headers
Kevin Horton | 1 Sep 2004 01:25
Favicon

Re: system-xfree86 doesn't provide x11

At 15:25 -0400 31/8/04, Alexander K. Hansen wrote:
>On Aug 31, 2004, at 2:53 PM, Peter Taylor wrote:
>
>># COLUMNS=76 fink list x11\* system-xfree86\*
>>Information about 1833 packages read in 1 seconds.
>>  i   system-xfree86  2:4.3-2     [placeholder for user installed x11]
>>  i   system-xfre...  2:4.3-2     [placeholder for user installed x11 dev...
>>  i   system-xfre...  2:4.3-2     [placeholder for user installed x11 sha...
>>      x11                         [virtual package]
>>      x11-dev                     [virtual package]
>>      x11-shlibs                  [virtual package]
>>
>x11, x11-dev and x11-shlibs are in your "fink list", so they are 
>being provided.  Run "fink-virtual-pkgs" if you don't believe me.
>
>Since x11 is a provided virtual package rather than one that is 
>generated directly by fink, it doesn't have a version number or show 
>up with an i flag if it exists.  If it weren't being provided, "fink 
>list" wouldn't show it.
>

This design certainly is confusing to the user though.  I have been 
confused in the past by the fact that x11 didn't have an "i" beside 
it, but I didn't complain as everything seemed to be working 
correctly.  Shouldn't it be possible to code a special case so that 
virtual packages look like they are installed?

Kevin Horton

-------------------------------------------------------
(Continue reading)

Permalink | Reply | 1 comment |
headers
Darian Lanx | 1 Sep 2004 02:34
Picon
Gravatar

Re: system-xfree86 doesn't provide x11

Kevin Horton wrote:
<snip>
>>
> 
> This design certainly is confusing to the user though.  I have been 
> confused in the past by the fact that x11 didn't have an "i" beside it, 
> but I didn't complain as everything seemed to be working correctly.  
> Shouldn't it be possible to code a special case so that virtual packages 
> look like they are installed?
> 
Well, virtual packages are not "installed". However it might be smart to 
provide a "p" instead of "i" which would stand for "provided"

-d

-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
Permalink | Reply | 0 comments |
headers
Andrea Riciputi | 1 Sep 2004 11:36

Re: Building and installation directory.

I's exactly what I'm looking for! But I've not found any mention in 
Python docs about how I can tell distutils to pass that argument to 
compile_dir(). Any hints?

Thanks,
	Andrea.

On 30 Aug 2004, at 20:57, Daniel Macks wrote:

> That's caused by using compileall.compile_dir() with only one
> argument. See /sw/lib/python2.3/compileall.py for details.
>
> dan

-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
Permalink | Reply | 4 comments |
headers
Daniel E. Macks | 1 Sep 2004 14:14

Re: Building and installation directory.

I wonder if, in general, 'python setup.py install --prefix=%i'
is the wrong idiom for the case where one wants to install in a
temp location and then copy to the actual runtime location later.

Looking at the output of 'python2.2 setup.py --help install', I see:
  --prefix            installation prefix
  --root              install everything relative to this alternate root
                      directory

I just tried 'python setup.py install --root=%d' and wound up with
files in %d%p, and the messages embedded in those .pyc contain the
string %p, not %d%p (where before with --prefix=%i, they contained
%i).

dan

Andrea Riciputi <ariciputi <at> pito.com> said:
> I's exactly what I'm looking for! But I've not found any mention in 
> Python docs about how I can tell distutils to pass that argument to 
> compile_dir(). Any hints?
> 
> On 30 Aug 2004, at 20:57, Daniel Macks wrote:
> 
>> That's caused by using compileall.compile_dir() with only one
>> argument. See /sw/lib/python2.3/compileall.py for details.

--

-- 
Daniel Macks
dmacks <at> netspace.org
http://www.netspace.org/~dmacks
(Continue reading)

Permalink | Reply | 3 comments |
headers
Andrea Riciputi | 1 Sep 2004 15:28

Re: Re: Building and installation directory.

I've tried --root=%d and I've got the same result for my packages. But 
before changing all of them I'd prefer to know the opinion of someone 
more experienced in Python...

Any comment?

	Andrea.

On 1 Sep 2004, at 14:14, Daniel E. Macks wrote:

> I wonder if, in general, 'python setup.py install --prefix=%i'
> is the wrong idiom for the case where one wants to install in a
> temp location and then copy to the actual runtime location later.
>
> Looking at the output of 'python2.2 setup.py --help install', I see:
>   --prefix            installation prefix
>   --root              install everything relative to this alternate 
> root
>                       directory
>
> I just tried 'python setup.py install --root=%d' and wound up with
> files in %d%p, and the messages embedded in those .pyc contain the
> string %p, not %d%p (where before with --prefix=%i, they contained
> %i).
>
> dan

-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
(Continue reading)

Permalink | Reply | 2 comments |
headers
Martin Costabel | 1 Sep 2004 16:57
Picon

Re: Re: Building and installation directory.

Andrea Riciputi wrote:

> I've tried --root=%d and I've got the same result for my packages. But 
> before changing all of them I'd prefer to know the opinion of someone 
> more experienced in Python...

Well, I'm not experienced, but I know how to run 'grep' :-)
It seems there are already some packages that use --root=%d (two of 
mine, for example.)

OTOH Jeff, who *is* experienced, said it doesn't really matter. I think 
the worst thing that can happen is that execution slows down, because 
the precompiled scripts are not used.

--

-- 
Martin

-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
Permalink | Reply | 1 comment |
headers
linux-lists | 1 Sep 2004 17:26
Picon

X11 Cut'n Paste

Hi,

I've XFree86 installed on OSX 10.3.5
Is there any way to get Cut'n `Paste between XFree and Aqua working, or
do I need to take Apples X11? Is it possible to run e.g. sylpheed, gimp2
or freeciv from fink with Apple X11?

regards,
alex

-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
Permalink | Reply | 1 comment |
headers
Alexander K. Hansen | 1 Sep 2004 17:53
Picon

Re: X11 Cut'n Paste


On Sep 1, 2004, at 11:26 AM, linux-lists <at> web.de wrote:

> Hi,
>
> I've XFree86 installed on OSX 10.3.5
> Is there any way to get Cut'n `Paste between XFree and Aqua working

You can use the autocutsel package.

> , or
> do I need to take Apples X11?

Only if you want to.

> Is it possible to run e.g. sylpheed, gimp2
> or freeciv from fink with Apple X11?
>

Yes.

> regards,
> alex
>
>
>
>
--
Alexander Hansen
Fink Documentarian
(Continue reading)

Permalink | Reply | 0 comments |
headers
Timothy Reaves | 2 Sep 2004 01:30

ethereal

	Ethereal is in the list of packages over at the sf site; why does fink 
not see it when I run fink install?

-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
Permalink | Reply | 1 comment |
headers
Andrea Riciputi | 1 Sep 2004 17:19

Re: Re: Building and installation directory.

I see, but you must admit that having traceback messages that report 
directories that don't exist it's not very nice. So if there is not any 
real reason to avoid the --root flag, I think we have to prefer it to 
--prefix as suggested by Daniel Macks.

Cheers,
	Andrea.

On 1 Sep 2004, at 16:57, Martin Costabel wrote:

> Andrea Riciputi wrote:
>
>> I've tried --root=%d and I've got the same result for my packages. 
>> But before changing all of them I'd prefer to know the opinion of 
>> someone more experienced in Python...
>
> Well, I'm not experienced, but I know how to run 'grep' :-)
> It seems there are already some packages that use --root=%d (two of 
> mine, for example.)
>
> OTOH Jeff, who *is* experienced, said it doesn't really matter. I 
> think the worst thing that can happen is that execution slows down, 
> because the precompiled scripts are not used.
>
> -- 
> Martin
>

-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
(Continue reading)

Permalink | Reply | 0 comments |

Gmane