Re: [gentoo-user] no ebuilds to satisfy =dev-libs/grantlee-0.1

2011-04-24 Thread Alan McKinnon
Apparently, though unproven, at 06:32 on Sunday 24 April 2011, Dale did opine 
thusly:

 Well, I synced and got this interesting message:
 
 root@fireball / # emerge -uvDNp world
 
 These are the packages that would be merged, in order:
 
 Calculating dependencies... done!
 
 emerge: there are no ebuilds to satisfy =dev-libs/grantlee-0.1.
 (dependency required by kde-base/kjots-4.5.95 [ebuild])
 (dependency required by kde-base/kdepim-meta-4.5.95 [ebuild])
 (dependency required by kde-base/kde-meta-4.6.2[semantic-desktop]
 [installed])
 (dependency required by @selected [set])
 (dependency required by @world [argument])
 root@fireball / #
 
 So, it is not masked or keyworded, it just plain ain't there.  Since
 this is unstable, should I report it as a bug or wait on the Raid to
 show up on it's own?

so is it kjots from the kde overlays that's doing it?

-- 
alan dot mckinnon at gmail dot com



Re: [gentoo-user] no ebuilds to satisfy =dev-libs/grantlee-0.1

2011-04-24 Thread Dale

Alan McKinnon wrote:

Apparently, though unproven, at 06:32 on Sunday 24 April 2011, Dale did opine
thusly:

   

Well, I synced and got this interesting message:

root@fireball / # emerge -uvDNp world

These are the packages that would be merged, in order:

Calculating dependencies... done!

emerge: there are no ebuilds to satisfy =dev-libs/grantlee-0.1.
(dependency required by kde-base/kjots-4.5.95 [ebuild])
(dependency required by kde-base/kdepim-meta-4.5.95 [ebuild])
(dependency required by kde-base/kde-meta-4.6.2[semantic-desktop]
[installed])
(dependency required by @selected [set])
(dependency required by @world [argument])
root@fireball / #

So, it is not masked or keyworded, it just plain ain't there.  Since
this is unstable, should I report it as a bug or wait on the Raid to
show up on it's own?
 

so is it kjots from the kde overlays that's doing it?

   


I don't have any overlays on here that I know of, unless portage did 
something fancy that I'm not aware of.  It appears that something got 
moved into the tree but maybe one ebuild got missed.  That's the reason 
for my post and the question.  Am I correct or is there something fishy 
on my system?  I did wait a few hours and sync again, just in case I got 
it in the middle of the change or something.  This is the correct 
command for listing local overlays correct?


root@fireball / # layman -l
root@fireball / #

If that is correct, I shouldn't have any overlays.  I don't want to file 
a bug and it be me.  red face   If the tree has a boo boo, I'd like to 
report it so it can get fixed otherwise someone else will run into this 
little error.


Dale

:-)  :-)



Re: [gentoo-user] Building pygtk-2.22.0-r1 fails. Help, please!

2011-04-24 Thread Mick
On Saturday 23 April 2011 21:06:25 Alan Mackenzie wrote:
 On Sat, Apr 23, 2011 at 08:46:30PM +0100, Mick wrote:

  What do you get when you run:
  
  # eselect python list
 
 Available Python interpreters:
   [1]   python2.6 *
   [2]   python2.7
   [3]   python3.1

OK, the next stage would be to change your python to the latest stable:

eselect python set 2

and then remerge those packages that were linked against the old python:

python-updater -v -p

to get a list of these.

When you finish all this you can run:

emerge --depclean -v -p

It should now ask you to remove the old python, but check carefully the 
remaining packages in case something important is in the list and breaks your 
system.

-- 
Regards,
Mick


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-user] no ebuilds to satisfy =dev-libs/grantlee-0.1

2011-04-24 Thread Alan McKinnon
Apparently, though unproven, at 09:30 on Sunday 24 April 2011, Dale did opine 
thusly:

 Alan McKinnon wrote:
  Apparently, though unproven, at 06:32 on Sunday 24 April 2011, Dale did
  opine
  
  thusly:
  Well, I synced and got this interesting message:
  
  root@fireball / # emerge -uvDNp world
  
  These are the packages that would be merged, in order:
  
  Calculating dependencies... done!
  
  emerge: there are no ebuilds to satisfy =dev-libs/grantlee-0.1.
  (dependency required by kde-base/kjots-4.5.95 [ebuild])
  (dependency required by kde-base/kdepim-meta-4.5.95 [ebuild])
  (dependency required by kde-base/kde-meta-4.6.2[semantic-desktop]
  [installed])
  (dependency required by @selected [set])
  (dependency required by @world [argument])
  root@fireball / #
  
  So, it is not masked or keyworded, it just plain ain't there.  Since
  this is unstable, should I report it as a bug or wait on the Raid to
  show up on it's own?
  
  so is it kjots from the kde overlays that's doing it?
 
 I don't have any overlays on here that I know of, unless portage did
 something fancy that I'm not aware of.  It appears that something got
 moved into the tree but maybe one ebuild got missed.  That's the reason
 for my post and the question.  Am I correct or is there something fishy
 on my system?  I did wait a few hours and sync again, just in case I got
 it in the middle of the change or something.  This is the correct
 command for listing local overlays correct?

I just synced myself, and it's not an overlay issue.

kdepim-4.5.95 just got moved into the tree (masked) and you likely have one or 
more of those package unmasked. Or a mask is missing in the tree.

Either way, find the list in $PORTDIR/profiles/package.mask and comparing with 
what update world wants to do, you will see what needs to be done next.

kdepim-4.5.95 is not the version you want, latest useable is 4.4.11 (it's out 
of step with version numbers of everything else in kde)



-- 
alan dot mckinnon at gmail dot com



Re: [gentoo-user] Gentoo on Cygwin

2011-04-24 Thread Nils Andresen
Alex,
have you re-tried lately?
I fail during emerging baselayout-prefix. The error is
 start-stop-daemon.c:64:4: error: #error Unknown architecture - cannot build 
 start-stop-daemon

Any ideas?

Nils

2011/4/19 Alex Schuster wo...@wonkology.org:
 He was somewhat successful in emerging @system, then he wrote the wiki
 article :)

        Wonko





Re: [gentoo-user] zlib and WOFF

2011-04-24 Thread luis jure

on 2011-04-24 at 14:31 Adam Carter wrote:

 /usr/lib64/libz.so: invalid ELF header
 Could not find the zlib library which is needed to understand WOFF


 I'd re-emerge zlib - looks like its corrupted.


as i said in my previous post, i can compile lilypond with zlib-1.2.3-r1,
but it fails with 1.2.5-r2 (i also tried the dead zlib-1.2.4). obviously
i unmerged and re-emerged different versions of zlib several times, always
with the same result. that's why i'm asking on the list. according to
google, no-one is having this problem... 


best,

lj



Re: [gentoo-user] zlib and WOFF

2011-04-24 Thread Volker Armin Hemmann
On Sunday 24 April 2011 01:07:57 luis jure wrote:
 i've been trying to install a recent version of media-sound/lilypond (music
 typesetting software) for some time, but it always fails with this message:
 
 /usr/lib64/libz.so: invalid ELF header
 Could not find the zlib library which is needed to understand WOFF
 
 i had this same problem some time ago, and after much trial and error i
 managed to solve it, downgrading zlib to an older version (zlib-1.2.3-r1).
 but having such an old version of zlib is inconvenient for other reasons.
 
 any idea what's wrong here? who's the culprit? zlib? lilypond? me?
 
 best,
 
 
 lj

http://bugs.gentoo.org/show_bug.cgi?id=311147

you might add yourself to that bug mentioning lillypond.

Another case of devs not thinking big enough.



Re: [gentoo-user] zlib and WOFF

2011-04-24 Thread Stroller

On 24/4/2011, at 5:07am, luis jure wrote:
 ...
 any idea what's wrong here? who's the culprit? zlib? lilypond? me?

I looked at Lilypond myself a week or two ago, and installed their binary 
package under OS X.

Whilst glancing at their documentation again now the wording I'm finding is not 
so strong, I had the impression that Lilypond was a bit finicky and difficult 
to compile. Again, I can't immediately find the reference, but I thought that 
it was sufficiently demanding in terms of environment and versioning that they 
provided a VM for development purposes.

I would guess that the fault is simply versioning. I would suggest filing a 
bug, but also trying to resolve this though the Lilipond IRC channel /or 
mailing list - Lilipond's developers may well have seen this before.

Stroller.




[gentoo-user] WPA Supplicant

2011-04-24 Thread dhkuhl
I'm trying to get wireless working reliably on my laptop.  I have followed the 
documentation, but still have a lot of questions . . . especially since it only 
seems to work sometimes.
 
First, I'm using WPA Supplicant without the gui tools.
 
Second, I'm in NYC and there are a lot of networks I can use around town.
 
Third, Is there a way I can control the services I use from the Grub menu?  
Since the laptop has a wireless card and an RJ45 jack, I'd like to be able boot 
and not use one or the other.  Since I know if I'm not physically connected to 
a network, there's really no reason to even try eth0.
 
Forth, The problem.  I'm not sure how wpa_supplicant works or how it should 
work.  The wpa_supplicant man page gives a few examples on how to run it, but 
when I look at the process list it seems to be run by another program called 
wpa_cli.  There's also a shell script in /etc/wpa/supplicant that looks like it 
can start or stop it with CONNECT or DISCONNECT.
1) Do I need to enter networks in wpa_supplicant.conf or does wpa_supplicant 
scan for networks and connect to whatever's available?
2) If I have multiple networks available how does wpa_supplicant choose which 
to connect to and can I specify which one I want?
3) How should wpa_supplicant be started, stopped and restarted?  What should be 
used for this:  wpa_supplicant, wpa_cli, or wpa_cli.sh?  I don't see anything 
in /etc/init.d for that, but it looks like netmount may be doing it.
4) The documentation doesn't say to, but the way I got wireless working is by 
creating a link net.wlan0 - net.lo in the /etc/init.d directory.  Is this 
correct?  I think that's why it's starting automatically when I boot too, 
because I never added it with rc-update so netmount must be picking it up.
5) This is the most puzzling thing.  When wpa_supplicant starts even though I 
get a inet address I can't always get to the internet.  Why does the panel 
applet says I'm connected and ifconfig shows an inet address but firefox and 
ping can't reach a site like yahoo or google?
6) For networks where I have a password, should that go in wpa_supplicant.conf 
as plain text or should it be encrypted?
 
I think I have more questions, but this is good for starters.
 
Thanks,
 
dhk


Re: [gentoo-user] ssmtp rewriting From: in emails...

2011-04-24 Thread Indi
On Fri, Apr 22, 2011 at 09:00:02PM +0200, Jarry wrote:
 
 btw, I find /etc/ssmtp/ssmtp.conf a little confusing:
 
 # Set this to never rewrite the From: line (unless not given) and to
 # use that address in the from line of the envelope.
 #FromLineOverride=YES
 
 I always thought if a value in config file is commented out,
 it is a default value. So I never tried to uncomment it...


I always thought that too, until just the other day when it 
happened that simply uncommenting a line in privoxy's config 
fixed a problem for me. I always thought the commented out 
lines were set to default values before that. Pretty sure 
*most* are, but apparently not all. Who knew?

-- 
 /\   /\ 
   \   /
  ^  caveat utilitor 
'v-v'




Re: [gentoo-user] WPA Supplicant

2011-04-24 Thread Mick
On Sunday 24 April 2011 13:37:03 dhk...@optonline.net wrote:
 I'm trying to get wireless working reliably on my laptop.  I have followed
 the documentation, but still have a lot of questions . . . especially
 since it only seems to work sometimes.
 
 First, I'm using WPA Supplicant without the gui tools.

run wpa_gui from a terminal and a lot of what you're asking below will become 
self-explanatory.


 Second, I'm in NYC and there are a lot of networks I can use around town.
 
 Third, Is there a way I can control the services I use from the Grub menu? 
 Since the laptop has a wireless card and an RJ45 jack, I'd like to be able
 boot and not use one or the other.  Since I know if I'm not physically
 connected to a network, there's really no reason to even try eth0.

Check /etc/conf.d/rc and in particular:

# RC_NET_STRICT_CHECKING allows some flexibility with the 'net' service.
# The following values are allowed:
#  none  - The 'net' service is always considered up.
#  no- This basically means that at least one net.* service besides net.lo
#  must be up.  This can be used by notebook users that have a wifi 
and
#  a static nic, and only wants one up at any given time to have the
#  'net' service seen as up.
#  lo- This is the same as the 'no' option, but net.lo is also counted.
#  This should be useful to people that do not care about any specific
#  interface being up at boot.
#  yes   - For this ALL network interfaces MUST be up for the 'net' service to
#  be considered up.

RC_NET_STRICT_CHECKING=no

(or you can use lo)


 Forth, The problem.  I'm not sure how wpa_supplicant works or how it should
 work.  The wpa_supplicant man page gives a few examples on how to run it,
 but when I look at the process list it seems to be run by another program
 called wpa_cli.  There's also a shell script in /etc/wpa/supplicant that
 looks like it can start or stop it with CONNECT or DISCONNECT. 1) Do I
 need to enter networks in wpa_supplicant.conf or does wpa_supplicant scan
 for networks and connect to whatever's available? 

The latter.

You can however enter manually in /etc/wpa_supplicant/wpa_supplicant.conf 
particular parameters (keys and what not) of known networks to which you 
connect as a matter of preference.


 2) If I have multiple
 networks available how does wpa_supplicant choose which to connect to and
 can I specify which one I want? 

It'll connect to:

a) Any network you have specified in your 
/etc/wpa_supplicant/wpa_supplicant.conf according to the preference you have 
set up therein.

b) Any network it finds.

c) Any network you select with wpa_cli, or select/enable/disable in wpa_gui.


 3) How should wpa_supplicant be started,
 stopped and restarted?  What should be used for this:  wpa_supplicant,
 wpa_cli, or wpa_cli.sh?  I don't see anything in /etc/init.d for that, but
 it looks like netmount may be doing it. 

You need to define it in /etc/conf.d/net:

modules=( wpa_supplicant )
wpa_supplicant_wlan0=-Dwext  

(adjust this according to the name of your wireless iface and driver).


 4) The documentation doesn't say
 to, but the way I got wireless working is by creating a link net.wlan0 -
 net.lo in the /etc/init.d directory.  Is this correct?  

It depends which documentation you are looking at.  I am sure that this is 
explained in the gentoo Handbook and associated documentation.

This is the link you need:

lrwxrwxrwx  1 root root 6 Dec 16 14:26 net.wlan0 - net.lo

but you should have also configured /etc/conf.d/net with your desired settings 
or just defaults will run.


 I think that's why
 it's starting automatically when I boot too, because I never added it with
 rc-update so netmount must be picking it up. 


 5) This is the most puzzling
 thing.  When wpa_supplicant starts even though I get a inet address I
 can't always get to the internet.  Why does the panel applet says I'm
 connected and ifconfig shows an inet address but firefox and ping can't
 reach a site like yahoo or google? 

This could well be a dns server/repeater issue.

If you can ping the IP address of google, but not the domain name of it, then 
the problem is that you do not have access to a DNS repeater.  Look in your 
/etc/resolve.conf to see if there is a line saying:

nameserver XXX.XXX.XX.XX

if it is absent then you have not connected to a namesever.  This is a router 
issue and it could be controlled by some authentication scheme.  A lot of 
wireless services offered by coffee shops, libraries, etc. may give you an IP 
address automatically, but then require you use your browser to register with 
their authentication server (using a passwd that they provide after you pay 
them for the privilege).

Open access points with no encryption and no DNS authentication requirements 
should allow you to connect seamlessly to the Internet.


 6) For networks where I have a
 password, should that go in wpa_supplicant.conf as plain text or should it
 be 

Re: [gentoo-user] WPA Supplicant

2011-04-24 Thread dhkuhl


- Original Message -
From: Mick 
Date: Sunday, April 24, 2011 9:14 am
Subject: Re: [gentoo-user] WPA Supplicant
To: gentoo-user@lists.gentoo.org

 On Sunday 24 April 2011 13:37:03 dhk...@optonline.net wrote:
  I'm trying to get wireless working reliably on my laptop. I 
 have followed
  the documentation, but still have a lot of questions . . . 
 especially since it only seems to work sometimes.
  
  First, I'm using WPA Supplicant without the gui tools.
 
 run wpa_gui from a terminal and a lot of what you're asking 
 below will become 
 self-explanatory.
 
 
  Second, I'm in NYC and there are a lot of networks I can use 
 around town.
  
  Third, Is there a way I can control the services I use from 
 the Grub menu? 
  Since the laptop has a wireless card and an RJ45 jack, I'd 
 like to be able
  boot and not use one or the other. Since I know if I'm not 
 physically connected to a network, there's really no reason to 
 even try eth0.
 
 Check /etc/conf.d/rc and in particular:
 
 # RC_NET_STRICT_CHECKING allows some flexibility with the 'net' 
 service.# The following values are allowed:
 # none - The 'net' service is always considered up.
 # no - This basically means that at least one net.* service 
 besides net.lo
 # must be up. This can be used by notebook users that 
 have a wifi 
 and
 # a static nic, and only wants one up at any given time 
 to have the
 # 'net' service seen as up.
 # lo - This is the same as the 'no' option, but net.lo is 
 also counted.
 # This should be useful to people that do not care 
 about any specific
 # interface being up at boot.
 # yes - For this ALL network interfaces MUST be up for the 
 'net' service to
 # be considered up.
 
 RC_NET_STRICT_CHECKING=no
 
 (or you can use lo)
 
 
  Forth, The problem. I'm not sure how wpa_supplicant works or 
 how it should
  work. The wpa_supplicant man page gives a few examples on how 
 to run it,
  but when I look at the process list it seems to be run by 
 another program
  called wpa_cli. There's also a shell script in 
 /etc/wpa/supplicant that
  looks like it can start or stop it with CONNECT or DISCONNECT. 
 1) Do I
  need to enter networks in wpa_supplicant.conf or does 
 wpa_supplicant scan
  for networks and connect to whatever's available? 
 
 The latter.
 
 You can however enter manually in 
 /etc/wpa_supplicant/wpa_supplicant.conf 
 particular parameters (keys and what not) of known networks to 
 which you 
 connect as a matter of preference.
 
 
  2) If I have multiple
  networks available how does wpa_supplicant choose which to 
 connect to and
  can I specify which one I want? 
 
 It'll connect to:
 
 a) Any network you have specified in your 
 /etc/wpa_supplicant/wpa_supplicant.conf according to the 
 preference you have 
 set up therein.
 
 b) Any network it finds.
 
 c) Any network you select with wpa_cli, or select/enable/disable 
 in wpa_gui.
 
 
  3) How should wpa_supplicant be started,
  stopped and restarted? What should be used for this: 
 wpa_supplicant, wpa_cli, or wpa_cli.sh? I don't see anything 
 in /etc/init.d for that, but
  it looks like netmount may be doing it. 
 
 You need to define it in /etc/conf.d/net:
 
 modules=( wpa_supplicant )
 wpa_supplicant_wlan0=-Dwext 
 
 (adjust this according to the name of your wireless iface and driver).
 
 
  4) The documentation doesn't say
  to, but the way I got wireless working is by creating a link 
 net.wlan0 -
  net.lo in the /etc/init.d directory. Is this correct? 
 
 It depends which documentation you are looking at. I am sure 
 that this is 
 explained in the gentoo Handbook and associated documentation.
 
 This is the link you need:
 
 lrwxrwxrwx 1 root root 6 Dec 16 14:26 net.wlan0 - net.lo
 
 but you should have also configured /etc/conf.d/net with your 
 desired settings 
 or just defaults will run.
 
 
  I think that's why
  it's starting automatically when I boot too, because I never 
 added it with
  rc-update so netmount must be picking it up. 
 
 
  5) This is the most puzzling
  thing. When wpa_supplicant starts even though I get a inet 
 address I
  can't always get to the internet. Why does the panel applet 
 says I'm
  connected and ifconfig shows an inet address but firefox and 
 ping can't
  reach a site like yahoo or google? 
 
 This could well be a dns server/repeater issue.
 
 If you can ping the IP address of google, but not the domain 
 name of it, then 
 the problem is that you do not have access to a DNS repeater. 
 Look in your 
 /etc/resolve.conf to see if there is a line saying:
 
 nameserver XXX.XXX.XX.XX
 
 if it is absent then you have not connected to a namesever. 
 This is a router 
 issue and it could be controlled by some authentication scheme. 
 A lot of 
 wireless services offered by coffee shops, libraries, etc. may 
 give you an IP 
 address automatically, but then require you use your browser to 
 register with 
 their authentication server (using a passwd that they provide 
 after you pay 
 them for the 

Re: [gentoo-user] zlib and WOFF

2011-04-24 Thread luis jure
on 2011-04-24 at 13:48 Volker Armin Hemmann wrote:


 http://bugs.gentoo.org/show_bug.cgi?id=311147

you might add yourself to that bug mentioning lillypond.

reading the thread, i don't think it's the same bug. anyway, i'll follow
your advise (and stroller's) and i'll file a bug both for zlib and
lilypond. i'll also ask at the lilypond mailing list.

best,

lj



[gentoo-user] Reinstall Gentoo? [Was: Building pygtk-2.22.0-r1 fails. Help, please!]

2011-04-24 Thread Alan Mackenzie
Hi, Mick.

On Sun, Apr 24, 2011 at 09:17:45AM +0100, Mick wrote:
 On Saturday 23 April 2011 21:06:25 Alan Mackenzie wrote:
  On Sat, Apr 23, 2011 at 08:46:30PM +0100, Mick wrote:

   What do you get when you run:

   # eselect python list

  Available Python interpreters:
[1]   python2.6 *
[2]   python2.7
[3]   python3.1

 OK, the next stage would be to change your python to the latest stable:

 eselect python set 2

DONE.

 and then remerge those packages that were linked against the old
 python:

 python-updater -v -p

 to get a list of these.

That gives me a list of 24 packages.  Am I meant to actually run
python-updater without the -p, here?

 When you finish all this you can run:

 emerge --depclean -v -p

 It should now ask you to remove the old python, but check carefully the
 remaining packages in case something important is in the list and
 breaks your system.

I do emerge --depclean -v -p.  It says I should run emerge -uDN @world
first.  I'm a bit apprehensive about this, since the world update says it
would reemerge 138 packages (I'm not sure whether this is top-level
(whatever that means) packages or the real total).  In that list are 3
blockages I don't know wha do do with.  My experience suggests this will
not work smoothly, and I'll likely be left with a non-working (or even a
non-bootable) system.

How come?  Well, I started my installation in February 2010, and with one
thing and another, didn't get it finished, so it went into cold storage
until a month ago.  I've had so much trouble trying to get updated, that
it might be better to start again from scratch with a new stage3 (or even
a new installation CD).  This would surely leave my home directory and
suchlike untouched.  What do you think?

 -- 
 Regards,
 Mick

-- 
Alan Mackenzie (Nuremberg, Germany).



Re: [gentoo-user] Reinstall Gentoo? [Was: Building pygtk-2.22.0-r1 fails. Help, please!]

2011-04-24 Thread Mick
On Sunday 24 April 2011 14:25:58 Alan Mackenzie wrote:
 Hi, Mick.
 
 On Sun, Apr 24, 2011 at 09:17:45AM +0100, Mick wrote:
  On Saturday 23 April 2011 21:06:25 Alan Mackenzie wrote:
   On Sat, Apr 23, 2011 at 08:46:30PM +0100, Mick wrote:
What do you get when you run:

# eselect python list
   
   Available Python interpreters:
 [1]   python2.6 *
 [2]   python2.7
 [3]   python3.1
  
  OK, the next stage would be to change your python to the latest stable:
  
  eselect python set 2
 
 DONE.
 
  and then remerge those packages that were linked against the old
  python:
  
  python-updater -v -p
  
  to get a list of these.
 
 That gives me a list of 24 packages.  Am I meant to actually run
 python-updater without the -p, here?

That's correct.  As the man emerge say -p stands for --pretend.  Just to give 
a chance to see what it wants to do and think about it before you run it again 
without it for execution.

You need to do this next.


  When you finish all this you can run:
  
  emerge --depclean -v -p
  
  It should now ask you to remove the old python, but check carefully the
  remaining packages in case something important is in the list and
  breaks your system.
 
 I do emerge --depclean -v -p.  It says I should run emerge -uDN @world
 first.  I'm a bit apprehensive about this, since the world update says it
 would reemerge 138 packages (I'm not sure whether this is top-level
 (whatever that means) packages or the real total).  In that list are 3
 blockages I don't know wha do do with.  My experience suggests this will
 not work smoothly, and I'll likely be left with a non-working (or even a
 non-bootable) system.

At this stage you should only run:

python-updater -v

Nothing else.

Once it completes you can run --depclean which will ask you to remove the 
older 2.6 python package.


 How come?  Well, I started my installation in February 2010, and with one
 thing and another, didn't get it finished, so it went into cold storage
 until a month ago.  I've had so much trouble trying to get updated, that
 it might be better to start again from scratch with a new stage3 (or even
 a new installation CD).  This would surely leave my home directory and
 suchlike untouched.  What do you think?

Adding the -N flag will remerge any packages that are affected by changes to 
USE flags that you have made since they were first installed.  So the list 
will be longer than without it.
-- 
Regards,
Mick


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-user] Reinstall Gentoo? [Was: Building pygtk-2.22.0-r1 fails. Help, please!]

2011-04-24 Thread Mick
On Sunday 24 April 2011 16:44:05 Mick wrote:
 On Sunday 24 April 2011 14:25:58 Alan Mackenzie wrote:
  Hi, Mick.
  
  On Sun, Apr 24, 2011 at 09:17:45AM +0100, Mick wrote:
   On Saturday 23 April 2011 21:06:25 Alan Mackenzie wrote:
On Sat, Apr 23, 2011 at 08:46:30PM +0100, Mick wrote:
 What do you get when you run:
 
 # eselect python list

Available Python interpreters:
  [1]   python2.6 *
  [2]   python2.7
  [3]   python3.1
   
   OK, the next stage would be to change your python to the latest stable:
   
   eselect python set 2
  
  DONE.
  
   and then remerge those packages that were linked against the old
   python:
   
   python-updater -v -p
   
   to get a list of these.
  
  That gives me a list of 24 packages.  Am I meant to actually run
  python-updater without the -p, here?
 
 That's correct.  As the man emerge say -p stands for --pretend.  Just to
 give a chance to see what it wants to do and think about it before you run
 it again without it for execution.
 
 You need to do this next.
 
   When you finish all this you can run:
   
   emerge --depclean -v -p
   
   It should now ask you to remove the old python, but check carefully the
   remaining packages in case something important is in the list and
   breaks your system.
  
  I do emerge --depclean -v -p.  It says I should run emerge -uDN @world
  first.  I'm a bit apprehensive about this, since the world update says it
  would reemerge 138 packages (I'm not sure whether this is top-level
  (whatever that means) packages or the real total).  In that list are 3
  blockages I don't know wha do do with.  My experience suggests this will
  not work smoothly, and I'll likely be left with a non-working (or even a
  non-bootable) system.
 
 At this stage you should only run:
 
 python-updater -v
 
 Nothing else.
 
 Once it completes you can run --depclean which will ask you to remove the
 older 2.6 python package.
 
  How come?  Well, I started my installation in February 2010, and with one
  thing and another, didn't get it finished, so it went into cold storage
  until a month ago.  I've had so much trouble trying to get updated, that
  it might be better to start again from scratch with a new stage3 (or even
  a new installation CD).  This would surely leave my home directory and
  suchlike untouched.  What do you think?
 
 Adding the -N flag will remerge any packages that are affected by changes
 to USE flags that you have made since they were first installed.  So the
 list will be longer than without it.

Post any blockers shown if you don't know what you need to do about them.
-- 
Regards,
Mick


signature.asc
Description: This is a digitally signed message part.


[gentoo-user] Packages needing masked 'introspection' keyword

2011-04-24 Thread Graham Murray
After syncing a few minutes ago,

emerge -puDv --reinstall changed-use  --autounmask=y @world @system

These are the packages that would be merged, in order:

Calculating dependencies... done!
[ebuild U  ] media-libs/libpng-1.4.7 [1.4.5] USE=-static-libs 535 kB
[ebuild U  ] dev-libs/gobject-introspection-0.10.8 [0.10.7-r1] USE=doc 
-test 1,001 kB
[ebuild U  ] x11-libs/gdk-pixbuf-2.22.1-r1 [2.22.1] USE=X doc 
(introspection*) jpeg jpeg2k tiff -debug -test 0 kB
[ebuild   R] dev-libs/atk-1.32.0  USE=doc (introspection*) nls 0 kB
[ebuild   R] x11-libs/pango-1.28.4  USE=X doc (introspection*) -debug 
-test 0 kB
[ebuild   R] media-video/mjpegtools-1.9.0-r1  USE=dv gtk mmx png quicktime 
sdl v4l yv12 -dga (-X%*) 0 kB
[ebuild  N ] media-video/gpac-0.4.5-r4  USE=a52 aac alsa ffmpeg ipv6 jack 
javascript jpeg jpeg2k mad opengl oss png pulseaudio sdl ssl theora truetype 
vorbis xml xvid -debug -wxwidgets 0 kB 


[ebuild U  ] dev-db/mysql-5.1.56 [5.1.53] USE=community embedded perl ssl 
-big-tables -cluster -debug -extraengine -latin1 -max-idx-128 -minimal -pbxt 
-profiling (-selinux) -static -test -xtradb 0 kB   


[ebuild   R] x11-libs/gtk+-3.0.9  USE=cups doc examples introspection* 
vim-syntax (-aqua) -debug -test -xinerama 0 kB
[ebuild U  ] app-text/libwpd-0.8.14-r1 [0.8.14] USE=doc -test% -tools% 0 
kB
[ebuild  N ] sci-chemistry/openbabel-python-2.3.0  0 kB
[ebuild   R] sci-chemistry/openbabel-2.3.0  USE=doc perl* python* 
-wxwidgets 0 kB

Total: 12 packages (5 upgrades, 2 new, 5 reinstalls), Size of downloads: 1,536 
kB

The following USE changes are necessary to proceed:
#required by x11-libs/gtk+-3.0.9, required by dev-ada/gtkada-2.10.0, required 
by @selected, required by @world (argument)
=x11-libs/pango-1.28.4 introspection
#required by x11-libs/gtk+-3.0.9, required by dev-ada/gtkada-2.10.0, required 
by @selected, required by @world (argument)
=dev-libs/atk-1.32.0 introspection
#required by x11-libs/gtk+-3.0.9, required by dev-ada/gtkada-2.10.0, required 
by @selected, required by @world (argument)
=x11-libs/gdk-pixbuf-2.22.1-r1 introspection


As the introspection USE flag is shown in brackets for these
packages. it is not possible to enable them as required,



Re: [gentoo-user] WPA Supplicant

2011-04-24 Thread Neil Bothwick
On Sun, 24 Apr 2011 12:37:03 + (GMT), dhk...@optonline.net wrote:

 I'm trying to get wireless working reliably on my laptop.  I have
 followed the documentation, but still have a lot of questions . . .
 especially since it only seems to work sometimes. First, I'm using WPA
 Supplicant without the gui tools. 
 Second, I'm in NYC and there are a lot of networks I can use around
 town. 
 Third, Is there a way I can control the services I use from the Grub
 menu?  Since the laptop has a wireless card and an RJ45 jack, I'd like
 to be able boot and not use one or the other.  Since I know if I'm not
 physically connected to a network, there's really no reason to even try
 eth0. Forth, The problem.  I'm not sure how wpa_supplicant works or how
 it should work.

Wicd answers all of these questions. It connects via eth0 if a cable is
connected, otherwise it takes care of multiple wireless networks. It
avoids all the hassles with wpa_supplicant too.


-- 
Neil Bothwick

B?#$^f, said Pooh, as line noise garbled his transmission.


signature.asc
Description: PGP signature


Re: [gentoo-user] Reinstall Gentoo? [Was: Building pygtk-2.22.0-r1 fails. Help, please!]

2011-04-24 Thread Stroller

On 24/4/2011, at 4:44pm, Mick wrote:
 ...
 At this stage you should only run:
 
 python-updater -v
 
 Nothing else.


Doing this today I've had a couple of packages that needed to be emerged with 
USE=-doc when they failed.

Stroller.


 


Re: [gentoo-user] Reinstall Gentoo? [Was: Building pygtk-2.22.0-r1 fails. Help, please!]

2011-04-24 Thread David W Noon
On Sun, 24 Apr 2011 21:40:02 +0200, Stroller wrote about Re:
[gentoo-user] Reinstall Gentoo?  [Was: Building pygtk-2.22.0-r1 fails.
Help, please!]:

On 24/4/2011, at 4:44pm, Mick wrote:
 ...
 At this stage you should only run:
 
 python-updater -v
 
 Nothing else.


Doing this today I've had a couple of packages that needed to be
emerged with USE=-doc when they failed.

Those would be jinja and sphinx.  They are notorious for their circular
dependency, which requires USE='-doc' to bypass.
-- 
Regards,

Dave  [RLU #314465]
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
dwn...@ntlworld.com (David W Noon)
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*


signature.asc
Description: PGP signature


Re: [gentoo-user] Reinstall Gentoo? [Was: Building pygtk-2.22.0-r1 fails. Help, please!]

2011-04-24 Thread Neil Bothwick
On Sun, 24 Apr 2011 21:17:23 +0100, David W Noon wrote:

 Doing this today I've had a couple of packages that needed to be
 emerged with USE=-doc when they failed.  
 
 Those would be jinja and sphinx.  They are notorious for their circular
 dependency, which requires USE='-doc' to bypass.

It's rarely desirable to enable doc globally. It is best to enable only
for those packages where you need extended documentation.


-- 
Neil Bothwick

A bit of tolerance is worth a megabyte of flaming. -- Henry Spencer


signature.asc
Description: PGP signature


Re: [gentoo-user] Reinstall Gentoo? [Was: Building pygtk-2.22.0-r1 fails. Help, please!]

2011-04-24 Thread Mick
On Sunday 24 April 2011 21:30:33 Neil Bothwick wrote:
 On Sun, 24 Apr 2011 21:17:23 +0100, David W Noon wrote:
  Doing this today I've had a couple of packages that needed to be
  emerged with USE=-doc when they failed.
  
  Those would be jinja and sphinx.  They are notorious for their circular
  dependency, which requires USE='-doc' to bypass.
 
 It's rarely desirable to enable doc globally. It is best to enable only
 for those packages where you need extended documentation.

@Alan Mackenzie:

What Neil is saying can be achieved by setting package specific USE flags in 
the file /etc/portage/package.use; e.g. use an entry like:

dev-python/jinja -doc i18n

to exclude USE flag doc, but include i18n.
-- 
Regards,
Mick


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-user] Gentoo on Cygwin

2011-04-24 Thread Alex Schuster

Nils Andresen writes:

 2011/4/19 Alex Schuster wo...@wonkology.org:
  He was somewhat successful in emerging @system, then he wrote the wiki
  article :)

 Alex,
 have you re-tried lately?

Although I like the idea very much, I never tried this myself. I only know 
that Al posted some questions here and told about his plan, and I see that 
he wrote the wiki article.

Wonko