Re: [gentoo-user] installing systemd

[gottlieb]

On Mon, Aug 05 2013, Canek Peláez Valdés wrote:

> On Mon, Aug 5, 2013 at 9:59 PM,   wrote:
>> On Mon, Aug 05 2013, Canek Peláez Valdés wrote:
>>
>>> On Mon, Aug 5, 2013 at 8:26 PM,   wrote:
 (I brought an old system upto the state of my real one and am trying to
 follow the wiki for converting to systemd).

 I am upto the part where the wiki says

emerge --ask systemd

 I believe the wiki left out unmerging or something since the emerge
 gives conflicts (as expected).

 What must I unmerge?

 The emerge --ask systemd, specifically mentions a conflict with
 consolekit, which it states is pulled in by polkit and pambase.

 I tried adding these last three to package.mask hoping that the emerge
 would propose uninstalling them while installing systemd, but that
 didn't happen.

 What is/are the step(s) I am missing.
>>>
>>> If you already compiled everything with USE=-consolekit, just
>>> uninstall consolekit. Otherwise, first set -consolekit in your USE
>>> flags, emerge --newuse world, and then uninstall it.
>>
>> That is it. I never set USE=-consolekit.  If it is in the wiki, I missed
>> it.  Thanks.
>>
>>> logind (included with systemd) provides all the functionality that
>>> consolekit provided, and more. They cannot be installed at the same
>>> time (hence the block).
>>>
>>> Be aware, if you had some packages with USE=consolekit (pambase and
>>> polkit in particular), you cannot use GDM nor GNOME after you
>>> uninstall it and until you install systemd and reboot with it.
>>
>> I don't understand this last point.  You said that I should
>> 1.  set USE="-consolekit ..."
>> 2.  emerge --newuse world
>> 3.  unmerge consolekit
>> Since I do steps 1 and 2 before 3, by the time I get to uninstalling
>> consolekit, I can be sure that neither pambase and polkit (nor anything
>> else) will not be installed with USE=consolekit.  So I don't see how
>> your last sentence can apply.
>
> Mmmh. You are missing "4. reboot with systemd". I'm assuming you
> haven't booted with systemd yet (you seemed hesitant to do it); logind
> will not work under anything different than systemd. The problem is
> not uninstalling or not CK (that doesn't matter after "USE=-consolekit
> emerge --newuse world", as you correctly pointed); the problem is that
> you need to be running systemd for logind to work.
>
> If logind doesn't work, gdm (and the whole GNOME stack) will acty
> strangely, if at all.
>
>>> Strange things will happen if you do.
>>> Login via ssh and VT will work as usual.
>>
>> ssh and VT will be sufficient.
>
> Regards.

I see.  I misunderstood you and though the problems were to arise before
step 4.

I get it now.  I don't know if I can work on this right away as we are
traveling but will report back when I do try it.

Thanks,
allan

Re: [gentoo-user] installing systemd

[Canek Peláez Valdés]

On Mon, Aug 5, 2013 at 9:59 PM,   wrote:
> On Mon, Aug 05 2013, Canek Peláez Valdés wrote:
>
>> On Mon, Aug 5, 2013 at 8:26 PM,   wrote:
>>> (I brought an old system upto the state of my real one and am trying to
>>> follow the wiki for converting to systemd).
>>>
>>> I am upto the part where the wiki says
>>>
>>>emerge --ask systemd
>>>
>>> I believe the wiki left out unmerging or something since the emerge
>>> gives conflicts (as expected).
>>>
>>> What must I unmerge?
>>>
>>> The emerge --ask systemd, specifically mentions a conflict with
>>> consolekit, which it states is pulled in by polkit and pambase.
>>>
>>> I tried adding these last three to package.mask hoping that the emerge
>>> would propose uninstalling them while installing systemd, but that
>>> didn't happen.
>>>
>>> What is/are the step(s) I am missing.
>>
>> If you already compiled everything with USE=-consolekit, just
>> uninstall consolekit. Otherwise, first set -consolekit in your USE
>> flags, emerge --newuse world, and then uninstall it.
>
> That is it. I never set USE=-consolekit.  If it is in the wiki, I missed
> it.  Thanks.
>
>> logind (included with systemd) provides all the functionality that
>> consolekit provided, and more. They cannot be installed at the same
>> time (hence the block).
>>
>> Be aware, if you had some packages with USE=consolekit (pambase and
>> polkit in particular), you cannot use GDM nor GNOME after you
>> uninstall it and until you install systemd and reboot with it.
>
> I don't understand this last point.  You said that I should
> 1.  set USE="-consolekit ..."
> 2.  emerge --newuse world
> 3.  unmerge consolekit
> Since I do steps 1 and 2 before 3, by the time I get to uninstalling
> consolekit, I can be sure that neither pambase and polkit (nor anything
> else) will not be installed with USE=consolekit.  So I don't see how
> your last sentence can apply.

Mmmh. You are missing "4. reboot with systemd". I'm assuming you
haven't booted with systemd yet (you seemed hesitant to do it); logind
will not work under anything different than systemd. The problem is
not uninstalling or not CK (that doesn't matter after "USE=-consolekit
emerge --newuse world", as you correctly pointed); the problem is that
you need to be running systemd for logind to work.

If logind doesn't work, gdm (and the whole GNOME stack) will acty
strangely, if at all.

>> Strange things will happen if you do.
>> Login via ssh and VT will work as usual.
>
> ssh and VT will be sufficient.

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Re: [gentoo-user] installing systemd

[gottlieb]

On Mon, Aug 05 2013, Canek Peláez Valdés wrote:

> On Mon, Aug 5, 2013 at 8:26 PM,   wrote:
>> (I brought an old system upto the state of my real one and am trying to
>> follow the wiki for converting to systemd).
>>
>> I am upto the part where the wiki says
>>
>>emerge --ask systemd
>>
>> I believe the wiki left out unmerging or something since the emerge
>> gives conflicts (as expected).
>>
>> What must I unmerge?
>>
>> The emerge --ask systemd, specifically mentions a conflict with
>> consolekit, which it states is pulled in by polkit and pambase.
>>
>> I tried adding these last three to package.mask hoping that the emerge
>> would propose uninstalling them while installing systemd, but that
>> didn't happen.
>>
>> What is/are the step(s) I am missing.
>
> If you already compiled everything with USE=-consolekit, just
> uninstall consolekit. Otherwise, first set -consolekit in your USE
> flags, emerge --newuse world, and then uninstall it.

That is it. I never set USE=-consolekit.  If it is in the wiki, I missed
it.  Thanks.

> logind (included with systemd) provides all the functionality that
> consolekit provided, and more. They cannot be installed at the same
> time (hence the block).
>
> Be aware, if you had some packages with USE=consolekit (pambase and
> polkit in particular), you cannot use GDM nor GNOME after you
> uninstall it and until you install systemd and reboot with it.

I don't understand this last point.  You said that I should
1.  set USE="-consolekit ..."
2.  emerge --newuse world
3.  unmerge consolekit
Since I do steps 1 and 2 before 3, by the time I get to uninstalling
consolekit, I can be sure that neither pambase and polkit (nor anything
else) will not be installed with USE=consolekit.  So I don't see how
your last sentence can apply.

> Strange things will happen if you do.
> Login via ssh and VT will work as usual.

ssh and VT will be sufficient.

thanks again,
allan

Re: [gentoo-user] installing systemd

[Canek Peláez Valdés]

On Mon, Aug 5, 2013 at 8:26 PM,   wrote:
> (I brought an old system upto the state of my real one and am trying to
> follow the wiki for converting to systemd).
>
> I am upto the part where the wiki says
>
>emerge --ask systemd
>
> I believe the wiki left out unmerging or something since the emerge
> gives conflicts (as expected).
>
> What must I unmerge?
>
> The emerge --ask systemd, specifically mentions a conflict with
> consolekit, which it states is pulled in by polkit and pambase.
>
> I tried adding these last three to package.mask hoping that the emerge
> would propose uninstalling them while installing systemd, but that
> didn't happen.
>
> What is/are the step(s) I am missing.

If you already compiled everything with USE=-consolekit, just
uninstall consolekit. Otherwise, first set -consolekit in your USE
flags, emerge --newuse world, and then uninstall it.

logind (included with systemd) provides all the functionality that
consolekit provided, and more. They cannot be installed at the same
time (hence the block).

Be aware, if you had some packages with USE=consolekit (pambase and
polkit in particular), you cannot use GDM nor GNOME after you
uninstall it and until you install systemd and reboot with it. Strange
things will happen if you do. Login via ssh and VT will work as usual.

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

[gentoo-user] installing systemd

[gottlieb]

(I brought an old system upto the state of my real one and am trying to
follow the wiki for converting to systemd).

I am upto the part where the wiki says

   emerge --ask systemd

I believe the wiki left out unmerging or something since the emerge
gives conflicts (as expected).

What must I unmerge?

The emerge --ask systemd, specifically mentions a conflict with
consolekit, which it states is pulled in by polkit and pambase.

I tried adding these last three to package.mask hoping that the emerge
would propose uninstalling them while installing systemd, but that
didn't happen.

What is/are the step(s) I am missing.

thanks,
allan

Re: [gentoo-user] Moving from old udev to eudev

[Walter Dnes]

On Mon, Aug 05, 2013 at 10:10:45AM -0400, Anthony G. Basile wrote

> For now.  And you get a ton of bloat.  I removed over 300 unused 
> functions.

  Wonderful.  It reminds me of...  
http://www.quotationspage.com/quote/26979.html

> Perfection is achieved, not when there is nothing more to add,
> but when there is nothing left to take away.

> Antoine de Saint-Exupery
> French writer (1900 - 1944)

  That is a saying that should be taken to heart by more programmers,
both in the linux and Microsoft worlds.

-- 
Walter Dnes 
I don't run "desktop environments"; I run useful applications

Re: [gentoo-user] Moving from old udev to eudev

[Walter Dnes]

On Mon, Aug 05, 2013 at 09:18:38PM +0100, Neil Bothwick wrote
> On Mon, 05 Aug 2013 10:24:27 -0400, Tanstaafl wrote:
> 
> > > But there's not a lot of point as eudev isn't that different to udev
> > > now, AFAICT, and a recent update forced me to switch back to udev
> > > because eudev hadn't been updated (on ~amd64).  
> > 
> > Can you elaborate on what this update was that forced you to go back to 
> > regular udev? 
> 
> I can't remember what it was now, and it may have been avoidable by
> making virtual/udev-206 (or whichever version it was that needed a higher
> udev version than eudev could provide). It's moot now as eudev has been
> updated and portage is happy again, but it would be a concern if this
> happened regularly.

  I ran into this.  Here is what I think happened...

- I specified "sys-fs/eudev-1.2-r1-beta ~amd64" (or something similar)
  in my /etc/portage/package.keywords file
- I ran "emerge --sync".  On that particular day, it removed the beta
  version ebuild, and replaced it with eudev-1.2.ebuild
- "emerge --changed-use --deep --update @world" could no longer find an
  unmasked version of sys-fs/eudev that satisfied virtual/udev.  So it
  fell back to a version of sys-fs/udev
- My workaround, *UNTIL SUCH TIME AS EUDEV HITS STABLE AMD64*, is...

I don't run "desktop environments"; I run useful applications

Re: [gentoo-user] Moving from old udev to eudev

[Walter Dnes]

On Mon, Aug 05, 2013 at 01:19:34PM +0300, Samuli Suominen wrote

> That might be the systemd upstream view point, but definately isn't mine.

  Your view and mine don't matter.  Upstream's view matters.  That's how
we end up with fiascos like GNOME and Microsoft's Metro interface.

> Fact is that udev can be built and ran standalone without systemd and 
> you don't need eudev for that.

  Kay Sievers, *THE LEAD DEVELOPER* specifically says in
http://lists.freedesktop.org/archives/systemd-devel/2012-July/006065.html

> We promised to keep udev properly *running* as standalone, we never
> told that it can be *build* standalone. And that still stands

I.e. no promise of being able to build standalone.

> If udev upstream makes it impossible to build, or run it standalone
> then we need to patch or fork it -- but that's far from now.

[...deletia...]

> I mean, why the heck fork something too early when upstream still 
> supports udev on non-systemd init systems?!

  Let's say that that it happens 2 years from now, after udev has been
getting ever more tightly integrated into systemd.  At that point, it'll
be way too late.  The udev source will have all sorts of hooks into
systemd, at least at build-time.  Creating a stand-alone build in a few
weeks would be painful.  Another option is to dig up 2-year-old source
code for the last stand-alone version of udev and update it in a rush.
The old version would depend on libs no-longer in the tree, and other
apps would depend on a newer udev.  You yourself, pointed out in
http://www.mail-archive.com/gentoo-user@lists.gentoo.org/msg139485.html

> By udev maintainers forcing them to upgrade to the new keymap hwdb
> which required version to be raised to up-to-par with udev-206.

  Imagine 2 years of such updates to catch up with in a few weeks.  It's
too late to start building the fire-escapes when the fire-alarm goes
off.  Similarly, if we want a viable alternative udev, that means having
it (eudev) maintained and up-to-date and ready at all times.  I'm sorry
that it has come to this, but the current udev maintainers have made it
clear which way they want to go, and it's not the way that I and a lot
of other people want to go.  Don't blame us for getting out while the
getting out is still good.

-- 
Walter Dnes 
I don't run "desktop environments"; I run useful applications

[gentoo-user] Re: export LC_CTYPE=en_US.UTF-8

[Kai Krakow]

Chris Stankevitz  schrieb:

> 3. The handbook suggests that I should add this line to
> /etc/env.d/02locale: 'LC_COLLATE="C"', but I do not know if they are
> again talking about the language "DE".  RESOLUTION: I assumed
> LC_COLLATE=C refers to english and added the line without
> modification.

C refers to "as in C code"... Or something like that. What's essential: It 
tells the system to use the strings like they are recorded in the program 
file, without translating. In most cases that's English.

HTH
Kai

Re: [gentoo-user] export LC_CTYPE=en_US.UTF-8

[Chris Stankevitz]

On Mon, Aug 5, 2013 at 11:53 AM, Mike Gilbert  wrote:
> The handbook documents setting a system-wide default locale. You
> generally do this by setting the LANG variable in
> /etc/conf.d/02locale.
>
> http://www.gentoo.org/doc/en/handbook/handbook-amd64.xml?part=1&chap=8#doc_chap3_sect3

Mike,

Thank you for your help.  I attempted to follow these instructions and
ran into three problems.  Can you please confirm the fixes I employed
to deal with each of these issues:

1. The handbook suggests I should modify the file /etc/env.d/02locale,
but that file does not exist on my system.  RESOLUTION: create the
file

2. The handbook suggests I should add this line to
/etc/env.d/02locale: 'LANG="de_DE.UTF-8"', but I do not speak the
language "DE".  RESOLUTION: type instead 'LANG="en_US.UTF-8"' to match
/etc/locale.gen

3. The handbook suggests that I should add this line to
/etc/env.d/02locale: 'LC_COLLATE="C"', but I do not know if they are
again talking about the language "DE".  RESOLUTION: I assumed
LC_COLLATE=C refers to english and added the line without
modification.

Thank you again for your help,

Chris

Re: [gentoo-user] SLES or gentoo ... ?

[Stefan G. Weichinger]

Am 05.08.2013 23:55, schrieb Alan McKinnon:
> On 05/08/2013 23:20, Stefan G. Weichinger wrote:
>>
>> ! Pls don't flame me :-) !
>>
>> Don't misunderstand ... I happily trust gentoo on most of my customers
>> servers.
>>
>> A customer of mine chose SLES 10 back then to run a VMware-Server
>> installation (1.0.x back then) because they had some big
>> company-license-pool available.
>>
>> That server is to be replaced and I asked them if they still want to use
>> SLES because of that. Answer: no ... no more licenses available/paid.
>>
>> So they asked me for alternatives and I told them about gentoo.
>>
>> My question:
>>
>> how would you guys compare the 2 choices to report it back to them?
>>
>> People buy stuff like SLES to get/feel the feeling that all the choice
>> and review of changes is done for them  we didn't need one
>> support-call in the last few years. And the gentoo-community is a
>> helpful and competent one (yes, thank you!).
>>
>> So I tend to do the job with gentoo ... better they pay my work than
>> some never-used support-contract ;-)
>>
>> I just plan to use stable gentoo there, be conservative with changes and
>> keep the system up-to-date regularly ... as I use ~amd64 on my main
>> machines I think I am rather informed about any *bigger* or problematic
>> upgrades.
>>
>> It's gonna be a QEMU/KVM-host .. this and some rather powerful server
>> should speed up those smallish and dusty VMs.
>>
>> Any thoughts? How to professionally deploy gentoo linux as a one-man-show?
>>
>> ;-)
>>
>> Best regards, Stefan
>>
> 
> 
> I despise SLES intensely[1]. Even more than Windows. So this may be biased.
> 
> The primary question as I see it is
> 
> Who will maintain this installation?
> 
> If the answer is you and 1|2 guys you train yourself, by all means go
> right ahead and use gentoo. You already know it well so the quality of
> service you offer a customer is likely to be better than if you went say
> Centos.
> 
> If the answer is you plus other guys but you don't know who they are or
> how good they are or if you get to train them, then gentoo starts
> getting risky. You don't want a gentoo box where the admin is the
> "emerge world && reboot" and walk away kind of guy.
> 
> I think you fall in the first class. And our Infrastructure team has
> also never had to log a VMware service call. Our managed service team
> that faces clients - very different story and not applicable here.
> 
> I find that gentoo does not scale well in corporates where machines are
> a mix of everything. It takes too much brain power to update them. It
> also doesn't work well if you have to give admin rights to people of
> little skill.
> 
> Where gentoo shines is
> 
> - small installs that need something none standard
> - large pools of identical hosts that are somehow non-standard so you
> get to build what you want once and deploy it many times
> - embedded. Your tools let you automate the build end-to-end
> 
> 
> 
> [1] A predecessor used SLES 9 & 10 for everything coz he thought it was
> awesome. Nothing could ever get updated as it was always manual,
> SuSEconfig kept biting us in the teeth hard and it was just awful for
> anyone used to working on *nix at any leveol. Fine for Windows admins
> moving over though...


Thanks. That's what I wanted to read and fished for ;-)

Stefan

Re: [gentoo-user] SLES or gentoo ... ?

[Alan McKinnon]

On 05/08/2013 23:20, Stefan G. Weichinger wrote:
> 
> ! Pls don't flame me :-) !
> 
> Don't misunderstand ... I happily trust gentoo on most of my customers
> servers.
> 
> A customer of mine chose SLES 10 back then to run a VMware-Server
> installation (1.0.x back then) because they had some big
> company-license-pool available.
> 
> That server is to be replaced and I asked them if they still want to use
> SLES because of that. Answer: no ... no more licenses available/paid.
> 
> So they asked me for alternatives and I told them about gentoo.
> 
> My question:
> 
> how would you guys compare the 2 choices to report it back to them?
> 
> People buy stuff like SLES to get/feel the feeling that all the choice
> and review of changes is done for them  we didn't need one
> support-call in the last few years. And the gentoo-community is a
> helpful and competent one (yes, thank you!).
> 
> So I tend to do the job with gentoo ... better they pay my work than
> some never-used support-contract ;-)
> 
> I just plan to use stable gentoo there, be conservative with changes and
> keep the system up-to-date regularly ... as I use ~amd64 on my main
> machines I think I am rather informed about any *bigger* or problematic
> upgrades.
> 
> It's gonna be a QEMU/KVM-host .. this and some rather powerful server
> should speed up those smallish and dusty VMs.
> 
> Any thoughts? How to professionally deploy gentoo linux as a one-man-show?
> 
> ;-)
> 
> Best regards, Stefan
> 


I despise SLES intensely[1]. Even more than Windows. So this may be biased.

The primary question as I see it is

Who will maintain this installation?

If the answer is you and 1|2 guys you train yourself, by all means go
right ahead and use gentoo. You already know it well so the quality of
service you offer a customer is likely to be better than if you went say
Centos.

If the answer is you plus other guys but you don't know who they are or
how good they are or if you get to train them, then gentoo starts
getting risky. You don't want a gentoo box where the admin is the
"emerge world && reboot" and walk away kind of guy.

I think you fall in the first class. And our Infrastructure team has
also never had to log a VMware service call. Our managed service team
that faces clients - very different story and not applicable here.

I find that gentoo does not scale well in corporates where machines are
a mix of everything. It takes too much brain power to update them. It
also doesn't work well if you have to give admin rights to people of
little skill.

Where gentoo shines is

- small installs that need something none standard
- large pools of identical hosts that are somehow non-standard so you
get to build what you want once and deploy it many times
- embedded. Your tools let you automate the build end-to-end



[1] A predecessor used SLES 9 & 10 for everything coz he thought it was
awesome. Nothing could ever get updated as it was always manual,
SuSEconfig kept biting us in the teeth hard and it was just awful for
anyone used to working on *nix at any leveol. Fine for Windows admins
moving over though...

-- 
Alan McKinnon
alan.mckin...@gmail.com

[gentoo-user] SLES or gentoo ... ?

[Stefan G. Weichinger]

! Pls don't flame me :-) !

Don't misunderstand ... I happily trust gentoo on most of my customers
servers.

A customer of mine chose SLES 10 back then to run a VMware-Server
installation (1.0.x back then) because they had some big
company-license-pool available.

That server is to be replaced and I asked them if they still want to use
SLES because of that. Answer: no ... no more licenses available/paid.

So they asked me for alternatives and I told them about gentoo.

My question:

how would you guys compare the 2 choices to report it back to them?

People buy stuff like SLES to get/feel the feeling that all the choice
and review of changes is done for them  we didn't need one
support-call in the last few years. And the gentoo-community is a
helpful and competent one (yes, thank you!).

So I tend to do the job with gentoo ... better they pay my work than
some never-used support-contract ;-)

I just plan to use stable gentoo there, be conservative with changes and
keep the system up-to-date regularly ... as I use ~amd64 on my main
machines I think I am rather informed about any *bigger* or problematic
upgrades.

It's gonna be a QEMU/KVM-host .. this and some rather powerful server
should speed up those smallish and dusty VMs.

Any thoughts? How to professionally deploy gentoo linux as a one-man-show?

;-)

Best regards, Stefan

Re: [gentoo-user] export LC_CTYPE=en_US.UTF-8

[Mike Gilbert]

On Mon, Aug 5, 2013 at 2:57 PM, Bruce Hill
 wrote:
> On Mon, Aug 05, 2013 at 02:53:11PM -0400, Mike Gilbert wrote:
>> On Mon, Aug 5, 2013 at 2:25 PM, Chris Stankevitz
>>  wrote:
>> > Hello,
>> >
>> > I am using svn to update a repository.  Somebody added files to the
>> > repository with weird characters in the filename.  SVN refuses to
>> > update the respository unless I first:
>> >
>> > export LC_CTYPE=en_US.UTF-8
>> >
>> > I don't know or really care what that mumbo jumbo means, but I would
>> > like an answer to this question:
>> >
>> > Is my gentoo system properly setup?  If not, what step did I miss that
>> > is causing svn to want me to export LC_CTYPE?
>> >
>> > I suspect either my gentoo system is messed up or svn is messed up.
>> >
>>
>> Sparing you the details as requested: In general, you want to be using
>> a locale that ends with ".UTF-8" to avoid encoding issues with
>> software like python and subversion.
>>
>> The handbook documents setting a system-wide default locale. You
>> generally do this by setting the LANG variable in
>> /etc/conf.d/02locale.
>>
>> http://www.gentoo.org/doc/en/handbook/handbook-amd64.xml?part=1&chap=8#doc_chap3_sect3
>
> Without looking, shouldn't that be /etc/env.d/02locale ?

Yes.

Or /etc/locale.conf if you're on systemd.

Re: [gentoo-user] Moving from old udev to eudev

[Tanstaafl]

On 2013-08-05 4:18 PM, Neil Bothwick  wrote:

On Mon, 05 Aug 2013 10:24:27 -0400, Tanstaafl wrote:


But there's not a lot of point as eudev isn't that different to udev
now, AFAICT, and a recent update forced me to switch back to udev
because eudev hadn't been updated (on ~amd64).


Can you elaborate on what this update was that forced you to go back to
regular udev?


I can't remember what it was now, and it may have been avoidable by
making virtual/udev-206 (or whichever version it was that needed a higher
udev version than eudev could provide). It's moot now as eudev has been
updated and portage is happy again, but it would be a concern if this
happened regularly.


Agreed... Anthony, can you comment on the likelihood of this happening 
in the future? An occasional temporary issue wouldn't trouble me, as I 
already wait at least a few days before updating anything critical, and 
it isn't like this kind of thing hasn't happened for regular udev...


Thanks for the reply Neil...

Re: [gentoo-user] Moving from old udev to eudev

[Neil Bothwick]

On Mon, 05 Aug 2013 10:24:27 -0400, Tanstaafl wrote:

> > But there's not a lot of point as eudev isn't that different to udev
> > now, AFAICT, and a recent update forced me to switch back to udev
> > because eudev hadn't been updated (on ~amd64).  
> 
> Can you elaborate on what this update was that forced you to go back to 
> regular udev? 

I can't remember what it was now, and it may have been avoidable by
making virtual/udev-206 (or whichever version it was that needed a higher
udev version than eudev could provide). It's moot now as eudev has been
updated and portage is happy again, but it would be a concern if this
happened regularly.


-- 
Neil Bothwick

PC DOS Error #01: Windows loading, come back tomorrow


signature.asc
Description: PGP signature

Re: [gentoo-user] export LC_CTYPE=en_US.UTF-8

[Bruce Hill]

On Mon, Aug 05, 2013 at 02:53:11PM -0400, Mike Gilbert wrote:
> On Mon, Aug 5, 2013 at 2:25 PM, Chris Stankevitz
>  wrote:
> > Hello,
> >
> > I am using svn to update a repository.  Somebody added files to the
> > repository with weird characters in the filename.  SVN refuses to
> > update the respository unless I first:
> >
> > export LC_CTYPE=en_US.UTF-8
> >
> > I don't know or really care what that mumbo jumbo means, but I would
> > like an answer to this question:
> >
> > Is my gentoo system properly setup?  If not, what step did I miss that
> > is causing svn to want me to export LC_CTYPE?
> >
> > I suspect either my gentoo system is messed up or svn is messed up.
> >
> 
> Sparing you the details as requested: In general, you want to be using
> a locale that ends with ".UTF-8" to avoid encoding issues with
> software like python and subversion.
> 
> The handbook documents setting a system-wide default locale. You
> generally do this by setting the LANG variable in
> /etc/conf.d/02locale.
> 
> http://www.gentoo.org/doc/en/handbook/handbook-amd64.xml?part=1&chap=8#doc_chap3_sect3
 
Without looking, shouldn't that be /etc/env.d/02locale ?
-- 
Happy Penguin Computers   >')
126 Fenco Drive   ( \
Tupelo, MS 38801   ^^
supp...@happypenguincomputers.com
662-269-2706 662-205-6424
http://happypenguincomputers.com/

A: Because it messes up the order in which people normally read text.   

   
Q: Why is top-posting such a bad thing? 

   
A: Top-posting. 

   
Q: What is the most annoying thing in e-mail?

Don't top-post: http://en.wikipedia.org/wiki/Top_post#Top-posting

Re: [gentoo-user] /usr/locale vs /usr/share/locale

[Mike Gilbert]

On Mon, Aug 5, 2013 at 8:17 AM, Alexey Mishustin  wrote:
> Hello list,
>
> I have noticed today this folder on my laptop: /usr/locale. It
> contains multiple subfolders-languages with LC_MESSAGES.
>
> I wonder why these LC_MESSAGES are situated there, and not at
> /usr/share/locale, as on my other machine with Gentoo installed.
>
> Equery says that /usr/locale belongs to gnome-icon-theme 3.6.2. But
> the same gnome-icon-theme 3.6.2 didn't create /usr/locale on my other
> machine, created /usr/share/locale instead. Reemerging
> gnome-icon-theme didn't help.
>

Sounds like a bug somewhere. Can you provide a build log for
gnome-icon-theme? You will need to set the PORT_LOGDIR option in
make.conf to keep build logs after a successful merge.

Re: [gentoo-user] export LC_CTYPE=en_US.UTF-8

[Mike Gilbert]

On Mon, Aug 5, 2013 at 2:25 PM, Chris Stankevitz
 wrote:
> Hello,
>
> I am using svn to update a repository.  Somebody added files to the
> repository with weird characters in the filename.  SVN refuses to
> update the respository unless I first:
>
> export LC_CTYPE=en_US.UTF-8
>
> I don't know or really care what that mumbo jumbo means, but I would
> like an answer to this question:
>
> Is my gentoo system properly setup?  If not, what step did I miss that
> is causing svn to want me to export LC_CTYPE?
>
> I suspect either my gentoo system is messed up or svn is messed up.
>

Sparing you the details as requested: In general, you want to be using
a locale that ends with ".UTF-8" to avoid encoding issues with
software like python and subversion.

The handbook documents setting a system-wide default locale. You
generally do this by setting the LANG variable in
/etc/conf.d/02locale.

http://www.gentoo.org/doc/en/handbook/handbook-amd64.xml?part=1&chap=8#doc_chap3_sect3

[gentoo-user] Re: Browsers cannot access WWW while ping and host utilities work as expected.

[Grant Edwards]

On 2013-08-04, Mick  wrote:
> On Sunday 04 Aug 2013 19:56:08 gevisz wrote:

>> The ifconfig utility shows that instead of eth0 I have somewhat
>> strange enp2s15 but I have played with this and found no difference
>> after forcing udev to name it eth0.
>
> This is the new kernel naming scheme of NICs.

No, it's not.

It's the new _udev_ naming scheme.  The kernel names things the same
as always (eth0, wlan0, and so on depending depends somewhat on the
driver itself, which can provide the base name).

-- 
Grant Edwards   grant.b.edwardsYow! OVER the underpass!
  at   UNDER the overpass!
  gmail.comAround the FUTURE and
   BEYOND REPAIR!!

[gentoo-user] export LC_CTYPE=en_US.UTF-8

[Chris Stankevitz]

Hello,

I am using svn to update a repository.  Somebody added files to the
repository with weird characters in the filename.  SVN refuses to
update the respository unless I first:

export LC_CTYPE=en_US.UTF-8

I don't know or really care what that mumbo jumbo means, but I would
like an answer to this question:

Is my gentoo system properly setup?  If not, what step did I miss that
is causing svn to want me to export LC_CTYPE?

I suspect either my gentoo system is messed up or svn is messed up.

Thank you,

Chris

Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.

[Alan McKinnon]

On 05/08/2013 16:41, Bruce Hill wrote:
> On Mon, Aug 05, 2013 at 04:31:44PM +0200, Marc Joliet wrote:
>> Am Mon, 5 Aug 2013 07:59:09 -0500
>> schrieb Bruce Hill :
>>
>>> If this is "the new kernel naming scheme of NICs", why this in dmesg:
>>>
>>> [4.725902] systemd-udevd[1176]: renamed network interface wlan0 to 
>>> enp0s18f2u2
>>>
>>> It looks as if systemd-udev renamed the NIC to me. Can you explain?
>>
>> It already has been explained in the previous NIC renaming discussion: what's
>> broken is renaming a device within the kernels internal namespace, which
>> contains eth*, wlan* (and maybe others). The problem is that there is a race
>> condition with the kernel when renaming ethX to ethY. What you *can* do is
>> rename ethX to somethingelseX or somethingelseY, because then you are not 
>> racing
>> against the kernel to hand out device names.
>>
>> This is explained on the website that also explains the new default renaming
>> scheme used by udev. I (and IIRC others, too) already linked to it in in the 
>> old
>> thread, and the relevant news item also referenced it, but here it is again:
>>
>>   
>> http://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames/
> 
> The fact is that udev renamed the NIC. For the average Joe with one NIC (very
> large percentage of users) this is a non sequitur. For those of us with 2 or
> more NICs, myself included, we have already setup our systems to use multiple
> NICs for a purpose and configured the system so that nothing can/will/needs to
> rename subsequent NICs.
> 
> My point is don't say "the new kernel naming scheme of NICs", say "the new
> systemd naming scheme of NICs".
> 

Let me see if I can clarify somewhat.

eth0 can be considered the "kernel name" - the kernel named the NIC
according to it's own rules using the info it had available. Kernel
names depend on discovery order and to a lesser degree on the kernel
code (a dev could change how things are done for example)

enp0s18f2u2 can be considered the "userspace name" - it's derived from
the slot the card is plugged into, and is set by udev. The kernel
doesn't really care about this stuff, but you might. Most people think
of their NICS on multi-NIC machines in terms of positions i.e. "third
one on the left" and can't work with "whatever eth2 happens to be today"

So why change this? Because you can't rely on ethX always being the same
physical hardware. On a firewall or router, you absolutely need to rely
on this. The udev scheme works around this by letting you specify exact
rules that will always do what you want.

Why was this changed rammed down your throat? Well, that is political.

The udev maintainers (along with systemd) work for Red Hat. RH's market
is almost totally servers, and big multi-nic ones at that. They really
need consistent names, doubly so if the host is a virtualization host.

The catch: RH (or more exactly the udev maintainer employed by RH)
probably couldn't give a toss what you think or want, and went ahead and
fixed their problem expecting you to "deal with it or shove off"

Does all that fit better with what you see before you?

[All of this is what I've inferred over months, it's my opinion in my
words. You won't find this description with anyone else's name attached
:-) ]


-- 
Alan McKinnon
alan.mckin...@gmail.com

Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.

[Stroller]

On 5 August 2013, at 15:41, Bruce Hill wrote:
> … For those of us with 2 or
> more NICs, myself included, we have already setup our systems to use multiple
> NICs for a purpose and configured the system so that nothing can/will/needs to
> rename subsequent NICs.

Except you can't say that if you have a bank of a thousand servers all 
net-booting a shared system image. 

Sys admins on the Dell Poweredge list were very pleased when they learned of 
the persistent device names plan.

Stroller.

Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.

[Mick]

On Monday 05 Aug 2013 15:41:09 Bruce Hill wrote:
> On Mon, Aug 05, 2013 at 04:31:44PM +0200, Marc Joliet wrote:
> > Am Mon, 5 Aug 2013 07:59:09 -0500
> > 
> > schrieb Bruce Hill :
> > > If this is "the new kernel naming scheme of NICs", why this in dmesg:
> > > 
> > > [4.725902] systemd-udevd[1176]: renamed network interface wlan0 to
> > > enp0s18f2u2
> > > 
> > > It looks as if systemd-udev renamed the NIC to me. Can you explain?
> > 
> > It already has been explained in the previous NIC renaming discussion:
> > what's broken is renaming a device within the kernels internal
> > namespace, which contains eth*, wlan* (and maybe others). The problem is
> > that there is a race condition with the kernel when renaming ethX to
> > ethY. What you *can* do is rename ethX to somethingelseX or
> > somethingelseY, because then you are not racing against the kernel to
> > hand out device names.
> > 
> > This is explained on the website that also explains the new default
> > renaming scheme used by udev. I (and IIRC others, too) already linked to
> > it in in the old
> > 
> > thread, and the relevant news item also referenced it, but here it is 
again:
> >   http://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInte
> >   rfaceNames/
> 
> The fact is that udev renamed the NIC. For the average Joe with one NIC
> (very large percentage of users) this is a non sequitur. For those of us
> with 2 or more NICs, myself included, we have already setup our systems to
> use multiple NICs for a purpose and configured the system so that nothing
> can/will/needs to rename subsequent NICs.
> 
> My point is don't say "the new kernel naming scheme of NICs", say "the new
> systemd naming scheme of NICs".

Indeed!  Thanks for bringing this to my attention.  Here's my eth0:

[6.437527] systemd-udevd[1407]: starting version 204
[7.457924] systemd-udevd[1428]: renamed network interface eth0 to enp11s0

while my wireless NIC stays named as always was (wlan0):

[7.822350] b43-phy0: Broadcom 4312 WLAN found (core revision 15)
[7.838741] b43-phy0: Found PHY: Analog 6, Type 5 (LP), Revision 1
[7.838760] b43-phy0 debug: Found Radio: Manuf 0x17F, Version 0x2062, 
Revision 2
[   15.771370] b43-phy0: Loading firmware version 666.2 (2011-02-23 01:15:07)
[   15.775217] b43-phy0 debug: b2062: Using crystal tab entry 19200 kHz.
[   17.157109] b43-phy0 debug: Chip initialized
[   17.157427] b43-phy0 debug: 64-bit DMA initialized
[   17.157888] b43-phy0 debug: QoS disabled
[   17.167424] b43-phy0 debug: Wireless interface started
[   17.172410] b43-phy0 debug: Adding Interface type 2
[   17.173097] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready


BTW, I have no systemd installed, only udev-204 and udev-init-scripts-26.
-- 
Regards,
Mick


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

Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.

[Marc Joliet]

Am Mon, 5 Aug 2013 09:41:09 -0500
schrieb Bruce Hill :

[...]
> My point is don't say "the new kernel naming scheme of NICs", say "the new
> systemd naming scheme of NICs".

Oh, oops. I don't know how I managed to misread your email that spectacularly.
Sorry!

-- 
Marc Joliet
--
"People who think they know everything really annoy those of us who know we
don't" - Bjarne Stroustrup


signature.asc
Description: PGP signature

Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.

[Bruce Hill]

On Mon, Aug 05, 2013 at 04:31:44PM +0200, Marc Joliet wrote:
> Am Mon, 5 Aug 2013 07:59:09 -0500
> schrieb Bruce Hill :
> 
> > If this is "the new kernel naming scheme of NICs", why this in dmesg:
> > 
> > [4.725902] systemd-udevd[1176]: renamed network interface wlan0 to 
> > enp0s18f2u2
> > 
> > It looks as if systemd-udev renamed the NIC to me. Can you explain?
> 
> It already has been explained in the previous NIC renaming discussion: what's
> broken is renaming a device within the kernels internal namespace, which
> contains eth*, wlan* (and maybe others). The problem is that there is a race
> condition with the kernel when renaming ethX to ethY. What you *can* do is
> rename ethX to somethingelseX or somethingelseY, because then you are not 
> racing
> against the kernel to hand out device names.
> 
> This is explained on the website that also explains the new default renaming
> scheme used by udev. I (and IIRC others, too) already linked to it in in the 
> old
> thread, and the relevant news item also referenced it, but here it is again:
> 
>   
> http://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames/

The fact is that udev renamed the NIC. For the average Joe with one NIC (very
large percentage of users) this is a non sequitur. For those of us with 2 or
more NICs, myself included, we have already setup our systems to use multiple
NICs for a purpose and configured the system so that nothing can/will/needs to
rename subsequent NICs.

My point is don't say "the new kernel naming scheme of NICs", say "the new
systemd naming scheme of NICs".
-- 
Happy Penguin Computers   >')
126 Fenco Drive   ( \
Tupelo, MS 38801   ^^
supp...@happypenguincomputers.com
662-269-2706 662-205-6424
http://happypenguincomputers.com/

A: Because it messes up the order in which people normally read text.   

   
Q: Why is top-posting such a bad thing? 

   
A: Top-posting. 

   
Q: What is the most annoying thing in e-mail?

Don't top-post: http://en.wikipedia.org/wiki/Top_post#Top-posting

Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.

[Marc Joliet]

Am Mon, 5 Aug 2013 07:59:09 -0500
schrieb Bruce Hill :

> On Mon, Aug 05, 2013 at 11:06:52AM +0100, Mick wrote:
> > 
> > > Suggestions of  Michael Kintzios
> > 
> > > > This is the new kernel naming scheme of NICs.  Which-ever nomenclature
> > > > you decide to use, check that that's the only one having a symlink in
> > > > /etc/init.d to net.lo
> > > 
> > > Yes, there is only enp2s15 links to lo in /etc/init.d
> > 
> > The idea here is that you need consistent naming of your iface.  If you 
> > have 
> > settled on the kernel naming of enp2s15, then stick with this throughout 
> > your 
> > configuration.
> > -- 
> > Regards,
> > Mick
> 
> If this is "the new kernel naming scheme of NICs", why this in dmesg:
> 
> [4.725902] systemd-udevd[1176]: renamed network interface wlan0 to 
> enp0s18f2u2
> 
> It looks as if systemd-udev renamed the NIC to me. Can you explain?

It already has been explained in the previous NIC renaming discussion: what's
broken is renaming a device within the kernels internal namespace, which
contains eth*, wlan* (and maybe others). The problem is that there is a race
condition with the kernel when renaming ethX to ethY. What you *can* do is
rename ethX to somethingelseX or somethingelseY, because then you are not racing
against the kernel to hand out device names.

This is explained on the website that also explains the new default renaming
scheme used by udev. I (and IIRC others, too) already linked to it in in the old
thread, and the relevant news item also referenced it, but here it is again:

  
http://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames/

HTH
-- 
Marc Joliet
--
"People who think they know everything really annoy those of us who know we
don't" - Bjarne Stroustrup


signature.asc
Description: PGP signature

Re: [gentoo-user] Moving from old udev to eudev

[Tanstaafl]

On 2013-08-05 10:10 AM, Anthony G. Basile  wrote:

On 08/05/2013 06:19 AM, Samuli Suominen wrote:

That might be the systemd upstream view point, but definately isn't mine.
Fact is that udev can be built and ran standalone without systemd and
you don't need eudev for that.



For now.


And this is ultimately my primary concern. After reading the huge 
threads surrounding this debacle, I wouldn't trust Lennart (or the other 
systemd devs) on any promise to not remove this ability at a later date.



And you get a ton of bloat.


And this is the second concern. Both of these are why I decided to go 
with eudev. I looked at mdev, but it looked a lot more complicated to 
get right than eudev (correct me if I'm wrong someone).


Thanks again Anthony for your work maintaining eudev for 'the rest of 
us'... :)

Re: [gentoo-user] Moving from old udev to eudev

[Tanstaafl]

Going back and re-reading finds this answer to my other last question - 
also from you Neil (so thanks again!)...


But I'm curious...

On 2013-08-01 2:43 PM, Neil Bothwick  wrote:

On Thu, 01 Aug 2013 12:28:38 -0400, Tanstaafl wrote:


>I've googled until my fingers are blue, but cannot for the life of me
>find any explicit instructions for*how*  to switch from udev to eudev.

emerge -Ca udev
emerge -1a eudev

But there's not a lot of point as eudev isn't that different to udev now,
AFAICT, and a recent update forced me to switch back to udev because
eudev hadn't been updated (on ~amd64).


Can you elaborate on what this update was that forced you to go back to 
regular udev? This is the only 'concern' that I have right now, and 
this is the first comment I've seen from anyone about anything like this...


Thanks again,

Charles

Re: [gentoo-user] Moving from old udev to eudev

[Anthony G. Basile]

On 08/05/2013 06:19 AM, Samuli Suominen wrote:

On 04/08/13 05:56, Walter Dnes wrote:

On Fri, Aug 02, 2013 at 05:02:39AM +0300, Samuli Suominen wrote


Looking forward to lastrite sys-fs/eudev just like
sys-apps/module-init-tools already was removed as unnecessary later on.


   You want eudev removed, and Lennart Poettering wants udev on
non-systemd systems dropped.  Add those two items together, and we get
systemd rammed down our throats...

http://lists.freedesktop.org/archives/systemd-devel/2012-August/006066.html



(Yes, udev on non-systemd systems is in our eyes a dead end, in case
you haven't noticed it yet. I am looking forward to the day when we
can drop that support entirely.)




That might be the systemd upstream view point, but definately isn't mine.
Fact is that udev can be built and ran standalone without systemd and
you don't need eudev for that.


For now.  And you get a ton of bloat.  I removed over 300 unused 
functions.  Furthermore, there is a problem with iface renaming which 
Ian solved and legacy features are not there.  We will continue to 
support a bootable system with separate /usr without need for an intramfs.


But most importantly, you have a different upstream with a different 
attitude towards the users.  Even if the codebase were identical, this 
makes all the difference to those who want a system the way they want 
and not the way systemd upstream wants.  Your arguments have been 
ineffective at convincing people because you dismiss this critical point.




If udev upstream makes it impossible to build, or run it standalone then
we need to patch or fork it -- but that's far from now.
In any case there will always be sys-fs/udev and it will never require
sys-apps/systemd.
Futhermore sys-fs/udev will be the default for long as sys-apps/openrc
is the default.

I mean, why the heck fork something too early when upstream still
supports udev on non-systemd init systems?!

- Samuli



--
Anthony G. Basile, Ph. D.
Chair of Information Technology
D'Youville College
Buffalo, NY 14201
(716) 829-8197

Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.

[Bruce Hill]

On Mon, Aug 05, 2013 at 11:06:52AM +0100, Mick wrote:
> 
> > Suggestions of  Michael Kintzios
> 
> > > This is the new kernel naming scheme of NICs.  Which-ever nomenclature
> > > you decide to use, check that that's the only one having a symlink in
> > > /etc/init.d to net.lo
> > 
> > Yes, there is only enp2s15 links to lo in /etc/init.d
> 
> The idea here is that you need consistent naming of your iface.  If you have 
> settled on the kernel naming of enp2s15, then stick with this throughout your 
> configuration.
> -- 
> Regards,
> Mick

If this is "the new kernel naming scheme of NICs", why this in dmesg:

[4.725902] systemd-udevd[1176]: renamed network interface wlan0 to 
enp0s18f2u2

It looks as if systemd-udev renamed the NIC to me. Can you explain?

Cheers,
Bruce
-- 
Happy Penguin Computers   >')
126 Fenco Drive   ( \
Tupelo, MS 38801   ^^
supp...@happypenguincomputers.com
662-269-2706 662-205-6424
http://happypenguincomputers.com/

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

Don't top-post: http://en.wikipedia.org/wiki/Top_post#Top-posting

[gentoo-user] Re: /usr/locale vs /usr/share/locale

[Alexey Mishustin]

2013/8/5 Alexey Mishustin :
> Hello list,
>
> I have noticed today this folder on my laptop: /usr/locale. It
> contains multiple subfolders-languages with LC_MESSAGES.
>
> I wonder why these LC_MESSAGES are situated there, and not at
> /usr/share/locale, as on my other machine with Gentoo installed.
>
> Equery says that /usr/locale belongs to gnome-icon-theme 3.6.2. But
> the same gnome-icon-theme 3.6.2 didn't create /usr/locale on my other
> machine, created /usr/share/locale instead. Reemerging
> gnome-icon-theme didn't help.
>
> Gentoo Localization HowTo says: "The locales and their data are part
> of the system library and can be found at /usr/share/locale on most
> systems". I can't figure out why the laptop hasn't been included in
> "most systems".
>
> Which options related to locales may I have configured otherwise? In
> my opinion, installations are near equal. Arch - i686, locale - utf-8,
> desktop - Openbox. The world file is almost the same; some
> laptop-specific packages added. The only difference that I remember: I
> installed many packages on the laptop in another order.

The folder /usr/share/locale exists on the laptop too, and there are
much more files.

-- 
Regards,
Alex

Re: [gentoo-user] How much effort from udev-197-r3 to 206

[Nick Khamis]

please read the news it's a must. If your network card drivers are
built as modules ''in theory'' you are not effected. If they are built
into the kernel, you will have to delete a 70- something file, and
replace it with an 80- something empty file, to keep the same ethN
wlanN names. Sorry, i do not remember the actual file names. Please
google ''udev thay slut'' to see my original post about this.

N

On 8/3/13, Harry Putnam  wrote:
> I see an update to udev come up when investigating installing various
> other pkgs.  eix shows I'm on 197-r3 and the most recent is 206.
>
> Will that be a hefty amount of change... and concomittant amount of
> work? Or something a lazy slug can manage?
>
>
>

[gentoo-user] /usr/locale vs /usr/share/locale

[Alexey Mishustin]

Hello list,

I have noticed today this folder on my laptop: /usr/locale. It
contains multiple subfolders-languages with LC_MESSAGES.

I wonder why these LC_MESSAGES are situated there, and not at
/usr/share/locale, as on my other machine with Gentoo installed.

Equery says that /usr/locale belongs to gnome-icon-theme 3.6.2. But
the same gnome-icon-theme 3.6.2 didn't create /usr/locale on my other
machine, created /usr/share/locale instead. Reemerging
gnome-icon-theme didn't help.

Gentoo Localization HowTo says: "The locales and their data are part
of the system library and can be found at /usr/share/locale on most
systems". I can't figure out why the laptop hasn't been included in
"most systems".

Which options related to locales may I have configured otherwise? In
my opinion, installations are near equal. Arch - i686, locale - utf-8,
desktop - Openbox. The world file is almost the same; some
laptop-specific packages added. The only difference that I remember: I
installed many packages on the laptop in another order.

-- 
Regards,
Alex

Re: [gentoo-user] Moving from old udev to eudev

[Walter Dnes]

On Mon, Aug 05, 2013 at 06:12:02AM -0400, Anthony G. Basile wrote

> I am the current lead.  You may follow the activity here [1].
> 
> [1] https://github.com/gentoo/eudev/commits/master

  Thank you very much for your work on eudev, from an end-user who
benefits from your work.

-- 
Walter Dnes 
I don't run "desktop environments"; I run useful applications

Re: [gentoo-user] Complete list of USE flags?

[Daniel Troeder]

Am 04.08.2013 20:31, schrieb Pandu Poluan:
> Hello guys,
> 
> I'm a bit ashamed to ask this question, as it belies how long I haven't
> actually installed a 'lightweight' Gentoo system...
> 
> But I digress. On to my question:
> 
> Anyone knows an exhaustive list of USE flags?
> 
> And a related subquestion:
> 
> Is the USE flags list at znurt.org  up-to-date?
> 
> The reason I'm asking, is because I'm planning on building *very*
> lightweight systems with as small attack surface as possible.
> 
> Rgds,
> --
> 
I have two symlinks in all of my gentoo installations in $HOME:
use.desc   -> /usr/portage/profiles/use.desc
use.local.desc -> /usr/portage/profiles/use.local.desc

When I want to look "something" up, I use:
$ grep something ~/use*

To narrow results down, I sometimes prepend or append ':' to "something".

Greetings,
Daniel

Re: [gentoo-user] Moving from old udev to eudev

[Samuli Suominen]

On 05/08/13 13:27, Marc Stürmer wrote:

Why is was forked you ask? Because of the predictable Name stuff and
some People disliked the attitude of the udev programmer which was
"either my way or the high way." aside choice is always Good to have so
in the end IT was bound to happen sooner or later and is a Good thing to
have.



nope, the forking happened before predictable network interface names.
and forking udev was never the smart choice here, but it would be rather 
easy to port the old rule generator as a standalone udev helper and make 
it use free names like lan0, wireless0.

as in, you don't change whole car if your tire blows out

Re: [gentoo-user] Moving from old udev to eudev

[Marc Stürmer]

Why is was forked you ask? Because of the predictable Name stuff and some
People disliked the attitude of the udev programmer which was "either my
way or the high way." aside choice is always Good to have so in the end IT
was bound to happen sooner or later and is a Good thing to have.

Re: [gentoo-user] Moving from old udev to eudev

[Samuli Suominen]

On 04/08/13 05:56, Walter Dnes wrote:

On Fri, Aug 02, 2013 at 05:02:39AM +0300, Samuli Suominen wrote


Looking forward to lastrite sys-fs/eudev just like
sys-apps/module-init-tools already was removed as unnecessary later on.


   You want eudev removed, and Lennart Poettering wants udev on
non-systemd systems dropped.  Add those two items together, and we get
systemd rammed down our throats...

http://lists.freedesktop.org/archives/systemd-devel/2012-August/006066.html


(Yes, udev on non-systemd systems is in our eyes a dead end, in case
you haven't noticed it yet. I am looking forward to the day when we
can drop that support entirely.)




That might be the systemd upstream view point, but definately isn't mine.
Fact is that udev can be built and ran standalone without systemd and 
you don't need eudev for that.
If udev upstream makes it impossible to build, or run it standalone then 
we need to patch or fork it -- but that's far from now.
In any case there will always be sys-fs/udev and it will never require 
sys-apps/systemd.
Futhermore sys-fs/udev will be the default for long as sys-apps/openrc 
is the default.


I mean, why the heck fork something too early when upstream still 
supports udev on non-systemd init systems?!


- Samuli

Re: [gentoo-user] Freeze after suspend-to-ram with kernel 3.10

[Frank Steinmetzger]

On Fri, Aug 02, 2013 at 10:02:05AM -0500, Paul Hartman wrote:

> >> I tried to confirm my oberservation by deliberately hibernating it multiple
> >> times yesterday -- it always woke up with 3.9. Now I booted it with 3.10 
> >> and
> >> it didn't come up on the first try.
> >> Do you have any suggestion how I might debug this? I can’t simply report to
> >> kernel blokes “3.10 is crap on my netbook, you put in a regression 
> >> somewhere”.
> >>
> >> I don’t really have the time right now to go after hunches, such as the new
> >> tikless system, as building a kernel takes up to 45 minutes on that thing.
> > well, take your 3.9 config and don't change it.
> 
> If the config change doesn't reveal anything, you can do git bisect of
> the kernel to find out which patch broke it. When you do git bisect
> you don't need to recompile the whole kernel every time, it only
> compiles the changed files, which are usually not many. So even on a
> slower machine it's not so bad once the first compile is finished.
> 
> https://wiki.gentoo.org/wiki/Kernel_git-bisect

I didn't know one could have the git repo only for one subversion
(though I find it very handy for this huge codebase).

Anyhoo, I cleaned out my build dir and oldconfig'ed my 3.9 config, this
time saying no to everything new. As it turned out, the resulting config
was almost identical to my old 3.10 config (except for some crypto stuff
built as a module instead of built-in).
Now I don't have the freezes anymore. One possible reason is that I'm
now using 3.10.4 instead of 3.10.1. Another (more probable) reason is
that I may have forgotten to make clean the first time, which left over
some interfering cruft.
-- 
Gruß | Greetings | Qapla’
Please do not share anything from, with or about me with any Facebook service.

I am never drunk, I always keel over before that.


signature.asc
Description: Digital signature

Re: [gentoo-user] Moving from old udev to eudev

[Anthony G. Basile]

On 08/04/2013 11:56 AM, Dale wrote:

Anthony G. Basile wrote:


I have refrained from flamewars, but I want to reassure people, eudev
will not be dropped.



I noticed the other day, posted on this thread by the way, that it left
beta too.  I'm assuming you are involved in the project so allow me to
say this:  THANKS MUCH!!

Dale

:-)  :-)



I am the current lead.  You may follow the activity here [1].

[1] https://github.com/gentoo/eudev/commits/master

--
Anthony G. Basile, Ph. D.
Chair of Information Technology
D'Youville College
Buffalo, NY 14201
(716) 829-8197

Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.

[Mick]

On Monday 05 Aug 2013 07:06:08 gevisz wrote:
> My thanks to all who replied to my question.
> 
> The problem was with my local router, which I also used as DNS.
> After excluding it from /etc/resolv.config and /etc/init.d/net files,
> Firefox started to work as expected.

Hmm ... I wonder if this is related to my earlier comment about malformed 
packets.

May be worth trying a different firmware for this router.


> Suggestions of  Michael Kintzios

> > This is the new kernel naming scheme of NICs.  Which-ever nomenclature
> > you decide to use, check that that's the only one having a symlink in
> > /etc/init.d to net.lo
> 
> Yes, there is only enp2s15 links to lo in /etc/init.d

The idea here is that you need consistent naming of your iface.  If you have 
settled on the kernel naming of enp2s15, then stick with this throughout your 
configuration.


> After deleting all but my lan router DNS from /etc/conf.d/net and
> /etc/resolv.conf
> files, I had the same problem as before but in addition the host
> utility reports an
> additional error. Please, see the full response below.

You should not need to manually alter anything in your /etc/resolv.conf, which 
will be completed with the DNS server name(s) you have set up in your 
/etc/conf.d/net.


> # host www.google.com
> www.google.com has address 74.125.232.52
> www.google.com has address 74.125.232.48
> www.google.com has address 74.125.232.49
> www.google.com has address 74.125.232.50
> www.google.com has address 74.125.232.51
> ;; Warning: query response not set
> ;; Warning: query response not set

I think this means that the DNS server response is incorrectly formed (or that 
the server respond code does not include a 4 bit RCODE as it should - more 
detail for DNS geeks can be found here:  http://www.ietf.org/rfc/rfc2136.txt)


> Host www.google.com not found: 4(NOTIMP)

The RFC says:  The name server does not support the specified Opcode.  I would 
reflash the firmware, or try any OpenSource alternatives if available for your 
router.


> After leaving in /etc/conf.d/net and /etc/resolv.conf files only the
> DNS of my service
> provider, Firefox started to work as predicted. Thank you!

This may not be ideal (it will introduce some latency in your requests) but if 
you can't fix your router, it'll have to do for now.


> > Can you please show us:
> > ip route show
> > ip addr show
> > ip link show
> 
> $ ip route show
> default via 192.168.0.1 dev enp2s15  metric 2
> 127.0.0.0/8 via 127.0.0.1 dev lo  scope link
> 192.168.0.0/24 dev enp2s15  proto kernel  scope link  src 192.168.0.9

This says that your IP address us 192.168.0.9, but see below.


> $ ip addr show
[snip ...]

> 2: enp2s15:  mtu 1500 qdisc
> pfifo_fast state UP qlen 1000
> link/ether  brd ff:ff:ff:ff:ff:ff
> inet 192.168.0.7/24 brd 192.168.0.255 scope global enp2s15

This says that your ip address is 192.168.0.7 - did you get a different IP 
address between the two commands?  Your /etc/conf.d/net showed that you had 
set up a static address as config_enp2s15="192.168.0.9 ..."  so why is this 
here?


> $ ip link show
[snip ...]

> 2: enp2s15:  mtu 1500 qdisc
> pfifo_fast state UP mode DEFAULT qlen 1000
> link/ether  brd ff:ff:ff:ff:ff:ff

OK, this looks good.

 
> Suggestions of Kurian Thayil
> 
> > Can you do a ping and see if the resolv.conf DNS ips are reachable?
> 
> Yes, I can ping all my DNS. Moreover, I successfully use them from my
> Ubuntu installation on the same computer.
> 
> > do a
> > dig @8.8.8.8 www.google.com ## which will do a name resolution with
> > Google DNS servers.
> 
> Here is the output:
> 
> $ dig @8.8.8.8 www.google.co
> 
> ; <<>> DiG 9.9.2 <<>> @8.8.8.8 www.google.co
> ; (1 server found)
> ;; global options: +cmd
> ;; Got answer:
> ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 4036
> ;; flags: qr rd ra; QUERY: 1, ANSWER: 12, AUTHORITY: 4, ADDITIONAL: 5
> 
> ;; OPT PSEUDOSECTION:
> ; EDNS: version: 0, flags:; udp: 4096
> ;; QUESTION SECTION:
> ;www.google.co.   IN  A
> 
> ;; ANSWER SECTION:
> www.google.co.86400   IN  CNAME   www3.l.google.com.
> www3.l.google.com.13  IN  A   173.194.32.166
> www3.l.google.com.13  IN  A   173.194.32.167
> www3.l.google.com.13  IN  A   173.194.32.168
> www3.l.google.com.13  IN  A   173.194.32.169
> www3.l.google.com.13  IN  A   173.194.32.174
> www3.l.google.com.13  IN  A   173.194.32.160
> www3.l.google.com.13  IN  A   173.194.32.161
> www3.l.google.com.13  IN  A   173.194.32.162
> www3.l.google.com.13  IN  A   173.194.32.163
> www3.l.google.com.13  IN  A   173.194.32.164
> www3.l.google.com.13  IN  A   173.194.32.165
> 
> ;; AUTHORITY SECTION:
> google.com.   244594  IN  NS  ns3.google.com.
> google.com.   244594  IN  NS  ns2.google.com.
> google.com.   

[gentoo-user] layman issues

[András Csányi]

Good Morning All,

After update the layman repositories and portage tree I got this at
almost every package in every layman repository. Did I miss something
or I should inform the maintainer of the repository to do something to
the packages?

 Reading category  50|163 ( 30%): dev-util ..
Could not properly execute
/var/lib/layman/java/dev-util/visualvm/visualvm-1.3.3-r8.ebuild
 Reading category 151|163 ( 92%): www-servers .. * ERROR:
www-servers/tomcat-5.5.27-r4:: failed (depend phase):
 *   eutils.eclass could not be found by inherit()
 *
 * Call stack:
 * ebuild.sh, line 545:  Called source
'/var/lib/layman/java/www-servers/tomcat/tomcat-5.5.27-r4.ebuild'
 *   tomcat-5.5.27-r4.ebuild, line   9:  Called inherit 'eutils'
'java-pkg-2' 'java-ant-2' 'user'
 * ebuild.sh, line 257:  Called die
 * The specific snippet of code:
 *  [[ -z ${location} ]] && die "${1}.eclass could not be
found by inherit()"
 *
 * If you need support, post the output of `emerge --info
'=www-servers/tomcat-5.5.27-r4::'`,
 * the complete build log and the output of `emerge -pqv
'=www-servers/tomcat-5.5.27-r4::'`.
 * Working directory: '/usr/lib/portage/pym'
 * S: '/tomcat-5.5.27'

Thanks for any help in advance!

András

-- 
--  Csanyi Andras (Sayusi Ando)  -- http://sayusi.hu --
http://facebook.com/andras.csanyi
--  ""Trust in God and keep your gunpowder dry!" - Cromwell