gmane.os.apple.fink.devel

Re: not quite there yet with gdal 1.10.0

Kurt Schwehr — 2013-06-19T06:44:54

BABA and dmacks,

gdal 1.10.0 is in fink 10.[78]

Thanks for the help.  I ended up going with SetLDFLAGS: -lsqlite3.  That adds an extra shared library linked against the raster side of the tools, but that seems like a small cost to get it working.

gdal should be a lot less messy now...

-kurt

On Jun 17, 2013, at 2:57 PM, Daniel Macks <dmacks< at >netspace.org> wrote:



------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Re: Fink's mac-fsevents-pm5NNN fails to build on 10.6 and 10.7

Hanspeter Niederstrasser — 2013-06-18T18:26:06

I checked in 0.09 to the 10.7 tree.  I also tried adding it to the 
10.5/6 tree, but it failed there with errors like "Failed test 'history 
event received'", and without knowing what could fix those, I kept the 
10.5/6 tree at the previous version.

Hanspeter


------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

io-compress 2.061 fails bzip2 tests on OS X 10.7

Hanspeter Niederstrasser — 2013-06-18T14:45:04

While building io-compress-2.0.61 on OS X 10.7, all the bzip2 tests 
fail.  I am using Apple's perl 5.12.3.  compres-raw-bzip2 is at 2.061, 
and it passed all of its tests.


This is the build log truncated around the first couple failing tests. 
The full log is at 
<http://www.snaggledworks.com/fink/logs/fink-build-log_io-compress-pm5123_2.061-1_2013.06.18-10.37.50>


+ make test
PERL_DL_NONLAZY=1 /usr/bin/arch -x86_64 perl5.12 
"-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib', 
'blib/arch')" t/*.t
t/000prereq.t ................... ok
t/001bzip2.t .................... ok

....

t/100generic-zip.t .............. ok
bad bad bad at t/compress/CompTestUtils.pm line 647.
# Looks like your test died before it could output anything.
t/101truncate-bzip2.t ...........
Dubious, test returned 255 (wstat 65280, 0xff00)
Failed 4012/4012 subtests
t/101truncate-deflate.t ......... ok

....

t/103newtied-zip.t .............. ok

#     Failed test (t/compress/destroy.pl at line 52)
#          got: ''
#     expected: 'hello world
# this is a test
# '

#     Failed test (t/compress/destroy.pl at line 74)
# Looks like you failed 2 tests of 15.
t/104destroy-bzip2.t ............
Dubious, test returned 2 (wstat 512, 0x200)
Failed 2/15 subtests
t/104destroy-deflate.t .......... ok
t/104destroy-gzip.t ............. ok
t/104destroy-rawdeflate.t ....... ok
t/104destroy-zip.t .............. ok

#     Failed test (t/compress/oneshot.pl at line 305)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 319)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 334)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 349)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 367)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 385)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 404)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 420)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 440)
#          got: 'abcde'
#     expected: ''

#     Failed test (t/compress/oneshot.pl at line 460)
#          got: 'abcde'
#     expected: ''

#     Failed test (t/compress/oneshot.pl at line 477)
#          got: 'abcde'
#     expected: ''

#     Failed test (t/compress/oneshot.pl at line 500)
#          got: 'abcde'
#     expected: ''

#     Failed test (t/compress/oneshot.pl at line 305)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 319)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 334)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 349)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 367)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 385)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 404)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 420)
#          got: ''
#     expected: 'abcde'

#     Failed test (t/compress/oneshot.pl at line 440)
#          got: 'abcde'
#     expected: ''

#     Failed test (t/compress/oneshot.pl at line 460)
#          got: 'abcde'
#     expected: ''

#     Failed test (t/compress/oneshot.pl at line 477)
#          got: 'abcde'
#     expected: ''

#     Failed test (t/compress/oneshot.pl at line 500)
#          got: 'abcde'
#     expected: ''

#     Failed test (t/compress/oneshot.pl at line 551)
#     Structures begin differing at:
#          $got->[1] = ''
#     $expected->[1] = 'data1'
Error uncompressing -- Inflation Error: data error at 
t/compress/CompTestUtils.pm line 545.
main::anyUncompress('ARRAY(0x7fd529a72df8)') called at 
t/compress/oneshot.pl line 567
main::run() called at t/105oneshot-bzip2.t line 22
# Looks like you planned 989 tests but only ran 475.
# Looks like your test died just after 475.
t/105oneshot-bzip2.t ............
Dubious, test returned 255 (wstat 65280, 0xff00)
Failed 539/989 subtests
(less 2 skipped subtests: 448 okay)
t/105oneshot-deflate.t .......... ok

And it keeps going on like that, failing all bzip2 tests.

Hanspeter

------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Re: not quite there yet with gdal 1.10.0

Kurt Schwehr — 2013-06-18T00:28:33

Thanks dmacks and baba!  Gdal is most definitely a messy file. Will take a look soon

Sent from my iPhone

On Jun 17, 2013, at 2:57 PM, "Daniel Macks" <dmacks< at >netspace.org> wrote:


------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Re: not quite there yet with gdal 1.10.0

Daniel Macks — 2013-06-17T21:57:58

On Sun, 16 Jun 2013 22:13:49 -0700, Kurt Schwehr 
<kurtschwehr< at >yahoo.com> wrote:
having trouble getting gdal 1.10.0 to build with any of the 3 
libspatialite versions.  if any one has ideas, I'd appreciate any help. 

I left a mile of comments on gist (lots of misc cleanups, but also 
solving this linking problem). 

dan

  --
Daniel Macks
dmacks< at >netspace.org



------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Re: not quite there yet with gdal 1.10.0

BABA Yoshihiko — 2013-06-17T21:26:58

Kurt,

You may have known, but I googled around and got some information that may
be related.

"Current theory is that the MacOS SQLite package was not compiled with the
SQLITE_ENABLE_COLUMN_METADATA flag"
https://bugs.launchpad.net/zorba/+bug/1102549

"The final link line misses -lsqlite, and therefore the SQLite dependency
is not honored, whereby the error. It should be a very simple matter to
patch this."
http://trac.macports.org/ticket/39214

--
BABA Yoshihiko


2013/6/17 Kurt Schwehr <kurtschwehr< at >yahoo.com>

not quite there yet with gdal 1.10.0

Kurt Schwehr — 2013-06-17T05:13:49

having trouble getting gdal 1.10.0 to build with any of the 3 libspatialite versions.  if any one has ideas, I'd appreciate any help.

I've updated a few of the libraries and will add pcre once spatialite/rasterlite work

https://gist.github.com/schwehr/5794715

-kurt------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Re: Fink's mac-fsevents-pm5NNN fails to build on 10.6and 10.7

Steve Huff — 2013-06-16T21:05:50



looks like my previous message may not have gotten through.  anyway, i don't see that error on 10.7; i see a different one, concerning POD coverage, and the easy way to get around that is not not run the POD tests.  attached is a slightly updated .info that works for me on 10.7 x86_64.

-shuff

Re: Fink's mac-fsevents-pm5NNN fails to build on 10.6and 10.7

Steve Huff — 2013-06-16T20:36:43

On Jun 16, 2013, at 7:29 AM, Hanspeter Niederstrasser wrote:



huh - this is the error i see under 10.7 with Perl 5.12.3:

t/03podcoverage.t .......... 1/2
#   Failed test 'Pod coverage on Mac::FSEvents'
#   at /sw/lib/perl5/Test/Pod/Coverage.pm line 126.
# Coverage for Mac::FSEvents is 88.9%, with 1 naked subroutine:
# constant
# Looks like you failed 1 test of 2.

that seems like an upstream bug.  i've opened https://rt.cpan.org/Ticket/Display.html?id=86196 to track it.

i don't have a 10.6 system handy for testing, so the solution there may have to be just removing 10.6 from the list of supported distributions for that package.

-shuff

Fink's mac-fsevents-pm5NNN fails to build on 10.6 and10.7

Hanspeter Niederstrasser — 2013-06-16T11:29:39

The pkg mac-fsevents-pm5NNN fails to build on 10.6 with this error (10.7 
error is similar):

TEST_POD=1 make test || exit 2
PERL_DL_NONLAZY=1 /sw32/bin/perl5.8.8 "-MExtUtils::Command::MM" "-e" 
"test_harness(0, 'blib/lib', 'blib/arch')" t/*.t
t/01use............ok 

t/02pod............ok 

t/03podcoverage....ok 

t/04critic.........skipped
         all skipped: Critic test only for developers.
t/05event..........
#   Failed test 'event received (poll interface)'
#   in t/05event.t at line 58.
t/05event..........ok 3/7Use of uninitialized value in subroutine entry 
at t/05event.t line 109.

#   Failed test 'history event received'
#   in t/05event.t at line 139.
t/05event..........NOK 5# Looks like you planned 7 tests but only ran 5. 

# Looks like you failed 2 tests of 5 run.
t/05event..........dubious 

Test returned status 2 (wstat 512, 0x200)
DIED. FAILED tests 1, 5-7
Failed 4/7 tests, 42.86% okay
Failed Test Stat Wstat Total Fail  Failed  List of Failed
-------------------------------------------------------------------------------
t/05event.t    2   512     7    6  85.71%  1 5-7
1 test skipped.

Hanspeter

------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Re: ping on git

Daniel Johnson — 2013-06-16T01:17:19

On Jun 13, 2013, at 8:23 PM, Alexander Hansen <alexanderk.hansen< at >gmail.com> wrote:

You wouldn't be able to check in anything with it; it's read-only. :)

I looked at doing a repo using branches; converting the repo is easy but there would have to be a number of changes throughout the fink code to deal with it. It would also break our use of rsync since rsync doesn't understand branches. We would have to completely redo the rsync mirroring code to check out each branch and create a copy each.

Ok, here are the steps to make your very own git conversion.

I've found that the only reliable way to convert is with the cvs2git tool which is part of the cvs2svn project. You can install it with fink by using the cvs2svn-py27 package. The disadvantage of using cvs2git is that it can't do incremental updates of a converted repo. However, once the initial conversion is done, one can use 'git cvsimport' to do incremental updates. That's how I keep my mirror up-to-date.

To use cvs2git, you need an actual copy of the cvs repo, not a checked out working copy. You can do that with
rsync -av rsync://fink.cvs.sourceforge.net/cvsroot/fink/dists /path/to/fink-rsync

Next you'll need my customized cvs2git options file which is attached at the end of this email. You must edit the text '/Users/daniel/src/fink-rsync/dists' in it to point to whatever you used for /path/to/fink-rsync.

Now run the command thusly: cvs2git --options=cvs2git-fink-dists.options

That command should have created a directory called cvs2svn-tmp with two files in it: git-blob.dat and git-dump.dat.

Next create an empty git repo:
mkdir fink-dists.git
cd fink-dists.git
git init

Then we import the converted repo into git!
cat ../cvs2svn-tmp/git-blob.dat ../cvs2svn-tmp/git-dump.dat | git fast-import

Now we want to cleanup any tag fixup branches that might have been created by running /sw/share/doc/cvs2svn-py27/git-move-refs.py from within the repo.

Compact the new repo with 'git gc --prune=now' and checkout a working copy with 'git checkout'.

And now you should have a complete dists repo in git! Setting things up for incremental updates is a bit more complicated but I can go into it in another email. This should be enough to play around with though. You wouldn't want to keep updating it anyway if you want to play around with checking in updates via git.

To use your new repo with fink's selfupdate-git, you'll need to upload it to your own github account or wherever else you want, and also have to edit the git.pm file to point to your own repo instead of mine.

Daniel

------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Re: tcltk threading

Daniel Macks — 2013-06-14T20:24:39

Any chance that you'd be okay with adding threading to the tcltk 
client?  I've got a bug in displaying images in PIL/ImageTk.py that 
goes away if tcltk is build with --enable-threads?

I have no objections. Sounds like a good idea because "someone wants 
it", and I don't know anything about it to have an opinion either way. 

dan

  --
Daniel Macks
dmacks< at >netspace.org



------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Re: java on 10.8 and later

Alexander Hansen — 2013-06-14T17:33:20

The runtime part of JavaVM.framework (legacy) comes with 10.8 by 
default.  The compile-time elements are installed with the Xcode 
Command-Line Tools.

Re: java on 10.8 and later

Jack Howarth — 2013-06-14T15:30:15

Depends how you define subsumed. The installed JDK does have...

/Library/Java/JavaVirtualMachines/jdk1.7.0_21.jdk/Contents/Home/jre

which wouldn't be a huge pain if they had created a symlink like...

/Library/Java/JavaVirtualMachines/jdk1.7.jdk 

to point at the location of the currently installed JDK 1.7...

/Library/Java/JavaVirtualMachines/jdk1.7.0_21.jdk

such that the symlink would get updated when an updated JDK 1.7 was installed.
          Jack


------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Re: java on 10.8 and later

Alexander Hansen — 2013-06-14T15:22:06

On 10.8, I _think_ the legacy Java framework is still installed by 
default.  Rather than making assumptions, though, I can create a virtual 
machine and find out.

Re: java on 10.8 and later

Alexander Hansen — 2013-06-14T15:07:29

Knock yourself out.  I'm not a subscriber.


Oracle _claimed_ that the JDK subsumed the JRE, but after checking on a 
system with a clean 10.8 install where I only installed the JRE, I don't 
have /Library/Java/Home.

Re: java on 10.8 and later

Jack Howarth — 2013-06-14T14:43:33

 

It also may be help to bring this issue up on the macports-dev mailing list to
see what information they have on the problem and how it is to be addressed there.
The only obvious approach I see would be to force users to install both the jre and jdk
but have the builds use the headers from the jdk but link against the libs in
the jre.
              Jack


------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

java on 10.8 and later

Alexander Hansen — 2013-06-14T14:21:12

The discussion below also applies to 10.8.

-------- Original Message --------
Subject: Re: [fink-seed] 10.9
Date: Fri, 14 Jun 2013 07:06:55 -0700
From: Alexander Hansen <alexanderk.hansen< at >gmail.com>
Organization: Fink Core Team
To: fink ADC/seed NDA list <fink-seed< at >lists.sourceforge.net>
CC: Jack Howarth <howarth< at >bromo.med.uc.edu>

On 6/14/13 6:52 AM, Jack Howarth wrote:

<snip>


According to Oracle, the JDK includes the JRE, and this appears to be
the case on my system, too.


eww, right.  A specific example being:

$ ls -l
/Library/Java/JavaVirtualMachines/jdk1.7.0_21.jdk/Contents/Home/jre/lib/libawt.dylib 


-rwxrwxr-x  1 root  wheel  563940 Apr  7 15:10
/Library/Java/JavaVirtualMachines/jdk1.7.0_21.jdk/Contents/Home/jre/lib/libawt.dylib
$ ls -l /Library/Java/Home/bundle/Libraries/libawt.jnilib
-rwxr-xr-x  1 root  wheel  4652208 Apr 16 14:10
/Library/Java/Home/bundle/Libraries/libawt.jnilib

We've got a current issue on fink about the system-java virtual packages 
for 10.8 related to this, too.  It took me this long to figure out 
_what_ the heck I was actually supposed to download and install so that 
I could actually experience the issue.  This only affects dependencies 
and not actual package builds.
--
Alexander Hansen, Ph.D.
Fink User Liaison
My package updates: http://finkakh.wordpress.com/



------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Re: python packagers - please switch from using distribute back to setuptools

Alexander Hansen — 2013-06-14T02:09:22

I found something which looks like it works.  The gist is:

1) Make distribute a real package again but have it really be 
setuptools-0.7.2.  This has the following:
Conflicts: setuptools-py%type_pkg[python] (<< 0.7.2-3)
Replaces: setuptools-py%type_pkg[python]
Provides: setuptools-py%type_pkg[python]

The unversioned Replaces: will allow it to replace setuptools (>= 
0.7.2-3) without removing the package entry, and the Provides: keeps 
current users of setuptools (via the prior Provides from the old 
distribute) happy.

2) Do the following in setuptools:

Conflicts: distribute-py%type_pkg[python] (<< 0.7.2-3)
Replaces: distribute-py%type_pkg[python]
Provides: distribute-py%type_pkg[python]

This allows folks who update distribute automatically to get setuptools 
(under the distribute label) without any nasty dependency hell.

Then, if setuptools is updated and installed, it overwrites the 
overlapping files from distribute with identical ones, but does _not_ 
remove the distribute package (since there is no Conflict), so packages 
with dependencies on it are happy.

It's not a complete solution, probably, but at least we can tell users 
to update distribute first.

.infos attached.

ping on git

Alexander Hansen — 2013-06-14T00:23:33

I'd like to try to move forward on getting the distribution files moved 
over to git.  Right now there is a selfupdate-git branch on fink's 
github repository, but I don't know where we stand on functionality. 
Last time I used it I was able to update from Daniel J's dists/ mirror, 
but I didn't try operations like checking in updates and the like.

Max had proposed at one point using git branches to handle distributions 
instead of the current separate directory trees.  Unfortunately, I don't 
know if at our current manpower levels (especially manpower with 
sufficient knowledge of git) will allow us to implement this in a timely 
manner.  We definitely are going to want to get _something_ in place.

Daniel, if you could reply to this to let folks know how to translate 
the existing CVS repository so that folks can set up their own github 
copies of it to test the ability to check updates in and the like, 
that'd be great.

tcltk threading

Kurt Schwehr — 2013-06-13T22:46:34

Any chance that you'd be okay with adding threading to the tcltk client?  I've got a bug in displaying images in PIL/ImageTk.py that goes away if tcltk is build with --enable-threads?

-kurt
(goatbar)

RCS file: /cvsroot/fink/dists/10.7/stable/main/finkinfo/languages/tcltk-x86_64.info,v
retrieving revision 1.9
diff -r1.9 tcltk-x86_64.info
4c4
< Revision: 1
---
30c30
< ConfigureParams: --enable-shared --disable-corefoundation --exec-prefix=%p --mandir=%p/share/man --x-includes=/usr/X11R6/include --x-libraries=/usr/X11R6/lib tcl_cv_type_64bit="long long"
---


------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Fink-devel mailing list
Fink-devel< at >lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel