Re: NEW: audio/last.fm

[Rodolfo Gouveia]

Compiled and installed fine on i386.
Tested with xmms, but scrobbling doesn't appear to be working.
Also if I start last.fm first, it blocks the audio card so xmms can't use it.
When exitting it spits this error out:
last.fm in realloc(): error: chunk is already free
Abort trap 

Rodolfo Gouveia

Re: NEW: devel/py-nose

[steven mestdagh]

Niall O'Higgins [2007-04-27, 23:25:34]:
> On Fri, Apr 27, 2007 at 04:56:24PM -0500, Will Maier wrote:
> > On Fri, Apr 27, 2007 at 10:33:39PM +0100, Niall O'Higgins wrote:
> > > Port for 'nose' python unittest framework.  Required for regress of
> > > py-simplejson (which I will post next).  Tested on amd64. 
> > 
> > I posted a version[0] of this for 0.9.1 on 4 January:
> > 
> > http://marc.info/?l=openbsd-ports&m=116791585912447&w=2
> 
> Ah, sorry.  I tend to just port this stuff when I need to install it,
> since its so simple, and then send it off to ports@ to share/archive. 
> 
> > I like my version better, though the differences are mostly
> > stylistic. ;) Some suggestions:
> > 
> > * shorten COMMENT (I used "discovery-based unittests for
> >   Python")
> > * V=NNN variables seem unnecessary
> > * ${DISTNAME:S/^/py-/} seems more idiomatic
> 
> I have no real preference here, just from another Python port.  Yours is
> shorter so I'll use that. 
> 
> > * why add www to CATEGORIES?
> 
> Good point.  I missed that, replaced it with devel/.
> 
> Port updated @ http://unworkable.org/~niallo/py-nose.tar.gz

some errors occur in the tests, log below (amd64).

===>  Regression check for py-nose-0.9.3
running test
running egg_info
writing nose.egg-info/PKG-INFO
writing top-level names to nose.egg-info/top_level.txt
writing dependency_links to nose.egg-info/dependency_links.txt
writing entry points to nose.egg-info/entry_points.txt
reading manifest file 'nose.egg-info/SOURCES.txt'
reading manifest template 'MANIFEST.in'
warning: no files found matching 'unit_tests/*/*/*/*/*.py'
writing manifest file 'nose.egg-info/SOURCES.txt'
running build_ext
collector in ['/a/ports/obj/py-nose-0.9.3/nose-0.9.3'] ... test module nose in 
/a/ports/obj/py-nose-0.9.3/nose-0.9.3/lib ... test directory 
/a/ports/obj/py-nose-0.9.3/nose-0.9.3/lib/nose in nose ... test module 
nose.plugins in /a/ports/obj/py-nose-0.9.3/nose-0.9.3/lib ... test directory 
/a/ports/obj/py-nose-0.9.3/nose-0.9.3/lib/nose/plugins in nose.plugins ... test 
module nose in /a/ports/obj/py-nose-0.9.3/nose-0.9.3 ... test directory 
/a/ports/obj/py-nose-0.9.3/nose-0.9.3/nose in nose ... test module nose.plugins 
in /a/ports/obj/py-nose-0.9.3/nose-0.9.3 ... test directory 
/a/ports/obj/py-nose-0.9.3/nose-0.9.3/nose/plugins in nose.plugins ... test 
module test_bug105 in /a/ports/obj/py-nose-0.9.3/nose-0.9.3/unit_tests ... test 
class  ... test_load_in_def_order 
(test_bug105.TestBug105) ... ERROR
test module test_cases in /a/ports/obj/py-nose-0.9.3/nose-0.9.3/unit_tests ... 
test class  ... test_function_test_case 
(test_cases.TestNoseCases) ... ERROR
test_function_test_case_fixtures (test_cases.TestNoseCases) ... ERROR
test_method_test_case (test_cases.TestNoseCases) ... ERROR
test module test_collector in /a/ports/obj/py-nose-0.9.3/nose-0.9.3/unit_tests 
... test class  ... 
test_basic_collection (test_collector.TestNoseCollector) ... ok
test_deep_collection (test_collector.TestNoseCollector) ... ok
test module test_config in /a/ports/obj/py-nose-0.9.3/nose-0.9.3/unit_tests ... 
test class  ... test_defaults 
(test_config.TestNoseConfig) ... ERROR
test_multiple_include (test_config.TestNoseConfig) ... ERROR
test_reset (test_config.TestNoseConfig) ... ERROR
test_single_include (test_config.TestNoseConfig) ... ERROR
test_update (test_config.TestNoseConfig) ... ERROR
test class  ... test_empty_files 
(test_config.TestNoseConfigFile) ... ERROR
test_missing_file (test_config.TestNoseConfigFile) ... ERROR
test_more_config (test_config.TestNoseConfigFile) ... ERROR
test_no_files (test_config.TestNoseConfigFile) ... ERROR
test class  ... test module test_core in 
/a/ports/obj/py-nose-0.9.3/nose-0.9.3/unit_tests ... test class  ... test_restore_stdout (test_core.TestAPI_run) ... 
ERROR
test class  ... test_init_arg_defaultTest 
(test_core.TestTestProgram) ... ERROR
test_init_arg_module (test_core.TestTestProgram) ... ERROR
test module test_doctest_error_handling in 
/a/ports/obj/py-nose-0.9.3/nose-0.9.3/unit_tests ... ERROR
test module test_importer in /a/ports/obj/py-nose-0.9.3/nose-0.9.3/unit_tests 
... test class  ... test_add_paths 
(test_importer.TestImporter) ... ERROR
test_import (test_importer.TestImporter) ... ERROR
test_module_no_file (test_importer.TestImporter) ... ERROR
test module test_inspector in /a/ports/obj/py-nose-0.9.3/nose-0.9.3/unit_tests 
... test class  ... Test that inspector 
can handle multi-line docstrings ... ok
test_get_tb_source_func (test_inspector.TestExpander) ... ok
test_get_tb_source_simple (test_inspector.TestExpander) ... ok
test_inspect_traceback_continued (test_inspector.TestExpander) ... ok
test_pick_tb_lines (test_inspector.TestExpander) ... ok
test_simple_inspect_frame (test_inspector.TestExpander) ... ok
test module test_isolation_plugin in 
/a/ports/obj/py-nose-0.9.3/nose-0.9.3/unit_tests ... 
test_isolation_plugin.test_lint ... ok
test module test_issue_006 in /a/ports/obj/py-nose-0.9.3/nose-0.9.3/uni

Re: NEW: devel/py-nose

[Niall O'Higgins]

On Fri, Apr 27, 2007 at 04:56:24PM -0500, Will Maier wrote:
> On Fri, Apr 27, 2007 at 10:33:39PM +0100, Niall O'Higgins wrote:
> > Port for 'nose' python unittest framework.  Required for regress of
> > py-simplejson (which I will post next).  Tested on amd64. 
> 
> I posted a version[0] of this for 0.9.1 on 4 January:
> 
> http://marc.info/?l=openbsd-ports&m=116791585912447&w=2

Ah, sorry.  I tend to just port this stuff when I need to install it,
since its so simple, and then send it off to ports@ to share/archive. 

> I like my version better, though the differences are mostly
> stylistic. ;) Some suggestions:
> 
> * shorten COMMENT (I used "discovery-based unittests for
>   Python")
> * V=NNN variables seem unnecessary
> * ${DISTNAME:S/^/py-/} seems more idiomatic

I have no real preference here, just from another Python port.  Yours is
shorter so I'll use that. 

> * why add www to CATEGORIES?

Good point.  I missed that, replaced it with devel/.

Port updated @ http://unworkable.org/~niallo/py-nose.tar.gz




> 
> Otherwise, looks good to me -- I'd like this in the tree no matter
> who maintains it. Your pkg/DESCR is much better than mine, too. ;)
> 
> [0] http://www.lfod.us/files/ports/py-nose.tgz
> 
> -- 
> 
> o--{ Will Maier }--o
> | web:...http://www.lfod.us/ | [EMAIL PROTECTED] |
> *--[ BSD Unix: Live Free or Die ]--*

Re: NEW: devel/py-simplejson

[Niall O'Higgins]

On Fri, Apr 27, 2007 at 10:35:24PM +0100, Niall O'Higgins wrote:
> Hi,
> 
> Port of Python 'simplejson' library.  I am using this primarily to
> access www.freebase.com database from OpenBSD, but its useful for many
> other things.  Regress target requires devel/py-nose port I just sent.
> 
> Tested on amd64.
> 
> Tests?  Comments?  OKs?
> 
> DESCR:
> simplejson is a simple, fast, complete, correct and extensible JSON
>  encoder and decoder for Python 2.3+. It is pure Python
> code with no dependencies, but includes an optional C extension for a
> serious speed boost.
> 
> Port attached and also available @
> http://unworkable.org/~niallo/py-simplejson.tar.gz

Port updated at above link.  Fixed category and made Makefile a little
nicer. 

Re: NEW: devel/py-nose

[Will Maier]

On Fri, Apr 27, 2007 at 10:33:39PM +0100, Niall O'Higgins wrote:
> Port for 'nose' python unittest framework.  Required for regress of
> py-simplejson (which I will post next).  Tested on amd64. 

I posted a version[0] of this for 0.9.1 on 4 January:

http://marc.info/?l=openbsd-ports&m=116791585912447&w=2

I like my version better, though the differences are mostly
stylistic. ;) Some suggestions:

* shorten COMMENT (I used "discovery-based unittests for
  Python")
* V=NNN variables seem unnecessary
* ${DISTNAME:S/^/py-/} seems more idiomatic
* why add www to CATEGORIES?

Otherwise, looks good to me -- I'd like this in the tree no matter
who maintains it. Your pkg/DESCR is much better than mine, too. ;)

[0] http://www.lfod.us/files/ports/py-nose.tgz

-- 

o--{ Will Maier }--o
| web:...http://www.lfod.us/ | [EMAIL PROTECTED] |
*--[ BSD Unix: Live Free or Die ]--*

NEW: devel/py-simplejson

[Niall O'Higgins]

Hi,

Port of Python 'simplejson' library.  I am using this primarily to
access www.freebase.com database from OpenBSD, but its useful for many
other things.  Regress target requires devel/py-nose port I just sent.

Tested on amd64.

Tests?  Comments?  OKs?

DESCR:
simplejson is a simple, fast, complete, correct and extensible JSON
 encoder and decoder for Python 2.3+. It is pure Python
code with no dependencies, but includes an optional C extension for a
serious speed boost.

Port attached and also available @
http://unworkable.org/~niallo/py-simplejson.tar.gz


py-simplejson.tar.gz
Description: application/tar-gz

NEW: devel/py-nose

[Niall O'Higgins]

Hi,

Port for 'nose' python unittest framework.  Required for regress of
py-simplejson (which I will post next).  Tested on amd64. 

Tests?  Comments?  OKS?

DESCR:

nose extends the test loading and running features of unittest, making
it easier to write, find and run tests.

By default, nose will run tests in files or directories under the
current working directory whose names include "test" or "Test" at a word
boundary (like "test_this" or "functional_test" or "TestClass" but not
"libtest"). Test output is similar to that of unittest, but also
includes captured stdout output from failing tests, for easy print-style
debugging.

These features, and many more, are customizable through the use of
plugins. Plugins included with nose provide support for doctest, code
coverage and profiling, and flexible attribute-based test selection.

Port attached and also available @
http://unworkable.org/~niallo/py-nose.tar.gz


py-nose.tar.gz
Description: application/tar-gz

Re: Where to download courier-authlib

[Marc Balmer]

* satimis wrote:
> 
> Hi folks,
> 
> 
> OpenBSD 4.0 amd64
> 
> 
> I need to install courier-authlib and courier-imap from source

why that?  is the port not ok or lacking features you need?

> 
> On;
> ftp://mirror.pacific.net.au/OpenBSD/4.0/packages/amd64/
> 
> only found.
> File: courier-imap-3.0.5p4.tgz378 KB  09/24/0600:00:00
> File: courier-ldap-3.0.5p1.tgz41 KB   09/24/0600:00:00
> File: courier-mysql-3.0.5p1.tgz   44 KB   09/24/0600:00:00
> File: courier-pgsql-3.0.5p1.tgz   38 KB   09/24/0600:00:00
> File: courier-pop3-3.0.5p1.tgz43 KB   09/24/0600:00:00
> File: courier-utils-1.7.0p2.tgz
> 
> courier-authlib NOT available.  Where can I download it.  Or has its name
> been changes?
> 
> TIA
> 
> 
> B.R.
> Stephen Liu
> -- 
> View this message in context: 
> http://www.nabble.com/Where-to-download-courier-authlib-tf3658643.html#a10222415
> Sent from the openbsd user - ports mailing list archive at Nabble.com.
> 

Where to download courier-authlib

[satimis]

Hi folks,


OpenBSD 4.0 amd64


I need to install courier-authlib and courier-imap from source

On;
ftp://mirror.pacific.net.au/OpenBSD/4.0/packages/amd64/

only found.
File: courier-imap-3.0.5p4.tgz  378 KB  09/24/0600:00:00
File: courier-ldap-3.0.5p1.tgz  41 KB   09/24/0600:00:00
File: courier-mysql-3.0.5p1.tgz 44 KB   09/24/0600:00:00
File: courier-pgsql-3.0.5p1.tgz 38 KB   09/24/0600:00:00
File: courier-pop3-3.0.5p1.tgz  43 KB   09/24/0600:00:00
File: courier-utils-1.7.0p2.tgz

courier-authlib NOT available.  Where can I download it.  Or has its name
been changes?

TIA


B.R.
Stephen Liu
-- 
View this message in context: 
http://www.nabble.com/Where-to-download-courier-authlib-tf3658643.html#a10222415
Sent from the openbsd user - ports mailing list archive at Nabble.com.

NEW: net/netrate

[Stuart Henderson]

Here's a port of the netrate tools from FreeBSD's src tree;
including UDP packet generators (netsend and netblast) and sink
(netreceive), useful for testing forwarding performance at
high pps rates and seeing what breaks down.

As far as I'm aware it's only distributed in-tree, so
I packaged the code into a tar.gz. Works for me on i386,
amd64, sparc64.

If you are testing this, in particular netblast, expect it
to cause high network load. Should go without saying, but I'll
say it anyway, don't blindly run it on a production network!

The port is attached and at:
http://spacehopper.org/openbsd/netrate.tar.gz



netrate.tar.gz
Description: application/tar-gz

NEW: audio/last.fm

[Jacob Meuser]

Here's a port of the official last.fm client.

pkg/DESCR:
With Last.fm on your computer you can scrobble your tracks, share
your music taste, listen to personalised radio streams, and discover
new music and people.

There are still a couple issues:

1) clicking on the hyperlinks in the artists dscriptions opens the
   links in firefox but hangs last.fm.  not sure what's happening
   to cause that.

2) sometimes dumps core on exit.  something is getting free()d that
   shouldn't, but I can't find where it is.

tested on amd64.

-- 
[EMAIL PROTECTED]
SDF Public Access UNIX System - http://sdf.lonestar.org


last.fm-1.1.3-port.tgz
Description: application/tar-gz

Re: UPDATE: net/scapy to version 1.1.1

[Claudio Jeker]

On Thu, Apr 26, 2007 at 06:16:35PM +0200, Claudio Jeker wrote:
> This diff brings the scapy port to the newest version.
> This is needed to make scapy6 work -- I'll create a port for that too.
> 
> I had to fix a few issues -- grrr grumble. Lesson to learn: never try to
> parse netstat output assuming that you can simply split it to seven
> fields! Locked MTU entries add a 8 field and interfaces with the name 'L'
> will not work on OpenBSD. Took me about 1-2 hours of head scratching to
> understand why scapy suddenly crashed on startup.
> 
> Before aplying the diff you need to "mkdir patches"

This is an updated version that fixes a obvious programming error in the
PcapWriter module.

-- 
:wq Claudio

Index: Makefile
===
RCS file: /cvs/ports/net/scapy/Makefile,v
retrieving revision 1.2
diff -u -p -r1.2 Makefile
--- Makefile3 Jun 2006 09:43:03 -   1.2
+++ Makefile25 Apr 2007 22:16:34 -
@@ -2,8 +2,7 @@
 
 COMMENT=   "powerful interactive packet manipulation in python"
 
-DISTNAME=  scapy-1.0.4
-PKGNAME=   ${DISTNAME}p0
+DISTNAME=  scapy-1.1.1
 CATEGORIES=net
 
 HOMEPAGE=  http://secdev.org/projects/scapy/
@@ -30,7 +29,6 @@ PKG_ARCH= *
 NO_REGRESS=Yes
 
 do-build:
-   gunzip < ${WRKSRC}/scapy.1.gz > ${WRKBUILD}/scapy.1
perl -p -i -e "s,^(\#\S+ \S+ )python\$$,#!${MODPY_BIN}," \
${WRKSRC}/*.py
perl -p -i -e "s,/etc/ethertypes,${SYSCONFDIR}/ethertypes,g" \
Index: distinfo
===
RCS file: /cvs/ports/net/scapy/distinfo,v
retrieving revision 1.2
diff -u -p -r1.2 distinfo
--- distinfo5 Apr 2007 16:20:15 -   1.2
+++ distinfo25 Apr 2007 22:15:40 -
@@ -1,10 +1,10 @@
 MD5 (ethertypes) = lL/96PdaG4+JH7eAv+Fcog==
-MD5 (scapy-1.0.4.tar.gz) = wVZmjPq0wf3idvgeC4W12Q==
+MD5 (scapy-1.1.1.tar.gz) = iscgob6kMEeXxxPvHtBj8Q==
 RMD160 (ethertypes) = KWuI/tkRslOYtfZXzWu1yR5y4og=
-RMD160 (scapy-1.0.4.tar.gz) = JhhUgEL2Ww0CRhYj/s6T1MJYiVk=
+RMD160 (scapy-1.1.1.tar.gz) = IL2x6lmgX0UqUV5DjiMOHYW4EBY=
 SHA1 (ethertypes) = btD+JEZmaHbXe4lx/Y7fSVqcC+M=
-SHA1 (scapy-1.0.4.tar.gz) = q+fMJCr08g7wFlV4AOLlIcLhShs=
+SHA1 (scapy-1.1.1.tar.gz) = hw2o5uKoeGsDoAVfwswdExfzCOQ=
 SHA256 (ethertypes) = 4UAHHkYt+Kq597Y6OWW/IGeX7Lz/M0o+xJQlbe/jXcs=
-SHA256 (scapy-1.0.4.tar.gz) = NAvrsmu/jLN728+3uN+SJa04Sgs/CzO1XKC4umyaOwI=
+SHA256 (scapy-1.1.1.tar.gz) = KhG6BfNLKXivOt1iOmQeA0DTkyK0yivSa6rnQ8eN6VY=
 SIZE (ethertypes) = 1317
-SIZE (scapy-1.0.4.tar.gz) = 132677
+SIZE (scapy-1.1.1.tar.gz) = 147401
--- /dev/null   Fri Apr 27 11:24:24 2007
+++ patches/patch-scapy_py  Thu Apr 26 23:15:05 2007
@@ -0,0 +1,38 @@
+$OpenBSD$
+--- scapy.py.orig  Mon Apr  9 15:17:35 2007
 scapy.py   Thu Apr 26 23:14:39 2007
+@@ -1152,7 +1152,7 @@ if not LINUX:
+ dest,mask,gw,netif,mxfrg,rtt,ref,flg = l.split()[:8]
+ else:
+ if mtu_present:
+-dest,gw,flg,ref,use,mtu,netif = l.split()[:7]
++dest,gw,flg,ref,use,mtu,netif = l.split(None, 6)[:7]
+ else:
+ dest,gw,flg,ref,use,netif = l.split()[:6]
+ if flg.find("Lc") >= 0:
+@@ -1172,6 +1172,8 @@ if not LINUX:
+ dest, = struct.unpack("I",inet_aton(dest))
+ if not "G" in flg:
+ gw = '0.0.0.0'
++if netif.find("L") >= 0:
++  dummy,netif = netif.split()
+ ifaddr = get_if_addr(netif)
+ routes.append((dest,netmask,gw,netif,ifaddr))
+ f.close()
+@@ -9847,7 +9849,8 @@ class PcapWriter:
+ """
+ 
+ if self.header_done == 0:
+-if self.linktype == None:
++linktype = self.linktype
++if linktype == None:
+ if isinstance(pkt,Packet):
+ linktype = LLNumTypes.get(pkt.__class__,1)
+ else:
+@@ -12190,4 +12193,5 @@ def read_config_file(configfile):
+ if __name__ == "__main__":
+ interact()
+ else:
+-read_config_file(DEFAULT_CONFIG_FILE)
++if DEFAULT_CONFIG_FILE:
++read_config_file(DEFAULT_CONFIG_FILE)

Re: Embedding MODPY_VERSION in PKGNAME

[Damien Miller]

On Fri, 27 Apr 2007, Joerg Zinke wrote:

> hi,
> 
> on-topic: having MODPY_VERSION in PKGNAME would be very nice and
> very useful.
> 
> On Fri, 27 Apr 2007 14:55:20 +1000 (EST)
> Damien Miller <[EMAIL PROTECTED]> wrote:
> 
> > On Fri, 27 Apr 2007, Lars Hansson wrote:
> > 
> > > Why are there different versions of python in the tree in the first
> > > place?
> > 
> > - Because major revisions of Python are not guaranteed to be 100%
> > backwards compatible
> 
> off-topic: ^^^ do you have source that refers to this statement? 

The note at the start of http://docs.python.org/lib/module-exceptions.html
is one example.

I believe the general practice for handling doomed language features and
APIs is to raise DeprecationWarning for one major release and then remove
them in the next major release.

-d

Re: Embedding MODPY_VERSION in PKGNAME

[Joerg Zinke]

hi,

on-topic: having MODPY_VERSION in PKGNAME would be very nice and
very useful.

On Fri, 27 Apr 2007 14:55:20 +1000 (EST)
Damien Miller <[EMAIL PROTECTED]> wrote:

> On Fri, 27 Apr 2007, Lars Hansson wrote:
> 
> > Why are there different versions of python in the tree in the first
> > place?
> 
> - Because major revisions of Python are not guaranteed to be 100%
> backwards compatible

off-topic: ^^^ do you have source that refers to this statement? 

regards,

joerg

Re: Embedding MODPY_VERSION in PKGNAME

[Marc Espie]

On Fri, Apr 27, 2007 at 03:12:57PM +1000, Damien Miller wrote:
> On Fri, 27 Apr 2007, Mathieu Sauve-Frankel wrote:
> 
> > > Most Python modules install under /usr/local/lib/python${MODPY_VERSION}
> > > anyway, so they are fully capable of existing in parallel.
> > 
> > But this won't be true of anything that installs anything outside 
> > of /usr/local/lib/python${MODPY_VERSION}. If the python stuff is
> > anything like the perl stuff then there are probably quite a few py-*
> > packages that do this. I believe this is what the concern is about. 
> 
> Sure, and there are a couple of simple solutions:
> 
> 1) Don't include MODPY_VERSION in the PKGNAME of these packages
> 
> 2) Include MODPY_VERSION but let conflicting files conflict
> 
> 3) Include MODPY_VERSION and suffix conflicting files with MODPY_VERSION
>too (e.g. /usr/local/bin/foobinator-2.4) like we do with python itself
>right now
> 
> Personally, I'd lean to #3.
> 
> There isn't that much stuff that conflicts like this:
> 
> cat /usr/ports/*/py-*/pkg/P[LF]* | grep -v '' | grep -v '^share/doc/py-' | 
> grep -v '^share/examples/py-' | grep -v '[EMAIL PROTECTED]'
> 
> will give you an idea of how many files conflict.
> 
> -d
I think I'd like to see it as an option, e.g., MODPY_AUTOVERSIONING=Yes or
some such, so that you can turn it on explicitly and check for breakage.
As you said, there are bugs. So I expect this makes little sense for some
modules...

There's the fact that it's going to `change' all python modules names, which
plays hell with updates.

So, I think the `default' python version should have no version markers,
and only non default python versions should have markers. That way, people
who don't play with python versions won't see a thing, and when we change
the `default', pkg_add -u will do the right thing, e.g., bump packages to
the new version.