
> /bin/sh: -c: line 1: syntax error: unexpected end of file
> make[3]: *** [match.o] Error 2

Here is a different type of guess as to what could be the problem:

What do you get from

grep "ASCPPPOST " /sw/src/fink.build/gzip-1.4-2/gzip-1.4/lib/Makefile

It should be

ASCPPPOST = sed '/^ *#/d; s,//.*,,; s/% /%/g; s/\. /./g'

If this line is truncated before the # sign, then there is indeed a
problem with the configure script as Alex assumes. If it is the same as
mine, then the problem would be with your "make" command.

But there is something else that I don't understand anyway: If I look at
my build of gzip, I don't see any sign of "match.o" being created. That
is, the lines that lead to the error in your case are never executed in
my case at all.

Can you post your build log and the config.log file (off-list)?

--
Martin

This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately by e-mail if you have received this e-mail by mistake and delete this e-mail from your system. E-mail transmission cannot be guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. The sender therefore does not accept liability for any errors or omissions in the contents of this message, which arise as a result of e-mail transmission.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 8/11/10 7:26 PM, Abe, Sumiko wrote:
> Hi Martin,
> 
> Thank you so much for your great information. I run the grep command and it show same result with yours.
> 
> sh-3.2#
> sh-3.2# grep "ASCPPPOST " /sw/src/fink.build/gzip-1.4-2/gzip-1.4/lib/Makefile
> ASCPPPOST = sed '/^ *#/d; s,//.*,,; s/% /%/g; s/\. /./g'
> sh-3.2#
> 
> Also, Please take a look at the attached config.log file of gzip package.
> 
> Any advice would be great appreciated!
> 
> Sumiko
> 

While we're collecting data, since Martin mentioned make, what do you
get from

which make
make --version

?

- -- 
Alexander Hansen
Fink User Liaison
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkxjTRAACgkQB8UpO3rKjQ9pJQCgnGtwNISq6DONXuf1XzT/rZ76
4ucAnj6Nm1xj9y8j/snFm6R3mk61W9aM
=y3qR
-----END PGP SIGNATURE-----

Abe, Sumiko wrote:
> Hi Martin,
> 
> Thank you so much for your great information. I run the grep command and it show same result with yours.
> 
> sh-3.2#
> sh-3.2# grep "ASCPPPOST " /sw/src/fink.build/gzip-1.4-2/gzip-1.4/lib/Makefile
> ASCPPPOST = sed '/^ *#/d; s,//.*,,; s/% /%/g; s/\. /./g'

So your sed is OK, and I suppose even your make is OK, and this is 
simply bad makefile syntax. Make apparently considers # as comment 
character, even inside quotes, and interprets the above line as 
equivalent to

  ASCPPPOST = sed '/^ *

So this is a real bug in the gzip build system. They should have escaped 
the # character. Indeed, writing \# instead of # in lib/Makefile makes 
the bug go away. But the bug usually does not show up, see below.

> sh-3.2#
> 
> Also, Please take a look at the attached config.log file of gzip package.

The main difference appears to be that you get

configure:22627: checking for an assembler syntax supported by this package
configure:22657: result: yes

whereas I have result: no.

This accounts for the fact that later on your build process tries to 
handle this "match" stuff where it fails. It looks like this part really 
is buggy, but the bug has never been encountered, because everyone else 
has "no" at the above test.

Staring at the configure script for a while, I get the following strong 
suspicion:

You have a file named "config.h" in either /sw/include/ or 
/usr/local/include/ . If you remove this, you will get the same result 
(no) for the assembler syntax test as me and everyone else, and the 
build will proceed.

This is a situation where one bug (#include <config.h> in lib/match.c) 
masks another (the make syntax error), and you only see the latter 
because of a third bug, namely your stray config.h.

Since gzip is so basic for fink, I think something should be done about 
this, like predefining the result of the above assembler test as "no", 
so that it does not depend on the existence of a stray config.h. Or 
define it as "yes", and fix lib/Makefile and lib/match.c.

Better yet, we should remove the completely superfluous gzip package 
from Fink, together with the other equally superfluous unzip and bzip2 
packages. This would speed up bootstrapping considerably (but I know 
that this is a dream...)

-- 
Martin

Am 12.08.2010 um 11:42 schrieb Martin Costabel:

[...]
> Better yet, we should remove the completely superfluous gzip package 
> from Fink, together with the other equally superfluous unzip and bzip2 
> packages. This would speed up bootstrapping considerably (but I know 
> that this is a dream...)

Are they completely superfluous?

On my 10.6 machine, I get this:

$ /sw/bin/gzip --version
gzip 1.4
[...]
$ /usr/bin/gzip --version
gzip 1.3.12
[...]

Likewise, our unzip is UnZip 6.00 of 20 April 2009, while the one shipped by OS X is UnZip 5.52 of 28 February 2005. Only bzip2 have identical version.

However, there seems to be no need anymore to have them marked as essential packages. That would still save bootstrap time.

Bye,
Max

Hi All,

Thank you so much for your great help.
I remove the file /usr/include/config.h  and the building process in passing through!
I am so happy to get it here!

Thank you so much for all of your help!

Sumiko


-----Original Message-----
From: Martin Costabel [mailto:costabel@...]
Sent: Thursday, August 12, 2010 2:43 AM
To: Abe, Sumiko
Cc: Fink Core Group
Subject: Re: [fink-core] DCMTK and xmedcon unstable package install...

Abe, Sumiko wrote:
> Hi Martin,
>
> Thank you so much for your great information. I run the grep command and it show same result with yours.
>
> sh-3.2#
> sh-3.2# grep "ASCPPPOST " /sw/src/fink.build/gzip-1.4-2/gzip-1.4/lib/Makefile
> ASCPPPOST = sed '/^ *#/d; s,//.*,,; s/% /%/g; s/\. /./g'

So your sed is OK, and I suppose even your make is OK, and this is
simply bad makefile syntax. Make apparently considers # as comment
character, even inside quotes, and interprets the above line as
equivalent to

  ASCPPPOST = sed '/^ *

So this is a real bug in the gzip build system. They should have escaped
the # character. Indeed, writing \# instead of # in lib/Makefile makes
the bug go away. But the bug usually does not show up, see below.

> sh-3.2#
>
> Also, Please take a look at the attached config.log file of gzip package.

The main difference appears to be that you get

configure:22627: checking for an assembler syntax supported by this package
configure:22657: result: yes

whereas I have result: no.

This accounts for the fact that later on your build process tries to
handle this "match" stuff where it fails. It looks like this part really
is buggy, but the bug has never been encountered, because everyone else
has "no" at the above test.

Staring at the configure script for a while, I get the following strong
suspicion:

You have a file named "config.h" in either /sw/include/ or
/usr/local/include/ . If you remove this, you will get the same result
(no) for the assembler syntax test as me and everyone else, and the
build will proceed.

This is a situation where one bug (#include <config.h> in lib/match.c)
masks another (the make syntax error), and you only see the latter
because of a third bug, namely your stray config.h.

Since gzip is so basic for fink, I think something should be done about
this, like predefining the result of the above assembler test as "no",
so that it does not depend on the existence of a stray config.h. Or
define it as "yes", and fix lib/Makefile and lib/match.c.

Better yet, we should remove the completely superfluous gzip package
from Fink, together with the other equally superfluous unzip and bzip2
packages. This would speed up bootstrapping considerably (but I know
that this is a dream...)

--
Martin






This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately by e-mail if you have received this e-mail by mistake and delete this e-mail from your system. E-mail transmission cannot be guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. The sender therefore does not accept liability for any errors or omissions in the contents of this message, which arise as a result of e-mail transmission.

Hi Martin,

Thank you so much for all of your great help.
The gzip package are built successfully. Also, I have built xmedcon too, it works on my macbook now, very happy, :-)

But for DCMTK, not happy, :. It still could not be built...

The following is the output error, and the attached if is the config.log from dcmtk.

................
c++     -DHAVE_CONFIG_H  -DNDEBUG  -c -I. -I. -I../include -I../../config/include   \
                -I/sw/include -O -I/sw/include/libxml2 -I/sw/include -D_REENTRANT -D_XOPEN_SOURCE_EXTENDED -D_XOPEN_SOURCE=500 -D_BSD_SOURCE -D_BSD_COMPAT -D_OSF_SOURCE -D_POSIX_C_SOURCE=199506L -Wall  oflist.cc
c++     -DHAVE_CONFIG_H  -DNDEBUG  -c -I. -I. -I../include -I../../config/include   \
                -I/sw/include -O -I/sw/include/libxml2 -I/sw/include -D_REENTRANT -D_XOPEN_SOURCE_EXTENDED -D_XOPEN_SOURCE=500 -D_BSD_SOURCE -D_BSD_COMPAT -D_OSF_SOURCE -D_POSIX_C_SOURCE=199506L -Wall  ofstring.cc
In file included from /usr/include/sys/attr.h:42,
                 from /usr/include/sys/mount.h:76,
                 from /usr/include/libc.h:45,
                 from ../include/dcmtk/ofstd/ofstdinc.h:320,
                 from ../include/dcmtk/ofstd/ofstring.h:61,
                 from ofstring.cc:47:
/usr/include/sys/ucred.h:91: error: 'u_long' does not name a type
/usr/include/sys/ucred.h:137: error: 'u_int' does not name a type
In file included from /usr/include/sys/mount.h:76,
                 from /usr/include/libc.h:45,
                 from ../include/dcmtk/ofstd/ofstdinc.h:320,
                 from ../include/dcmtk/ofstd/ofstring.h:61,
                 from ofstring.cc:47:
/usr/include/sys/attr.h:77: error: 'u_short' does not name a type
/usr/include/sys/attr.h:430: error: 'u_long' does not name a type
/usr/include/sys/attr.h:442: error: 'u_char' does not name a type
make[2]: *** [ofstring.o] Error 1
make[1]: *** [libsrc-all] Error 2
make: *** [ofstd-all] Error 2
### execution of make failed, exit code 2
Removing runtime build-lock...
Removing build-lock package...
/sw/bin/dpkg-lockwait -r fink-buildlock-dcmtk-3.5.4-1002
(Reading database ... 25650 files and directories currently installed.)
Removing fink-buildlock-dcmtk-3.5.4-1002 ...
Failed: phase compiling: dcmtk-3.5.4-1002 failed

Before reporting any errors, please run "fink selfupdate" and
try again.  If you continue to have issues, please check to see if the
FAQ on fink's website solves the problem.  If not, ask on the fink-users
or fink-beginners mailing lists, with a carbon copy to the maintainer:

        Bernd Kuemmerlen <bkuemmer@...>

Note that this is preferable to emailing the maintainer directly, since
most fink package maintainers do not have access to all possible
hardware and software configurations.


Sorry for interrupt you again,

Sumiko

-----Original Message-----
From: Martin Costabel [mailto:costabel@...]
Sent: Thursday, August 12, 2010 2:43 AM
To: Abe, Sumiko
Cc: Fink Core Group
Subject: Re: [fink-core] DCMTK and xmedcon unstable package install...

Abe, Sumiko wrote:
> Hi Martin,
>
> Thank you so much for your great information. I run the grep command and it show same result with yours.
>
> sh-3.2#
> sh-3.2# grep "ASCPPPOST " /sw/src/fink.build/gzip-1.4-2/gzip-1.4/lib/Makefile
> ASCPPPOST = sed '/^ *#/d; s,//.*,,; s/% /%/g; s/\. /./g'

So your sed is OK, and I suppose even your make is OK, and this is
simply bad makefile syntax. Make apparently considers # as comment
character, even inside quotes, and interprets the above line as
equivalent to

  ASCPPPOST = sed '/^ *

So this is a real bug in the gzip build system. They should have escaped
the # character. Indeed, writing \# instead of # in lib/Makefile makes
the bug go away. But the bug usually does not show up, see below.

> sh-3.2#
>
> Also, Please take a look at the attached config.log file of gzip package.

The main difference appears to be that you get

configure:22627: checking for an assembler syntax supported by this package
configure:22657: result: yes

whereas I have result: no.

This accounts for the fact that later on your build process tries to
handle this "match" stuff where it fails. It looks like this part really
is buggy, but the bug has never been encountered, because everyone else
has "no" at the above test.

Staring at the configure script for a while, I get the following strong
suspicion:

You have a file named "config.h" in either /sw/include/ or
/usr/local/include/ . If you remove this, you will get the same result
(no) for the assembler syntax test as me and everyone else, and the
build will proceed.

This is a situation where one bug (#include <config.h> in lib/match.c)
masks another (the make syntax error), and you only see the latter
because of a third bug, namely your stray config.h.

Since gzip is so basic for fink, I think something should be done about
this, like predefining the result of the above assembler test as "no",
so that it does not depend on the existence of a stray config.h. Or
define it as "yes", and fix lib/Makefile and lib/match.c.

Better yet, we should remove the completely superfluous gzip package
from Fink, together with the other equally superfluous unzip and bzip2
packages. This would speed up bootstrapping considerably (but I know
that this is a dream...)

--
Martin






This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately by e-mail if you have received this e-mail by mistake and delete this e-mail from your system. E-mail transmission cannot be guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. The sender therefore does not accept liability for any errors or omissions in the contents of this message, which arise as a result of e-mail transmission.

Sorry, I forgot to attach the log file.



-----Original Message-----
From: Abe, Sumiko
Sent: Thursday, August 12, 2010 2:31 PM
To: 'Martin Costabel'
Cc: Fink Core Group; 'Bernd Kuemmerlen'
Subject: RE: [fink-core] DCMTK and xmedcon unstable package install...

Hi Martin,

Thank you so much for all of your great help.
The gzip package are built successfully. Also, I have built xmedcon too, it works on my macbook now, very happy, :-)

But for DCMTK, not happy, :. It still could not be built...

The following is the output error, and the attached if is the config.log from dcmtk.

................
c++     -DHAVE_CONFIG_H  -DNDEBUG  -c -I. -I. -I../include -I../../config/include   \
                -I/sw/include -O -I/sw/include/libxml2 -I/sw/include -D_REENTRANT -D_XOPEN_SOURCE_EXTENDED -D_XOPEN_SOURCE=500 -D_BSD_SOURCE -D_BSD_COMPAT -D_OSF_SOURCE -D_POSIX_C_SOURCE=199506L -Wall  oflist.cc
c++     -DHAVE_CONFIG_H  -DNDEBUG  -c -I. -I. -I../include -I../../config/include   \
                -I/sw/include -O -I/sw/include/libxml2 -I/sw/include -D_REENTRANT -D_XOPEN_SOURCE_EXTENDED -D_XOPEN_SOURCE=500 -D_BSD_SOURCE -D_BSD_COMPAT -D_OSF_SOURCE -D_POSIX_C_SOURCE=199506L -Wall  ofstring.cc
In file included from /usr/include/sys/attr.h:42,
                 from /usr/include/sys/mount.h:76,
                 from /usr/include/libc.h:45,
                 from ../include/dcmtk/ofstd/ofstdinc.h:320,
                 from ../include/dcmtk/ofstd/ofstring.h:61,
                 from ofstring.cc:47:
/usr/include/sys/ucred.h:91: error: 'u_long' does not name a type
/usr/include/sys/ucred.h:137: error: 'u_int' does not name a type
In file included from /usr/include/sys/mount.h:76,
                 from /usr/include/libc.h:45,
                 from ../include/dcmtk/ofstd/ofstdinc.h:320,
                 from ../include/dcmtk/ofstd/ofstring.h:61,
                 from ofstring.cc:47:
/usr/include/sys/attr.h:77: error: 'u_short' does not name a type
/usr/include/sys/attr.h:430: error: 'u_long' does not name a type
/usr/include/sys/attr.h:442: error: 'u_char' does not name a type
make[2]: *** [ofstring.o] Error 1
make[1]: *** [libsrc-all] Error 2
make: *** [ofstd-all] Error 2
### execution of make failed, exit code 2
Removing runtime build-lock...
Removing build-lock package...
/sw/bin/dpkg-lockwait -r fink-buildlock-dcmtk-3.5.4-1002
(Reading database ... 25650 files and directories currently installed.)
Removing fink-buildlock-dcmtk-3.5.4-1002 ...
Failed: phase compiling: dcmtk-3.5.4-1002 failed

Before reporting any errors, please run "fink selfupdate" and
try again.  If you continue to have issues, please check to see if the
FAQ on fink's website solves the problem.  If not, ask on the fink-users
or fink-beginners mailing lists, with a carbon copy to the maintainer:

        Bernd Kuemmerlen <bkuemmer@...>

Note that this is preferable to emailing the maintainer directly, since
most fink package maintainers do not have access to all possible
hardware and software configurations.


Sorry for interrupt you again,

Sumiko

-----Original Message-----
From: Martin Costabel [mailto:costabel@...]
Sent: Thursday, August 12, 2010 2:43 AM
To: Abe, Sumiko
Cc: Fink Core Group
Subject: Re: [fink-core] DCMTK and xmedcon unstable package install...

Abe, Sumiko wrote:
> Hi Martin,
>
> Thank you so much for your great information. I run the grep command and it show same result with yours.
>
> sh-3.2#
> sh-3.2# grep "ASCPPPOST " /sw/src/fink.build/gzip-1.4-2/gzip-1.4/lib/Makefile
> ASCPPPOST = sed '/^ *#/d; s,//.*,,; s/% /%/g; s/\. /./g'

So your sed is OK, and I suppose even your make is OK, and this is
simply bad makefile syntax. Make apparently considers # as comment
character, even inside quotes, and interprets the above line as
equivalent to

  ASCPPPOST = sed '/^ *

So this is a real bug in the gzip build system. They should have escaped
the # character. Indeed, writing \# instead of # in lib/Makefile makes
the bug go away. But the bug usually does not show up, see below.

> sh-3.2#
>
> Also, Please take a look at the attached config.log file of gzip package.

The main difference appears to be that you get

configure:22627: checking for an assembler syntax supported by this package
configure:22657: result: yes

whereas I have result: no.

This accounts for the fact that later on your build process tries to
handle this "match" stuff where it fails. It looks like this part really
is buggy, but the bug has never been encountered, because everyone else
has "no" at the above test.

Staring at the configure script for a while, I get the following strong
suspicion:

You have a file named "config.h" in either /sw/include/ or
/usr/local/include/ . If you remove this, you will get the same result
(no) for the assembler syntax test as me and everyone else, and the
build will proceed.

This is a situation where one bug (#include <config.h> in lib/match.c)
masks another (the make syntax error), and you only see the latter
because of a third bug, namely your stray config.h.

Since gzip is so basic for fink, I think something should be done about
this, like predefining the result of the above assembler test as "no",
so that it does not depend on the existence of a stray config.h. Or
define it as "yes", and fix lib/Makefile and lib/match.c.

Better yet, we should remove the completely superfluous gzip package
from Fink, together with the other equally superfluous unzip and bzip2
packages. This would speed up bootstrapping considerably (but I know
that this is a dream...)

--
Martin






This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately by e-mail if you have received this e-mail by mistake and delete this e-mail from your system. E-mail transmission cannot be guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. The sender therefore does not accept liability for any errors or omissions in the contents of this message, which arise as a result of e-mail transmission.