Re: [gentoo-dev] rfc: openrc use flag

2011-04-21 Thread William Hubbs
On Thu, Apr 21, 2011 at 10:03:42PM +0200, Pacho Ramos wrote:
> El jue, 21-04-2011 a las 14:30 -0500, William Hubbs escribió:
> > On Wed, Apr 20, 2011 at 08:20:32PM +0200, Pacho Ramos wrote:
> > > El mié, 20-04-2011 a las 22:02 +0400, Peter Volkov escribió:
> > > > В Срд, 20/04/2011 в 12:24 -0500, William Hubbs пишет:
> > > > > The author of the bug feels that the way to fix this is for us to put 
> > > > > a
> > > > > check in openrc that makes it refuse to run services if it was not 
> > > > > used
> > > > > in the boot process.
> > > > 
> > > > This is good idea to have in any case since I remember my system went
> > > > crazy after I've tried to start some service inside chroot.
> > > > 
> > > > > This may work; however, I do not feel that it addresses the root cause
> > > > > of the bug. I feel that the root cause is packages unconditionally
> > > > > installing udev rules which assume everyone uses openrc.
> > > > 
> > > > I'd voted to have both implemented.
> > > > 
> > > 
> > > I would vote for the first one, I still don't like "openrc" USE flag
> > > approach much because:
> > > 1. Would need to rebuild some packages when switching between init
> > > systems.
> > 
> > I don't think you can get away from this, no matter how you approach
> > it. The other approach I thought of is to include the udev pieces
> > directly in openrc and make it possible to build openrc with or without
> > udev integration. That will still mean you have to rebuild openrc though
> > if you want udev support.
> > 
> 
> With mgorny's approach looks like recompiling wouldn't be needed :-/

His approach doesn't stop udev from attempting to run the services; it
just makes openrc print an error message and abort each service we try
to run.

With the integration into udev that we have right now, some of our udev
rules expect openrc to be usable.

William



pgpJbcbS7O7Jw.pgp
Description: PGP signature


Re: [gentoo-dev] rfc: openrc use flag

2011-04-21 Thread Pacho Ramos
El jue, 21-04-2011 a las 14:30 -0500, William Hubbs escribió:
> On Wed, Apr 20, 2011 at 08:20:32PM +0200, Pacho Ramos wrote:
> > El mié, 20-04-2011 a las 22:02 +0400, Peter Volkov escribió:
> > > В Срд, 20/04/2011 в 12:24 -0500, William Hubbs пишет:
> > > > The author of the bug feels that the way to fix this is for us to put a
> > > > check in openrc that makes it refuse to run services if it was not used
> > > > in the boot process.
> > > 
> > > This is good idea to have in any case since I remember my system went
> > > crazy after I've tried to start some service inside chroot.
> > > 
> > > > This may work; however, I do not feel that it addresses the root cause
> > > > of the bug. I feel that the root cause is packages unconditionally
> > > > installing udev rules which assume everyone uses openrc.
> > > 
> > > I'd voted to have both implemented.
> > > 
> > 
> > I would vote for the first one, I still don't like "openrc" USE flag
> > approach much because:
> > 1. Would need to rebuild some packages when switching between init
> > systems.
> 
> I don't think you can get away from this, no matter how you approach
> it. The other approach I thought of is to include the udev pieces
> directly in openrc and make it possible to build openrc with or without
> udev integration. That will still mean you have to rebuild openrc though
> if you want udev support.
> 

With mgorny's approach looks like recompiling wouldn't be needed :-/


> > 2. I remember (from "logrotate" USE flag case) that using an USE flag
> > for simply installing or not a file is not usually preferred :-/
> 
> In the logrotate use flag case, that decision was made because a user
> can use INSTALL_MASK="/etc/logrotate.d" in make.conf to block those
> files. But that argument definitely does not apply here. If the user
> doesn't want this support what should he set INSTALL_MASK to?
>  

That's true, thanks for the explanation :)

>  William
> 




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


Re: [gentoo-dev] rfc: openrc use flag

2011-04-21 Thread William Hubbs
On Wed, Apr 20, 2011 at 08:20:32PM +0200, Pacho Ramos wrote:
> El mié, 20-04-2011 a las 22:02 +0400, Peter Volkov escribió:
> > В Срд, 20/04/2011 в 12:24 -0500, William Hubbs пишет:
> > > The author of the bug feels that the way to fix this is for us to put a
> > > check in openrc that makes it refuse to run services if it was not used
> > > in the boot process.
> > 
> > This is good idea to have in any case since I remember my system went
> > crazy after I've tried to start some service inside chroot.
> > 
> > > This may work; however, I do not feel that it addresses the root cause
> > > of the bug. I feel that the root cause is packages unconditionally
> > > installing udev rules which assume everyone uses openrc.
> > 
> > I'd voted to have both implemented.
> > 
> 
> I would vote for the first one, I still don't like "openrc" USE flag
> approach much because:
> 1. Would need to rebuild some packages when switching between init
> systems.

I don't think you can get away from this, no matter how you approach
it. The other approach I thought of is to include the udev pieces
directly in openrc and make it possible to build openrc with or without
udev integration. That will still mean you have to rebuild openrc though
if you want udev support.

> 2. I remember (from "logrotate" USE flag case) that using an USE flag
> for simply installing or not a file is not usually preferred :-/

In the logrotate use flag case, that decision was made because a user
can use INSTALL_MASK="/etc/logrotate.d" in make.conf to block those
files. But that argument definitely does not apply here. If the user
doesn't want this support what should he set INSTALL_MASK to?
 
 William



pgpzVS3OLZ7u8.pgp
Description: PGP signature


Re: [gentoo-dev] rfc: openrc use flag

2011-04-21 Thread William Hubbs
On Wed, Apr 20, 2011 at 11:34:59PM -0500, William Hubbs wrote:
> On Thu, Apr 21, 2011 at 04:31:46AM +0200, Jeroen Roovers wrote:
> > How should the udev rules be changed to match /any/ init system?
> 
> For an example of the problem, take a look on your system at
> /lib/udev/rules.d/90-network.rules. This is part of openrc's hotplug
> functionality. Basically it tries to run
> /etc/init.d/net.$INTERFACE start when a network interface is added to
> the system and rc_hotplug allows it.
>  
> In net-wireless/bluez, you will see similar integration between openrc
> and udev. Once bluez is installed, /lib/udev/rules.d/70-bluetooth.rules
> runs bluetooth.sh which tries to run a service in /etc/init.d.
> 
> To make things work with /any/ init system, the best way to go would be
> to make it a practice not to run services from within udev rules or
> external run scripts like net.sh and bluetooth.sh in /lib/udev.

Another option for this would be to include this level of udev
integration in the openrc package itself instead of spreading it through
the other packages.

If I do that, when you install openrc, you would optionally get those
pieces that integrate it with udev instead of those pieces coming from
the individual packages.

Comments?

William



pgpTgyLblWyvb.pgp
Description: PGP signature


[gentoo-dev] Last rites: gnustep-libs/camaelon and x11-themes/camaelon-themes

2011-04-21 Thread Bernard Cafarelli

# camaelon does not compile with current gnustep versions in tree,
# deprecated by upstream, use gnustep built-in theming suppport instead
# Also remove associated themes, they do not work with built-in engine
# Masked for removal in 30 days, bug #328559
gnustep-libs/camaelon
x11-themes/camaelon-themes

--
Bernard Cafarelli (Voyageur)
Gentoo developer (NX, GNUstep, net-misc, ...)



[gentoo-dev] RFC: gnustep layout change news item

2011-04-21 Thread Bernard Cafarelli
Upstream has changed the default layout for gnustep applications:
standard FHS layout is now recommended (as opposed to our
current prefix /usr/GNUstep).

Switching to this layout has some advantages for us: staying close to
upstream, standard paths, gnustep apps running fine without too many
environment variables, ... However it does imply remerging all
installed gnustep packages, hence this news item as an early warning to
users.

I plan to commit it on 2011-04-26, and drop the package.mask for the
corresponding ~arch packages a few days later

Comments and reviews welcome!
-- 
Bernard Cafarelli (Voyageur)
Gentoo developer (NX, GNUstep, net-misc, ...)
Title: GNUstep packages new layout
Author: Bernard Cafarelli 
Content-Type: text/plain
Posted: 2011-04-26
Revision: 1
News-Item-Format: 1.0
Display-If-Installed: =2.6.0.

This change means that you have to re-emerge all installed packages
depending on GNUstep to move them to the new layout. You can use
gnustep-base/gnustep-updater for this step