Re: [gentoo-user] Re: [OT] How to be a penguin.

[R0b0t1]

Newfags. Newfags everywhere.

If you're going to fly by the seat of your pants you do it so other people
don't have to. Complaining [too] loudly is counterproduxtive.

[gentoo-user] kde-apps/libksane blocker

[Dale]

Howdy,

I'm trying to do a update and have this last blocker that I can't figure
out.  I fixed another one but this one doesn't make sense to me.  Here
is the relevant output.



[ebuild U ~] kde-apps/kde-l10n-16.04.1:5::gentoo [15.12.3:5::gentoo]
USE="handbook -debug" LINGUAS="-ar -ast% -bg -bs -ca -ca@valencia -cs
-da -de -el -en_GB -eo -es -et -eu -fa -fi -fr -ga -gl -he -hi -hr -hu
-ia -id -is -it -ja -kk -km -ko -lt -lv -mr -nb -nds -nl -nn -pa -pl -pt
-pt_BR -ro -ru -sk -sl -sr -sv -tr -ug -uk -wa -zh_CN -zh_TW" 0 KiB
[ebuild U ~] kde-apps/kde4-l10n-16.04.1:4/16.04::gentoo
[15.12.3-r1:4/15.12::gentoo] USE="handbook minimal (-aqua) {-test%}"
LINGUAS="-ar -ast% -bg -bs -ca -ca@valencia -cs -da -de -el -en_GB -eo
-es -et -eu -fa -fi -fr -ga -gl -he -hi -hr -hu -ia -id -is -it -ja -kk
-km -ko -lt -lv -mr -nb -nds -nl -nn -pa -pl -pt -pt_BR -ro -ru -sk -sl
-sr -sv -tr -ug -uk -wa -zh_CN -zh_TW" 0 KiB
[blocks b  ] =kde-apps/libksane-4.14.3:4[aqua=]
(>=kde-apps/libksane-4.14.3:4[-aqua]) required by
(kde-apps/ksaneplugin-16.04.1:4/16.04::gentoo, ebuild scheduled for merge)
>=kde-apps/libksane-4.14.3:4[aqua=]
(>=kde-apps/libksane-4.14.3:4[-aqua]) required by
(kde-apps/kolourpaint-16.04.1:4/16.04::gentoo, ebuild scheduled for merge)

  (kde-apps/libksane-16.04.1:5/5::gentoo, ebuild scheduled for merge)
pulled in by
>=kde-apps/libksane-14.12.0:5 required by
(kde-misc/skanlite-2.0:5/5::gentoo, ebuild scheduled for merge)
>=kde-apps/libksane-16.04.1 required by
(kde-apps/kdegraphics-meta-16.04.1:5/5::gentoo, ebuild scheduled for merge)


I'm not so worried about the procps thing right now, although it keeps
popping up.  The libksane one has me puzzled.  Here is some info on the
packages that are in question:


root@fireball / # equery list -p ksaneplugin skanlite kolourpaint
kdegraphics-meta
 * Searching for ksaneplugin ...
[-P-] [  ] kde-apps/ksaneplugin-4.14.3:4/4.14
[-P-] [  ] kde-apps/ksaneplugin-15.08.3:4/15.08
[IP-] [  ] kde-apps/ksaneplugin-15.12.3:4/15.12
[-P-] [  ] kde-apps/ksaneplugin-16.04.1:4/16.04

 * Searching for skanlite ...
[IP-] [  ] kde-misc/skanlite-1.1-r1:4
[-P-] [  ] kde-misc/skanlite-2.0:5

 * Searching for kolourpaint ...
[-P-] [  ] kde-apps/kolourpaint-4.14.3:4/4.14
[-P-] [  ] kde-apps/kolourpaint-15.08.3:4/15.08
[IP-] [  ] kde-apps/kolourpaint-15.12.3:4/15.12
[-P-] [  ] kde-apps/kolourpaint-16.04.1:4/16.04

 * Searching for kdegraphics-meta in kde-apps ...
[-P-] [  ] kde-apps/kdegraphics-meta-4.14.3:4
[-P-] [ ~] kde-apps/kdegraphics-meta-15.08.3:5
[IP-] [  ] kde-apps/kdegraphics-meta-15.12.3:5
[-P-] [  ] kde-apps/kdegraphics-meta-16.04.1:5
root@fireball / #
 

As you can see, none of them are keyworded so it should be able to use
the latest packages.  I just synced a hour or so ago so this may be more
up to date than you may have on your system.  It's why I am including
the info here. 

Is it possible I caught the tree in the middle of a update and I need to
sync again?  Is my old eyes missing something obvious?  Or, is it what I
think, a conflict that can't yet be solved because of a hard dependency? 

Looking for thoughts, ideas, suggestions. 

Dale

:-)  :-) 

Re: [gentoo-user] Re: Heads up perl-5.24 and virtual/

[covici]

walt  wrote:

> On Sun, 29 May 2016 12:33:55 +0200
> Alan McKinnon  wrote:
> 
> > > There's been another thread a week ago which mentioned a
> > > sys-devel/make-4.2 bug and the recommendation was to emerge perl
> > > with MAKEOPTS=-j1.  Did you try this?  
> > 
> > It's not perl failing to build, it's portage claiming it can't find a
> > clean upgrade path to offer. The merge doesn't start as the dep graph
> > resolution never completes, so MAKEOPTS isn't even in the picture yet.
> 
> I found that --with-bdeps=y fixed some problems.

I use the with-bdeps=y all the time and had nho real problems emerging
5.24.  It rebuilt about 200 packages and portage detected most of the
mess, but perlcleaner detected about 40 more and one I had to remove,
but it went pretty well for portage.  It seems to have gotten more
intelligent last few months, slow as heck, but I would rather use its
time than mine.

-- 
Your life is like a penny.  You're going to lose it.  The question is:
How do
you spend it?

 John Covici
 cov...@ccs.covici.com

[gentoo-user] Re: Heads up perl-5.24 and virtual/

[walt]

On Sun, 29 May 2016 12:33:55 +0200
Alan McKinnon  wrote:

> > There's been another thread a week ago which mentioned a
> > sys-devel/make-4.2 bug and the recommendation was to emerge perl
> > with MAKEOPTS=-j1.  Did you try this?  
> 
> It's not perl failing to build, it's portage claiming it can't find a
> clean upgrade path to offer. The merge doesn't start as the dep graph
> resolution never completes, so MAKEOPTS isn't even in the picture yet.

I found that --with-bdeps=y fixed some problems.

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Gregory Woodbury]

On Sun, May 29, 2016 at 8:13 AM, Alan Grimes  wrote:

> Gregory Woodbury wrote:
> >
> >
> > On Sat, May 28, 2016 at 4:53 PM, Neil Bothwick  > > wrote:
> >
> > On Sat, 28 May 2016 21:54:09 +0200, Volker Armin Hemmann wrote:
> >
> > > thanks a lot. My eyes are bleeding.
> >
> > Serves you right for being daft enough to read it again!
> >
> > I'd suggest that Alan RTFM for the commands he uses, but that
> > would be a
> > waste of keystrokes.
> >
> >
> > I have to agree with ng0
> >
> > WOW!
> >
> > Alan just wants to start it and walk away, as if Gentoo was a binary
> > distribution
> > that handles it all upstream.  He doesn't want to take the time to
> > review what
> > emerge is proposing and see if changes are needed first.
>
> You know what? fuck you.  That's what.
>

You really shouldn't say things like that. Some might consider it
harassment, and others
might take you up on the offer; you never can tell.


>
> The update list it's proposing is 403 packages, or roughly 25% of my
> system.
>

What the...?
Do you have every package there is installed?  The worst updates I have seen
are only 40-50 packages.  403 being 25% implies around 1600 - 1700 packages.
Actually that seems about right, but why are you getting that many updates?
You
may have maladjusted USE flags or keywords to make portage think that many
need an update.  I'm running with ~amd64 set and I check updates once a day.


> > Hey Alan: Gentoo is NOT a start an update and walk away setup. Some human
> > mind needs to be involved if troubles arise.  Also, read make.conf(5)
> > and set up
> > the various variables correctly; PYTHON_SINGLE_TARGET should only have
> > one python version set.
>
> DO YOU SERIOUSLY THINK I'M THAT STUPID??? SERIOUSLY
>

The output from the script that Dale re-posted clearly shows emerge
complaining
about PYTHON_SINGLE_TARGET being wrong for a dependency expression
in the ebuild involved.  That speaks for itself.

> Furthermore, the current portage doesn't require the revdep-rebuild
> > step because
> > of the @preserved-rebuild set creation.
>
> That missfeature is incompatible with how I use my system. I have not
> reformatted my hard drive in six years.
>

What does reformatting a hard drive have to do with it? I have an
installation that
has been running on one partition since 2011, when I had to build a new
computer
system after a flood.  It does look like I'm going to do a reinstall for
KDE5 Plasma
because too much KDE4 stuff hanging around seems to make the desktop
unstable.  It might take a few hours to do that.


> The principle way I accomplish that is by prohibiting the growth of
> cruft in the system. I cannot tolerate the accumulation of back versions
> anywhere in the system except where absolutely necessary. So if it is
> possible to re-build broken packages against new versions, I demand that
> take place
> as quickly as possible such that the system is left in the most pristine
> and self-consistent state possible. --- secret of immortality, dude. =\
>
> Gentoo used to be superlatively excellent at that.
>
> > In any case, to try and force things through without looking at what
> > problems are occuring
> > is just (excuse my language) batshit crazy stupid.
>
> You
> dumb.
> shit.
>
> You literally have no fucking clue do you?
>

You have no idea...
I have been using UNIX and UNIX-like systems since 1977. I've been involved
in UNIX and Linux for 40 years. I've written drivers, scripts, init systems
and lots
of other stuff, some of which is still hanging around yet.

I guess it tickles your fancy to tell people who know more than you that
they
don't have a "clue" when it is clear that you are ignoring the advice and
guidance
they are offering.

Others in this thread have offered advice and commentary, clearly explaining
why your methods are out-of-whack.  But I do like watching the logs and
builds
scroll by -- one does learn some interesting interactions occasionally.
[For example:
don't build virtualbox with more than -j2 because there is a missing
dependency in
the Makefile.]  For long builds, tailing or lessing the logfiles are easy,
and can be
done in another tab of the terminal program. I also have keep-going in my
default
Emerge options, along with --ask and --verbose; but when a build falters I
can
go look at the saved logfile and see exactly what went wrong, and usually
figure out
how to fix it. Most of the time is is a USE flag mixup, and occasionally a
missing
RDEPEND or DEPEND -- those get reported on Bugzilla.

I add a fair number of USE flags to the profile via make.conf, but some
don't belong
there, but rather in package.use.  I take advantage of the ability for
package.use to
be a directory, and place package specific USE flags in appropriately named
files
within that directory.  Interestingly, the KDE5 Plasma re-install (which is
almost done)
has fewer package specific flags than KDE4. [I prefer KDE and its
flexibility to the
"thi

Re: [gentoo-user] Re: Dolphin peculiarities with plasma5

[Mick]

On Sunday 29 May 2016 22:20:56 Alan McKinnon wrote:
> On 29/05/2016 20:46, Mick wrote:
> > On Sunday 29 May 2016 16:58:29 Mick wrote:
> >> On Monday 30 May 2016 01:09:26 Michael Palimaka wrote:
> >>> On 29/05/16 23:42, Mick wrote:
>  Following the latest round of KDE updates I noticed two undesirable
>  effects
>  and I would be grateful if you could point me to a way of reverting
>  them
>  to
>  their old behaviour:
>  
>  1. Konqueror is no longer integrated with dolphin as a file manager.
>  When
>  I click on Konqueror's Home Folder on its introductory page, a new
>  window
>  is launched with Dolphin.
> >>> 
> >>> I believe this is due to latest Konqueror being KDE4-based and latest
> >>> Dolphin (and therefore dolphinpart) being KF5-based. As a workaround,
> >>> try downgrading Dolphin to 4.14.3-r1.
> >>> 
>  2. Dolphin now requires a double click to open directories.  Where can
>  I
>  set this to a single click?  I've looked in dolphin Preferences and
>  systemsettings5 but I can't find where to change it to a single click
>  operation.
> >>> 
> >>> Try System Settings -> Input Devices -> Mouse -> Single-click to open
> >>> files and folders.
> >> 
> >> Thank you Michael, I already had this setting enabled.  Clicking on the
> >> arrow at the left of the directory expands the contents of the tree, but
> >> to
> >> get into it, to descend into that level on the filesystem, you have to
> >> double click on the directory itself.
> >> 
> >> Similarly, files do not open unless I double-click on them.  :-/
> > 
> > I've noticed that 'emerge --depclean' removed kde-apps/kdebase-kioslaves
> > and revdep-rebuild didn't bring anything else in (yet).  Could this be
> > why konqueror-dolphin integration is borked?
> 
> You might have a point there. I have kde-apps/kdebase-kioslaves
> installed here and my konqueror works as you want your's to. My deps for
> that package:
> 
> # equery depends kde-apps/kdebase-kioslaves
>  * These packages depend on kde-apps/kdebase-kioslaves:
> kde-apps/mplayerthumbs-15.12.3
> (>=kde-apps/kdebase-kioslaves-4.14.3:4[aqua=])
> media-gfx/digikam-4.14.0-r1 (kde-apps/kdebase-kioslaves:4)
> media-sound/amarok-2.8.90-r2 (>=kde-apps/kdebase-kioslaves-4.4:4[aqua=])
> 
> and I have these konquesror/dolphin packages installed:
> 
> kde-apps/konqueror
> kde-apps/konq-plugins
> kde-apps/libkonq
> kde-apps/dolphin
> kde-apps/dolphin-plugins

Thank you Alan, I just re-installed konq-plugins and then kdebase-kioslaves to 
see what happens.  Konqueror no longer opens dolphin in a separate window, nor 
does it open dolphin embedded in it as I want it.  Instead it opens FSView - 
showing colourful rectangles of sizes proportionate to the size of each file.  
Looking under Konqueror's menu/View/View Mode, I only have 'File Size View' 
available.  So I must be missing something else on this box.

I'm on stable versions:  

kde-apps/konqueror-15.12.3
kde-apps/dolphin-15.12.3-r1

Interestingly, the introductory page of Konqueror shows no images for the 
icons ... just empty placeholders.  Also, I just noticed something weird ... 
menu/Help/About Konqueror shows "Version 4.14.16" instead of 15.12.3.

Oh well, I am assuming KDE will get better over time. O_o

-- 
Regards,
Mick

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

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Volker Armin Hemmann]

Am 29.05.2016 um 22:15 schrieb Alan McKinnon:
> On 29/05/2016 19:54, Dale wrote:
>> Alan McKinnon wrote:
>>> On 29/05/2016 11:28, Dale wrote:
 Håkon Alstadheim wrote:
> Den 29. mai 2016 03:17, skrev Gregory Woodbury:
>> WOW!
>>
>> Alan just wants to start it and walk away, as if Gentoo was a binary
>> distribution
>> that handles it all upstream.  He doesn't want to take the time to
>> review what
>> emerge is proposing and see if changes are needed first.
>>
> It IS actually possible to do that, at least for non-critical systems.
> Just make sure to mail yourself with the emerge output, so you can fix
> things before the next automatic run.
>
> The point being that you don't have to sit and watch while emerge works.
> You can have the output of any blocks or failures waiting for you in
> your inbox at your convenience.
>
> For this to work in a timely fashion you need to stay on stable packages
> as much as possible, and also keep other customizations to a minimum,
> e.g. don't use --autounmask-write.
>
 Thing is, that's not what Alan seems to want.  Alan wants something like
 Ubuntu or something where you tell it to upgrade and then walk away
 without checking anything.
>>> You are both wrong. What Alan Grimes really wants is an excuse (any
>>> excuse) to whine, whinge and bitch about $STUFF.
>>>
>>> Notice how he never replies to any thread he starts?
>>>
>>>
>>
>> Well, he did reply.  Either his script is still doing it or he did and
>> he hasn't learned anything yet. 
>>
>> It's funny how he is the only one that has these problems and how he
>> keeps using that disaster of a script.  I don't think anyone has posted
>> a positive thing about that script.  I'm no script guru by any means but
>> even I can read that thing and see what a disaster it is. 
>>
>> Best of luck to him.  I'm about done trying to help.  Key word, trying.
>>
>> Dale
>>
>> :-)  :-) 
>>
>
> I agree Dale.
>
> Mr Grimes should probably move over to Linux From Scratch. It has none
> of portage's absurdities, makes no effort at all to be helpful to the
> user and allows anyone to write any build automation they feel is
> appropriate. LFS also requires you to watch all the compiler output all
> the time to catch problems; it all seems to match Mr Grimes'
> requirements right down to a tee.
>
>
> I'm going to STFU down, go to bed and finish an astonishingly intriguing
> Stephen King book and let the computer get on with doing whatever it
> thinks perl-cleaner fixes.
>

personally, I hope he stays around. His mails and the resulting threads
amuse me.

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Neil Bothwick]

On Sun, 29 May 2016 08:13:03 -0400, Alan Grimes wrote:

> You know what? fuck you.  That's what.

Helpful, really helpful.

> Packages are being updated at such a breakneck pace these days that it
> simply isn't humanly possible to review these manually, or even do
> anything intelligent if I tried.

Indeed, looking at the output of emerge -a before hitting enter is sooo
demanding.

> Then emerge
> got ornery and stopped letting the necessary, cathartic, inevitable,
> trainwreck take place, which is actually a good thing because the

You really think that breaking people's systems is a good thing?

> DO YOU SERIOUSLY THINK I'M THAT STUPID??? SERIOUSLY

I'll let others form their own opinions on this.

> > Furthermore, the current portage doesn't require the revdep-rebuild
> > step because
> > of the @preserved-rebuild set creation.  
> 
> That missfeature is incompatible with how I use my system. I have not
> reformatted my hard drive in six years.

Only six years? I am using installations that are older than that, even
if the hard drives aren't.

> The principle way I accomplish that is by prohibiting the growth of
> cruft in the system. I cannot tolerate the accumulation of back versions
> anywhere in the system except where absolutely necessary. So if it is
> possible to re-build broken packages against new versions, I demand that
> take place
> as quickly as possible such that the system is left in the most pristine
> and self-consistent state possible. --- secret of immortality, dude. =\

So a clean but broken system is preferable to having a few older
versions of libraries hanging around for half and hour, until they can be
safely, and automatically, removed?

Anyway, it seems you are well out of date in your understanding of how
portage works. preserved-rebuild is rarely needed these days, thanks to
the mysteries of subslots, which can do the necessary rebuilding at the
point of upgrade.

> You
> dumb.
> shit.
> 
> You literally have no fucking clue do you?

You seem to be echoing the thoughts of so many others here.
 
> Do you think I enabled that missfeature they introduced a few years ago
> that hid all of the build output so all I got to see was
> 
> installing package (1/400)
> installing package (2/400)
> installing package (3/400)

It's an options, use it or don't use it. If you really want your updates
to take longer just so you can see all the GCC output, which makes
little sense anyway when MAKEOPTS is set to anything but -j1.
Personally, I prefer the Unix approach of succeed quietly, fail noisily.
If I see compiler output on my monitor, i know something is wrong.
 
> I am typing this on my smaller monitor because I have the full verbose
> build process fullscreen on my 24". That's right, for the last 12 years,
> I have watched every single build take place in live time because I've
> watched it execute every single compiler invocation, I have watched
> every error and warning message.

What a sad life you must have. Watching compiler output is even more
tedious than reading Twitter!
 
> I do not need log files because I watch everything in live time.

And you retain everything you see and never need to refer back to what
scrolled off the screen ten minutes ago?

> By
> doing this, I have learned things about my packages that you fucking
> dipshits couldn't imagine. Back before when Gentoo jumped the shark
> (tried to force everyone onto libav), the system was completely
> self-correcting, If the system set was intact, literally every other
> problem would self correct. I didn't need to rant on this list because
> everything was perfect.

> The problem is that the portage people don't understand how the packages
> actually work

So there's the problem, all the Gentoo devs are dipshits, as well as all
of us. No wonder Gentoo is such a pile of steaming shite. I'm completely
flabbergasted that you still consider it worth using.

> It wouldn't let me do that because it was throwing a conflict message
> for libreoffice so I was forced to temporarily uninstall it to clear a
> block message, fuck you again portage...

Have you read the emerge and portage man pages anytime in the last five
years?

> IQ is a measure of how stupid you feel.

What measures how stupid you act?


-- 
Neil Bothwick

What do you have when you have six lawyers buried up to their necks in
sand? Not enough sand.


pgpHEeDWphQcf.pgp
Description: OpenPGP digital signature

Re: [gentoo-user] Re: Dolphin peculiarities with plasma5

[Alan McKinnon]

On 29/05/2016 20:46, Mick wrote:
> On Sunday 29 May 2016 16:58:29 Mick wrote:
>> On Monday 30 May 2016 01:09:26 Michael Palimaka wrote:
>>> On 29/05/16 23:42, Mick wrote:
 Following the latest round of KDE updates I noticed two undesirable
 effects
 and I would be grateful if you could point me to a way of reverting them
 to
 their old behaviour:

 1. Konqueror is no longer integrated with dolphin as a file manager. 
 When
 I click on Konqueror's Home Folder on its introductory page, a new
 window
 is launched with Dolphin.
>>>
>>> I believe this is due to latest Konqueror being KDE4-based and latest
>>> Dolphin (and therefore dolphinpart) being KF5-based. As a workaround,
>>> try downgrading Dolphin to 4.14.3-r1.
>>>
 2. Dolphin now requires a double click to open directories.  Where can I
 set this to a single click?  I've looked in dolphin Preferences and
 systemsettings5 but I can't find where to change it to a single click
 operation.
>>>
>>> Try System Settings -> Input Devices -> Mouse -> Single-click to open
>>> files and folders.
>>
>> Thank you Michael, I already had this setting enabled.  Clicking on the
>> arrow at the left of the directory expands the contents of the tree, but to
>> get into it, to descend into that level on the filesystem, you have to
>> double click on the directory itself.
>>
>> Similarly, files do not open unless I double-click on them.  :-/
> 
> I've noticed that 'emerge --depclean' removed kde-apps/kdebase-kioslaves and 
> revdep-rebuild didn't bring anything else in (yet).  Could this be why 
> konqueror-dolphin integration is borked?
> 

You might have a point there. I have kde-apps/kdebase-kioslaves
installed here and my konqueror works as you want your's to. My deps for
that package:

# equery depends kde-apps/kdebase-kioslaves
 * These packages depend on kde-apps/kdebase-kioslaves:
kde-apps/mplayerthumbs-15.12.3
(>=kde-apps/kdebase-kioslaves-4.14.3:4[aqua=])
media-gfx/digikam-4.14.0-r1 (kde-apps/kdebase-kioslaves:4)
media-sound/amarok-2.8.90-r2 (>=kde-apps/kdebase-kioslaves-4.4:4[aqua=])

and I have these konquesror/dolphin packages installed:

kde-apps/konqueror
kde-apps/konq-plugins
kde-apps/libkonq
kde-apps/dolphin
kde-apps/dolphin-plugins

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

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Alan McKinnon]

On 29/05/2016 19:54, Dale wrote:
> Alan McKinnon wrote:
>> On 29/05/2016 11:28, Dale wrote:
>>> Håkon Alstadheim wrote:
 Den 29. mai 2016 03:17, skrev Gregory Woodbury:
> WOW!
>
> Alan just wants to start it and walk away, as if Gentoo was a binary
> distribution
> that handles it all upstream.  He doesn't want to take the time to
> review what
> emerge is proposing and see if changes are needed first.
>
 It IS actually possible to do that, at least for non-critical systems.
 Just make sure to mail yourself with the emerge output, so you can fix
 things before the next automatic run.

 The point being that you don't have to sit and watch while emerge works.
 You can have the output of any blocks or failures waiting for you in
 your inbox at your convenience.

 For this to work in a timely fashion you need to stay on stable packages
 as much as possible, and also keep other customizations to a minimum,
 e.g. don't use --autounmask-write.

>>> Thing is, that's not what Alan seems to want.  Alan wants something like
>>> Ubuntu or something where you tell it to upgrade and then walk away
>>> without checking anything.
>>
>> You are both wrong. What Alan Grimes really wants is an excuse (any
>> excuse) to whine, whinge and bitch about $STUFF.
>>
>> Notice how he never replies to any thread he starts?
>>
>>
> 
> 
> Well, he did reply.  Either his script is still doing it or he did and
> he hasn't learned anything yet. 
> 
> It's funny how he is the only one that has these problems and how he
> keeps using that disaster of a script.  I don't think anyone has posted
> a positive thing about that script.  I'm no script guru by any means but
> even I can read that thing and see what a disaster it is. 
> 
> Best of luck to him.  I'm about done trying to help.  Key word, trying.
> 
> Dale
> 
> :-)  :-) 
> 


I agree Dale.

Mr Grimes should probably move over to Linux From Scratch. It has none
of portage's absurdities, makes no effort at all to be helpful to the
user and allows anyone to write any build automation they feel is
appropriate. LFS also requires you to watch all the compiler output all
the time to catch problems; it all seems to match Mr Grimes'
requirements right down to a tee.


I'm going to STFU down, go to bed and finish an astonishingly intriguing
Stephen King book and let the computer get on with doing whatever it
thinks perl-cleaner fixes.

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

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Dale]

»Q« wrote:
> On Sun, 29 May 2016 08:13:03 -0400
> Alan Grimes  wrote:
>
>> You know what? fuck you.  That's what.
> It's unclear whether you're attacking only Gregory or the entire
> list.  If the latter, would you consider unsubscribing?  It looks as
> though no one her is interested in helping you as long as you continue
> using the "jackhammer", so the only result of you posting here is that
> the replies make you more angry than your Gentoo problems alone did.
> (To be clear:  I certainly don't speak for the entire list -- this is
> just one guy's suggestion that you consider it.)
>

In my opinion, if he is not going to take advice from other users here,
why post to begin with?  If he isn't going to post, may as well
unsubscribe and be done with it.  He's not helping anyone else with his
posts and he doesn't take the help that has been provided numerous
times.  He's not giving or taking so what is the point? 

So, I 2nd the motion.  lol 

In all honesty, I wish he would go file bug reports on BGO since he
thinks everything Gentoo is broken.  I suspect a few devs would give him
a ear full and at some point, block him from filing bugs at all.  I
seriously doubt the devs would put up with even a small part of this mess. 

Dale

:-)  :-) 

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Dale]

Daniel Frey wrote:
> On 05/29/2016 02:20 AM, Dale wrote:
>> Neil Bothwick wrote:
>>> I have a weekly system health check cron job that includes revdep-rebuild
>>> -pi (hint to Alan: that's the correct way to have revdep-rebuild ignore
>>> the results of previous runs). It rarely finds anything. There's still
>>> the occasional glitch with preserved-libs, but I fons it works
>>> ninety-lots % of the time, and it is far better than the "let it break
>>> then try to fix it approach" of the days we needed to rely on
>>> revdep-rebuild.
>>>
>>>
>>
>> I haven't ran revdep-rebuild in likely over a year.  Just for giggles, I
>> ran it a bit ago.  The only thing it found was libreoffice.  That's not
>> exactly a critical package or anything.  Given that, I don't think it
>> really serves any point. 
>>
>> I wonder if me having backtrack set to 100 helps with that?  Of course,
>> unlike poor Alan, I also have a sane approach to upgrading.  I also run
>> the latest non- version of portage. 
>>
>> Dale
>>
>> :-)  :-) 
>>
>>
> I should correct myself, I always run revdep-rebuild after a depclean.
> The point it, portage doesn't catch everything, regardless of its
> importance or not.
>
> Dan
>
>

It seems to here.  Like I said, I haven't ran it in over a year and all
it found was libreoffice.  Everything else was fine.  I don't run
--depclean after every upgrade but do after either a big update or every
other update, whichever comes first. 

Dale

:-)  :-) 

Re: [gentoo-user] Re: Dolphin peculiarities with plasma5

[Mick]

On Sunday 29 May 2016 16:58:29 Mick wrote:
> On Monday 30 May 2016 01:09:26 Michael Palimaka wrote:
> > On 29/05/16 23:42, Mick wrote:
> > > Following the latest round of KDE updates I noticed two undesirable
> > > effects
> > > and I would be grateful if you could point me to a way of reverting them
> > > to
> > > their old behaviour:
> > > 
> > > 1. Konqueror is no longer integrated with dolphin as a file manager. 
> > > When
> > > I click on Konqueror's Home Folder on its introductory page, a new
> > > window
> > > is launched with Dolphin.
> > 
> > I believe this is due to latest Konqueror being KDE4-based and latest
> > Dolphin (and therefore dolphinpart) being KF5-based. As a workaround,
> > try downgrading Dolphin to 4.14.3-r1.
> > 
> > > 2. Dolphin now requires a double click to open directories.  Where can I
> > > set this to a single click?  I've looked in dolphin Preferences and
> > > systemsettings5 but I can't find where to change it to a single click
> > > operation.
> > 
> > Try System Settings -> Input Devices -> Mouse -> Single-click to open
> > files and folders.
> 
> Thank you Michael, I already had this setting enabled.  Clicking on the
> arrow at the left of the directory expands the contents of the tree, but to
> get into it, to descend into that level on the filesystem, you have to
> double click on the directory itself.
> 
> Similarly, files do not open unless I double-click on them.  :-/

I've noticed that 'emerge --depclean' removed kde-apps/kdebase-kioslaves and 
revdep-rebuild didn't bring anything else in (yet).  Could this be why 
konqueror-dolphin integration is borked?

-- 
Regards,
Mick

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

[gentoo-user] Re: [OT] How to be a penguin.

[»Q«]

On Sat, 28 May 2016 15:10:12 -0500
Dale  wrote:

> Volker Armin Hemmann wrote:

[ "jackhammer" script ]
> > thanks a lot. My eyes are bleeding.  
> 
> Well, I did warn you to skip it. 
> 
> I wonder, does he ever have a upgrade that goes smoothly?

Not only does he never have a smooth upgrade, he can't remember a time
when it took less than a week to recover from everything broken by the
"jackhammer".  (Where "recover" doesn't mean things are actually fixed,
just that he's stopped portage from giving him error messages until
the next upgrade "ordeal".)

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Dale]

Alan McKinnon wrote:
> On 29/05/2016 11:28, Dale wrote:
>> Håkon Alstadheim wrote:
>>> Den 29. mai 2016 03:17, skrev Gregory Woodbury:
 WOW!

 Alan just wants to start it and walk away, as if Gentoo was a binary
 distribution
 that handles it all upstream.  He doesn't want to take the time to
 review what
 emerge is proposing and see if changes are needed first.

>>> It IS actually possible to do that, at least for non-critical systems.
>>> Just make sure to mail yourself with the emerge output, so you can fix
>>> things before the next automatic run.
>>>
>>> The point being that you don't have to sit and watch while emerge works.
>>> You can have the output of any blocks or failures waiting for you in
>>> your inbox at your convenience.
>>>
>>> For this to work in a timely fashion you need to stay on stable packages
>>> as much as possible, and also keep other customizations to a minimum,
>>> e.g. don't use --autounmask-write.
>>>
>> Thing is, that's not what Alan seems to want.  Alan wants something like
>> Ubuntu or something where you tell it to upgrade and then walk away
>> without checking anything.
>
> You are both wrong. What Alan Grimes really wants is an excuse (any
> excuse) to whine, whinge and bitch about $STUFF.
>
> Notice how he never replies to any thread he starts?
>
>


Well, he did reply.  Either his script is still doing it or he did and
he hasn't learned anything yet. 

It's funny how he is the only one that has these problems and how he
keeps using that disaster of a script.  I don't think anyone has posted
a positive thing about that script.  I'm no script guru by any means but
even I can read that thing and see what a disaster it is. 

Best of luck to him.  I'm about done trying to help.  Key word, trying.

Dale

:-)  :-) 

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Dale]

Alan Grimes wrote:
> Gregory Woodbury wrote:
>>
>> On Sat, May 28, 2016 at 4:53 PM, Neil Bothwick > > wrote:
>>
>> On Sat, 28 May 2016 21:54:09 +0200, Volker Armin Hemmann wrote:
>>
>> > thanks a lot. My eyes are bleeding.
>>
>> Serves you right for being daft enough to read it again!
>>
>> I'd suggest that Alan RTFM for the commands he uses, but that
>> would be a
>> waste of keystrokes.
>>
>>
>> I have to agree with ng0
>>
>> WOW!
>>
>> Alan just wants to start it and walk away, as if Gentoo was a binary
>> distribution
>> that handles it all upstream.  He doesn't want to take the time to
>> review what
>> emerge is proposing and see if changes are needed first.
> You know what? fuck you.  That's what.
>
> The update list it's proposing is 403 packages, or roughly 25% of my
> system.
>
> Packages are being updated at such a breakneck pace these days that it
> simply isn't humanly possible to review these manually, or even do
> anything intelligent if I tried. Back in the golden age, for about ten
> years even! my approach to updating my system worked great. Then emerge
> got ornery and stopped letting the necessary, cathartic, inevitable,
> trainwreck take place, which is actually a good thing because the
> partial-good, update which seems nightmarish on first analysis,
> ***ACTUALLY CORRECTS ITSELF WHEN THE SCRIPT IS RUN REPEATEDLY UNTIL NO
> PROBLEMS REMAIN AND THE SYSTEM IS PRISTINE AND GOOD FOR REBOOT***. I
> have done this happily many many many many times. It actually works that
> way and I was gleefully singing gentoo's praise for many years.
>
> No, the penguins seem to think it is possible to get it perfect on the
> first iteration. It's not. It's not just that there are a handful of
> packages that just won't work no matter what, it's just that it is
> sometimes necessary to let a bootstrapping process take place. This
> actually works if you don't use such draconian checking.

So, you still think it is emerge that is the problem and not your script
which everyone that has seen it says is the wrong way to do a update?  ROFL

I did a emerge -e  world not to long ago.  Out of the over 1,000
packages on here, I only had one failure.  Just one. 

>> Hey Alan: Gentoo is NOT a start an update and walk away setup. Some human 
>> mind needs to be involved if troubles arise.  Also, read make.conf(5)
>> and set up 
>> the various variables correctly; PYTHON_SINGLE_TARGET should only have
>> one python version set.
> DO YOU SERIOUSLY THINK I'M THAT STUPID??? SERIOUSLY

Do you really want a answer to that?  Seriously?  You asked so I guess
you do.  Here it is.  Yes!  lol  The way you do things, and continue to
do things, even after having several VERY experienced users tell you
that the way you are doing things is wrong pretty much says it all.  If
I post that I was doing something and getting a bad result and someone
such as Neil and/or Alan McKinnon tells me I am doing it wrong, you can
bet your last dollar that I am about to change how I do things.  Why, if
either or both of them tell me I am doing something wrong, I'm doing it
wrong.  If I continue to do things the same way, well, that would be my
fault not members of this list or emerge/portage.  So far, I don't
recall seeing a single post that supports how your script is set up. 
Not one, except you of course. 

>
>> Furthermore, the current portage doesn't require the revdep-rebuild
>> step because
>> of the @preserved-rebuild set creation.
> That missfeature is incompatible with how I use my system. I have not
> reformatted my hard drive in six years.
>
> The principle way I accomplish that is by prohibiting the growth of
> cruft in the system. I cannot tolerate the accumulation of back versions
> anywhere in the system except where absolutely necessary. So if it is
> possible to re-build broken packages against new versions, I demand that
> take place
> as quickly as possible such that the system is left in the most pristine
> and self-consistent state possible. --- secret of immortality, dude. =\
>
> Gentoo used to be superlatively excellent at that.
>
>> In any case, to try and force things through without looking at what
>> problems are occuring
>> is just (excuse my language) batshit crazy stupid.
> You
> dumb.
> shit.
>
> You literally have no fucking clue do you?

Thing is, it would seem to me that others on this list do have a clue
and have tried several times to explain that to you.  The person who
doesn't have a clue is the one that comes here complaining about
problems that no one else here has. 


>
> Do you think I enabled that missfeature they introduced a few years ago
> that hid all of the build output so all I got to see was
>
> installing package (1/400)
> installing package (2/400)
> installing package (3/400)
>
> I am typing this on my smaller monitor because I have the full verbose
> build process fullscreen on my 24". That's right, for the last 12 years,
> I have watched e

[gentoo-user] Re: [OT] How to be a penguin.

[»Q«]

On Sun, 29 May 2016 08:13:03 -0400
Alan Grimes  wrote:

> You know what? fuck you.  That's what.

It's unclear whether you're attacking only Gregory or the entire
list.  If the latter, would you consider unsubscribing?  It looks as
though no one her is interested in helping you as long as you continue
using the "jackhammer", so the only result of you posting here is that
the replies make you more angry than your Gentoo problems alone did.
(To be clear:  I certainly don't speak for the entire list -- this is
just one guy's suggestion that you consider it.)

[gentoo-user] Re: How to be a penguin.

[Grant Edwards]

On 2016-05-28, Alan Grimes  wrote:

[plonk]

-- 
Grant

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Daniel Frey]

On 05/29/2016 02:20 AM, Dale wrote:
> Neil Bothwick wrote:
>> I have a weekly system health check cron job that includes revdep-rebuild
>> -pi (hint to Alan: that's the correct way to have revdep-rebuild ignore
>> the results of previous runs). It rarely finds anything. There's still
>> the occasional glitch with preserved-libs, but I fons it works
>> ninety-lots % of the time, and it is far better than the "let it break
>> then try to fix it approach" of the days we needed to rely on
>> revdep-rebuild.
>>
>>
> 
> 
> I haven't ran revdep-rebuild in likely over a year.  Just for giggles, I
> ran it a bit ago.  The only thing it found was libreoffice.  That's not
> exactly a critical package or anything.  Given that, I don't think it
> really serves any point. 
> 
> I wonder if me having backtrack set to 100 helps with that?  Of course,
> unlike poor Alan, I also have a sane approach to upgrading.  I also run
> the latest non- version of portage. 
> 
> Dale
> 
> :-)  :-) 
> 
> 

I should correct myself, I always run revdep-rebuild after a depclean.
The point it, portage doesn't catch everything, regardless of its
importance or not.

Dan

Re: [gentoo-user] Re: Dolphin peculiarities with plasma5

[Mick]

On Monday 30 May 2016 01:09:26 Michael Palimaka wrote:
> On 29/05/16 23:42, Mick wrote:
> > Following the latest round of KDE updates I noticed two undesirable
> > effects
> > and I would be grateful if you could point me to a way of reverting them
> > to
> > their old behaviour:
> > 
> > 1. Konqueror is no longer integrated with dolphin as a file manager.  When
> > I click on Konqueror's Home Folder on its introductory page, a new window
> > is launched with Dolphin.
> 
> I believe this is due to latest Konqueror being KDE4-based and latest
> Dolphin (and therefore dolphinpart) being KF5-based. As a workaround,
> try downgrading Dolphin to 4.14.3-r1.
> 
> > 2. Dolphin now requires a double click to open directories.  Where can I
> > set this to a single click?  I've looked in dolphin Preferences and
> > systemsettings5 but I can't find where to change it to a single click
> > operation.
> 
> Try System Settings -> Input Devices -> Mouse -> Single-click to open
> files and folders.

Thank you Michael, I already had this setting enabled.  Clicking on the arrow 
at the left of the directory expands the contents of the tree, but to get into 
it, to descend into that level on the filesystem, you have to double click on 
the directory itself.

Similarly, files do not open unless I double-click on them.  :-/

-- 
Regards,
Mick

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

[gentoo-user] Re: Dolphin peculiarities with plasma5

[Michael Palimaka]

On 29/05/16 23:42, Mick wrote:
> Following the latest round of KDE updates I noticed two undesirable effects 
> and I would be grateful if you could point me to a way of reverting them to 
> their old behaviour:
> 
> 1. Konqueror is no longer integrated with dolphin as a file manager.  When I 
> click on Konqueror's Home Folder on its introductory page, a new window is 
> launched with Dolphin.

I believe this is due to latest Konqueror being KDE4-based and latest
Dolphin (and therefore dolphinpart) being KF5-based. As a workaround,
try downgrading Dolphin to 4.14.3-r1.

> 2. Dolphin now requires a double click to open directories.  Where can I set 
> this to a single click?  I've looked in dolphin Preferences and 
> systemsettings5 but I can't find where to change it to a single click 
> operation. 
> 

Try System Settings -> Input Devices -> Mouse -> Single-click to open
files and folders.

Re: [gentoo-user] Re: undefined symbol "...[abi:cxx11]" (was: Do we have any Blender build experts in the house?)

[J.]

El dom, 29-05-2016 a las 08:35 -0600, J. García escribió:
> El dom, 29-05-2016 a las 09:02 +0200, David Haller escribió:
> > 
> > Yes, BUT...
> > 
> > a) use c++filt / nm -C (see below)
> > 
> > b) Won't help unless you actually recompile/emerge openexr with
> >    -std=c++11 or -std=gnu++11 (or -std={c,gnu}++14 or greater).
> >    reemerging as per default with -std=c++98 or less (what's the
> >    default std for c++ in g++?) won't help.
> > 
> Nice and informative post(I didn't know nm), I didn't said put 
> -std=c++11 in make.conf, or package.env, because, the news[1] item I
> pointed to in my first post says, quote
> "... GCC 5 uses the new C++ ABI by default." 
> I also said:
> > 
> > ... you might need to rebuild that and make sure(read the log) g++
> >  is being called with std=c++11
> So if you are using gcc:5 you don't need to specify the ABI, in the
> portage configuration files, to use the new one.
missed this:
[1]https://www.gentoo.org/support/news-items/2015-10-22-gcc-5-new-c++11
-abi.html

Re: [gentoo-user] Re: undefined symbol "...[abi:cxx11]" (was: Do we have any Blender build experts in the house?)

[J.]

El dom, 29-05-2016 a las 09:02 +0200, David Haller escribió:
> Yes, BUT...
> 
> a) use c++filt / nm -C (see below)
> 
> b) Won't help unless you actually recompile/emerge openexr with
>    -std=c++11 or -std=gnu++11 (or -std={c,gnu}++14 or greater).
>    reemerging as per default with -std=c++98 or less (what's the
>    default std for c++ in g++?) won't help.
> 
Nice and informative post(I didn't know nm), I didn't said put 
-std=c++11 in make.conf, or package.env, because, the news[1] item I
pointed to in my first post says, quote
"... GCC 5 uses the new C++ ABI by default." 
I also said:
> ... you might need to rebuild that and make sure(read the log) g++
>  is being called with std=c++11
So if you are using gcc:5 you don't need to specify the ABI, in the
portage configuration files, to use the new one.

Re: [gentoo-user] Do we have any Blender build experts in the house?

[Todd Goodman]

* Andrew Lowe  [160528 13:59]:
> Hi all,
>   I'm attempting to build Blender, 2,7,6 but am running into a problem 
> during linking. I am getting a series of errors referring to an, for 
> example:
> 
> unresolved reference to 'Imf_2_1::Header::view[abi:cxx11]() const
> 
> There are a whole series of these errors all referencing undefined 
> things contained within a library called "Imf_2_1". I've done some 
> googling and I think this is something to do with a library from 
> Industrial Light & Magic but other than that, I can't work things out. 
> Has anyone come across this sort of problem before I have to post the 
> whole gory details of this problem.
> 
>   Any thoughts greatly appreciated,
>   Andrew

When I've seen errors like that in the past my first step has been to
make sure all my libraries have been rebuilt with the new GCC.

Don't know if this is your issue and could be completely off, but a
recent news entry is:

2015-10-22-gcc-5-new-c++11-abi
  Title GCC 5 Defaults to the New C++11 ABI
  AuthorMike Frysinger 
  Posted2015-10-22
  Revision  2

  GCC 5 uses the new C++ ABI by default.  When building new code,
  you might run
  into link time errors that include lines similar to:
  ...: undefined reference to
  '_ZNSt6chrono12steady_clock3nowEv@GLIBCXX_3.4.17'

  Or you might see linkage failures with "std::__cxx11::string" in
  the output.

  These are signs that you need to rebuild packages using the new
  C++ ABI.
  You can quickly do so by using revdep-rebuild (from gentoolkit).

  For gentoolkit-0.3.1 or higher:
  # revdep-rebuild --library 'libstdc++.so.6' -- --exclude gcc

  For previous versions of gentoolkit:
  # revdep-rebuild --library 'libstdc\+\+\.so\.6' -- --exclude gcc

  For more details, feel free to peruse:
  https://developerblog.redhat.com/2015/02/05/gcc5-and-the-c11-abi/
  
https://blogs.gentoo.org/blueness/2015/03/10/the-c11-abi-incompatibility-problem-in-gentoo/

[gentoo-user] Dolphin peculiarities with plasma5

[Mick]

Following the latest round of KDE updates I noticed two undesirable effects 
and I would be grateful if you could point me to a way of reverting them to 
their old behaviour:

1. Konqueror is no longer integrated with dolphin as a file manager.  When I 
click on Konqueror's Home Folder on its introductory page, a new window is 
launched with Dolphin.

2. Dolphin now requires a double click to open directories.  Where can I set 
this to a single click?  I've looked in dolphin Preferences and 
systemsettings5 but I can't find where to change it to a single click 
operation. 

-- 
Regards,
Mick

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

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Alan Grimes]

Neil Bothwick wrote:
> On Sat, 28 May 2016 20:48:37 -0700, Daniel Frey wrote:
>
>>> Furthermore, the current portage doesn't require the revdep-rebuild
>>> step because
>>> of the @preserved-rebuild set creation.  
>> I beg to differ, portage still misses stuff more often than you think. I
>> always run revdep-rebuild after an emerge.
> I have a weekly system health check cron job that includes revdep-rebuild
> -pi (hint to Alan: that's the correct way to have revdep-rebuild ignore
> the results of previous runs). It rarely finds anything. There's still
> the occasional glitch with preserved-libs, but I fons it works
> ninety-lots % of the time, and it is far better than the "let it break
> then try to fix it approach" of the days we needed to rely on
> revdep-rebuild.

I respectfully disagree, my migraine quotient was much lower under that
approach than with what's going on now. =|

-- 
IQ is a measure of how stupid you feel.

Powers are not rights.

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Alan Grimes]

Gregory Woodbury wrote:
>
>
> On Sat, May 28, 2016 at 4:53 PM, Neil Bothwick  > wrote:
>
> On Sat, 28 May 2016 21:54:09 +0200, Volker Armin Hemmann wrote:
>
> > thanks a lot. My eyes are bleeding.
>
> Serves you right for being daft enough to read it again!
>
> I'd suggest that Alan RTFM for the commands he uses, but that
> would be a
> waste of keystrokes.
>
>
> I have to agree with ng0
>
> WOW!
>
> Alan just wants to start it and walk away, as if Gentoo was a binary
> distribution
> that handles it all upstream.  He doesn't want to take the time to
> review what
> emerge is proposing and see if changes are needed first.

You know what? fuck you.  That's what.

The update list it's proposing is 403 packages, or roughly 25% of my
system.

Packages are being updated at such a breakneck pace these days that it
simply isn't humanly possible to review these manually, or even do
anything intelligent if I tried. Back in the golden age, for about ten
years even! my approach to updating my system worked great. Then emerge
got ornery and stopped letting the necessary, cathartic, inevitable,
trainwreck take place, which is actually a good thing because the
partial-good, update which seems nightmarish on first analysis,
***ACTUALLY CORRECTS ITSELF WHEN THE SCRIPT IS RUN REPEATEDLY UNTIL NO
PROBLEMS REMAIN AND THE SYSTEM IS PRISTINE AND GOOD FOR REBOOT***. I
have done this happily many many many many times. It actually works that
way and I was gleefully singing gentoo's praise for many years.

No, the penguins seem to think it is possible to get it perfect on the
first iteration. It's not. It's not just that there are a handful of
packages that just won't work no matter what, it's just that it is
sometimes necessary to let a bootstrapping process take place. This
actually works if you don't use such draconian checking.

> Hey Alan: Gentoo is NOT a start an update and walk away setup. Some human 
> mind needs to be involved if troubles arise.  Also, read make.conf(5)
> and set up 
> the various variables correctly; PYTHON_SINGLE_TARGET should only have
> one python version set.

DO YOU SERIOUSLY THINK I'M THAT STUPID??? SERIOUSLY

> Furthermore, the current portage doesn't require the revdep-rebuild
> step because
> of the @preserved-rebuild set creation.

That missfeature is incompatible with how I use my system. I have not
reformatted my hard drive in six years.

The principle way I accomplish that is by prohibiting the growth of
cruft in the system. I cannot tolerate the accumulation of back versions
anywhere in the system except where absolutely necessary. So if it is
possible to re-build broken packages against new versions, I demand that
take place
as quickly as possible such that the system is left in the most pristine
and self-consistent state possible. --- secret of immortality, dude. =\

Gentoo used to be superlatively excellent at that.

> In any case, to try and force things through without looking at what
> problems are occuring
> is just (excuse my language) batshit crazy stupid.

You
dumb.
shit.

You literally have no fucking clue do you?

Do you think I enabled that missfeature they introduced a few years ago
that hid all of the build output so all I got to see was

installing package (1/400)
installing package (2/400)
installing package (3/400)

I am typing this on my smaller monitor because I have the full verbose
build process fullscreen on my 24". That's right, for the last 12 years,
I have watched every single build take place in live time because I've
watched it execute every single compiler invocation, I have watched
every error and warning message.

I do not need log files because I watch everything in live time. By
doing this, I have learned things about my packages that you fucking
dipshits couldn't imagine. Back before when Gentoo jumped the shark
(tried to force everyone onto libav), the system was completely
self-correcting, If the system set was intact, literally every other
problem would self correct. I didn't need to rant on this list because
everything was perfect.

The problem is that the portage people don't understand how the packages
actually work or what happens when these stupid, recently implemented
checks are disabled and emerge is run iteratively.

> I use my update generator script so make the emerge command(s) just so
> I can preview the packages and modify the sequences or leave out some
> updates
> if i need/want to do so. (E.g. I may want to defer a chromium or
> libreoffice update
> to after other updates are done and/or set them to occur with a lower
> niceness or
> an ionice idle class.)

It wouldn't let me do that because it was throwing a conflict message
for libreoffice so I was forced to temporarily uninstall it to clear a
block message, fuck you again portage...


-- 
IQ is a measure of how stupid you feel.

Powers are not rights.

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Neil Bothwick]

On Sun, 29 May 2016 12:26:08 +0200, Alan McKinnon wrote:

> You are both wrong. What Alan Grimes really wants is an excuse (any
> excuse) to whine, whinge and bitch about $STUFF.
> 
> Notice how he never replies to any thread he starts?

The script isn't that intelligent yet. ${DEITY} help us when it gains
that feature, although it will probably be based on repeating some
commands an arbitrary number of times than something sensible like a loop.


-- 
Neil Bothwick

Never argue with an idiot.
First, they bring you down to their level.
Then they beat you with experience.


pgpBiQWCg3Khm.pgp
Description: OpenPGP digital signature

Re: [gentoo-user] Heads up perl-5.24 and virtual/

[Alan McKinnon]

On 29/05/2016 10:13, Mick wrote:
> On Sunday 29 May 2016 09:40:10 Alan McKinnon wrote:
>> Heads up to any ~arch users who might run into this.
>>
>> I've just spent too many annoying hours dealing with perl-5.24 and it's
>> modules. As usual with recent perl upgrades, many modules have moved
>> around and now there's a whole whack of new virtuals. One or more are
>> causing problems like this:
>>
>> !!! Multiple package instances within a single package slot have been pulled
>> !!! into the dependency graph, resulting in a slot conflict:
>>
>> dev-lang/perl:0
>>
>>   (dev-lang/perl-5.22.2:0/5.22::gentoo, installed) pulled in by
>> =dev-lang/perl-5.22* required by
>> (virtual/perl-NEXT-0.650.0-r4:0/0::perl-experimental, installed)
>> ^  ^
>> dev-lang/perl:0/5.22=[-build(-)] required by
>> (perl-gcpan/GnuPG-0.19:0/0::splog-musicbrainz-mirror, installed)
>>  
>> (and 435 more with the same problems)
>>
>>   (dev-lang/perl-5.24.0:0/5.24::gentoo, ebuild scheduled for merge)
>> pulled in by
>> =dev-lang/perl-5.24* required by
>> (virtual/perl-Exporter-5.720.0-r1:0/0::gentoo, installed)
>> ^  ^
>> (and 59 more with the same problem)
>>
>> and portage bails out.
>> BUT IT'S NOT A HARD BLOCKER.
>>
>> I've given up trying to figure this out. Masking perl-5.24 and running
>> emerge with "--backtrack=99" gets portage back into a state where it's
>> willing to continue.
>>
>> Perhaps someone else with more patience can figure this one out.
> 
> Are you saying that the usual incantation of perl-cleaner, dep-clean and 
> @preserved rebuild will not arrive at a working system?

perl-cleaner won't help here. It was run when perl-5.22 was installed,
and everything was fixed up just hunky-dory.

perl-5.24 isn't installed yet so there's nothing to clean
> 
> There's been another thread a week ago which mentioned a sys-devel/make-4.2 
> bug and the recommendation was to emerge perl with MAKEOPTS=-j1.  Did you try 
> this?

It's not perl failing to build, it's portage claiming it can't find a
clean upgrade path to offer. The merge doesn't start as the dep graph
resolution never completes, so MAKEOPTS isn't even in the picture yet.

Basically, some modules have been made ready for 5.24 and other's
haven't. Normally what happens in this case is portage sees you still
need stuff that uses 5.22 so decides to not upgrade perl itself, leaves
all the modules at the current version, and gets on with doing the rest
of the non-perl upgrades.

With this upgrade, emerge itself fails implying that at least one perl
module is preventing the system from staying at 5.22 for $reasons.

But when I explicity mask 5.24 set set backtrack=99, portage searches
and searches and searches and eventually finds a way to complete and
still be consistent so the problem can't be a hard blocker. So something
is terribly non-optimum about portage's search paths though all these
perl modules.


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

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Alan McKinnon]

On 29/05/2016 11:28, Dale wrote:
> Håkon Alstadheim wrote:
>> Den 29. mai 2016 03:17, skrev Gregory Woodbury:
>>> WOW!
>>>
>>> Alan just wants to start it and walk away, as if Gentoo was a binary
>>> distribution
>>> that handles it all upstream.  He doesn't want to take the time to
>>> review what
>>> emerge is proposing and see if changes are needed first.
>>>
>> It IS actually possible to do that, at least for non-critical systems.
>> Just make sure to mail yourself with the emerge output, so you can fix
>> things before the next automatic run.
>>
>> The point being that you don't have to sit and watch while emerge works.
>> You can have the output of any blocks or failures waiting for you in
>> your inbox at your convenience.
>>
>> For this to work in a timely fashion you need to stay on stable packages
>> as much as possible, and also keep other customizations to a minimum,
>> e.g. don't use --autounmask-write.
>>
> 
> Thing is, that's not what Alan seems to want.  Alan wants something like
> Ubuntu or something where you tell it to upgrade and then walk away
> without checking anything.


You are both wrong. What Alan Grimes really wants is an excuse (any
excuse) to whine, whinge and bitch about $STUFF.

Notice how he never replies to any thread he starts?


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

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Dale]

Håkon Alstadheim wrote:
> Den 29. mai 2016 03:17, skrev Gregory Woodbury:
>> WOW!
>>
>> Alan just wants to start it and walk away, as if Gentoo was a binary
>> distribution
>> that handles it all upstream.  He doesn't want to take the time to
>> review what
>> emerge is proposing and see if changes are needed first.
>>
> It IS actually possible to do that, at least for non-critical systems.
> Just make sure to mail yourself with the emerge output, so you can fix
> things before the next automatic run.
>
> The point being that you don't have to sit and watch while emerge works.
> You can have the output of any blocks or failures waiting for you in
> your inbox at your convenience.
>
> For this to work in a timely fashion you need to stay on stable packages
> as much as possible, and also keep other customizations to a minimum,
> e.g. don't use --autounmask-write.
>

Thing is, that's not what Alan seems to want.  Alan wants something like
Ubuntu or something where you tell it to upgrade and then walk away
without checking anything.  Yea, you can script some things like
syncing, emailing a -p output to yourself and such but all that requires
a break in the process and Alan doesn't seem to get that part.  Alan
wants to type in one command and not put any effort into checking what
will be done or needs changing. 

I run a desktop here and it's not like my life or anything depends on
it.  There is no way I would do blind updates on this system for any
reason.  If I'm to busy to do my part, I wait until another time to do
the updates. 

Dale

:-)  :-) 

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Dale]

Neil Bothwick wrote:
> On Sat, 28 May 2016 20:48:37 -0700, Daniel Frey wrote:
>
>>> Furthermore, the current portage doesn't require the revdep-rebuild
>>> step because
>>> of the @preserved-rebuild set creation.  
>> I beg to differ, portage still misses stuff more often than you think. I
>> always run revdep-rebuild after an emerge.
> I have a weekly system health check cron job that includes revdep-rebuild
> -pi (hint to Alan: that's the correct way to have revdep-rebuild ignore
> the results of previous runs). It rarely finds anything. There's still
> the occasional glitch with preserved-libs, but I fons it works
> ninety-lots % of the time, and it is far better than the "let it break
> then try to fix it approach" of the days we needed to rely on
> revdep-rebuild.
>
>


I haven't ran revdep-rebuild in likely over a year.  Just for giggles, I
ran it a bit ago.  The only thing it found was libreoffice.  That's not
exactly a critical package or anything.  Given that, I don't think it
really serves any point. 

I wonder if me having backtrack set to 100 helps with that?  Of course,
unlike poor Alan, I also have a sane approach to upgrading.  I also run
the latest non- version of portage. 

Dale

:-)  :-) 

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Neil Bothwick]

On Sat, 28 May 2016 20:48:37 -0700, Daniel Frey wrote:

> > Furthermore, the current portage doesn't require the revdep-rebuild
> > step because
> > of the @preserved-rebuild set creation.  
> 
> I beg to differ, portage still misses stuff more often than you think. I
> always run revdep-rebuild after an emerge.

I have a weekly system health check cron job that includes revdep-rebuild
-pi (hint to Alan: that's the correct way to have revdep-rebuild ignore
the results of previous runs). It rarely finds anything. There's still
the occasional glitch with preserved-libs, but I fons it works
ninety-lots % of the time, and it is far better than the "let it break
then try to fix it approach" of the days we needed to rely on
revdep-rebuild.


-- 
Neil Bothwick

God is real, unless specifically declared integer.


pgpI2N7U7khek.pgp
Description: OpenPGP digital signature

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Peter Humphrey]

On Sunday 29 May 2016 09:45:57 I wrote:
> On Saturday 28 May 2016 12:06:20 Dale wrote:
> > What is also a surprise, Alan hasn't figured out that his script is the
> > problem.
> 
> It's called stubbornness, obduracy, obstinacy, wilful self-flagellation -
> choose your pick (as my Dad used to say when the rock cakes came out).

I missed the obvious one: pig-headedness.

-- 
Rgds
Peter

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Peter Humphrey]

On Saturday 28 May 2016 12:06:20 Dale wrote:

> What is also a surprise, Alan hasn't figured out that his script is the
> problem.

It's called stubbornness, obduracy, obstinacy, wilful self-flagellation - 
choose your pick (as my Dad used to say when the rock cakes came out).

-- 
Rgds
Peter

Re: [gentoo-user] Re: [OT] How to be a penguin.

[Håkon Alstadheim]

Den 29. mai 2016 03:17, skrev Gregory Woodbury:
>
> WOW!
>
> Alan just wants to start it and walk away, as if Gentoo was a binary
> distribution
> that handles it all upstream.  He doesn't want to take the time to
> review what
> emerge is proposing and see if changes are needed first.
>
It IS actually possible to do that, at least for non-critical systems.
Just make sure to mail yourself with the emerge output, so you can fix
things before the next automatic run.

The point being that you don't have to sit and watch while emerge works.
You can have the output of any blocks or failures waiting for you in
your inbox at your convenience.

For this to work in a timely fashion you need to stay on stable packages
as much as possible, and also keep other customizations to a minimum,
e.g. don't use --autounmask-write.

Re: [gentoo-user] Heads up perl-5.24 and virtual/

[Mick]

On Sunday 29 May 2016 09:40:10 Alan McKinnon wrote:
> Heads up to any ~arch users who might run into this.
> 
> I've just spent too many annoying hours dealing with perl-5.24 and it's
> modules. As usual with recent perl upgrades, many modules have moved
> around and now there's a whole whack of new virtuals. One or more are
> causing problems like this:
> 
> !!! Multiple package instances within a single package slot have been pulled
> !!! into the dependency graph, resulting in a slot conflict:
> 
> dev-lang/perl:0
> 
>   (dev-lang/perl-5.22.2:0/5.22::gentoo, installed) pulled in by
> =dev-lang/perl-5.22* required by
> (virtual/perl-NEXT-0.650.0-r4:0/0::perl-experimental, installed)
> ^  ^
> dev-lang/perl:0/5.22=[-build(-)] required by
> (perl-gcpan/GnuPG-0.19:0/0::splog-musicbrainz-mirror, installed)
>  
> (and 435 more with the same problems)
> 
>   (dev-lang/perl-5.24.0:0/5.24::gentoo, ebuild scheduled for merge)
> pulled in by
> =dev-lang/perl-5.24* required by
> (virtual/perl-Exporter-5.720.0-r1:0/0::gentoo, installed)
> ^  ^
> (and 59 more with the same problem)
> 
> and portage bails out.
> BUT IT'S NOT A HARD BLOCKER.
> 
> I've given up trying to figure this out. Masking perl-5.24 and running
> emerge with "--backtrack=99" gets portage back into a state where it's
> willing to continue.
> 
> Perhaps someone else with more patience can figure this one out.

Are you saying that the usual incantation of perl-cleaner, dep-clean and 
@preserved rebuild will not arrive at a working system?

There's been another thread a week ago which mentioned a sys-devel/make-4.2 
bug and the recommendation was to emerge perl with MAKEOPTS=-j1.  Did you try 
this?

-- 
Regards,
Mick

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

[gentoo-user] Heads up perl-5.24 and virtual/

[Alan McKinnon]

Heads up to any ~arch users who might run into this.

I've just spent too many annoying hours dealing with perl-5.24 and it's
modules. As usual with recent perl upgrades, many modules have moved
around and now there's a whole whack of new virtuals. One or more are
causing problems like this:

!!! Multiple package instances within a single package slot have been pulled
!!! into the dependency graph, resulting in a slot conflict:

dev-lang/perl:0

  (dev-lang/perl-5.22.2:0/5.22::gentoo, installed) pulled in by
=dev-lang/perl-5.22* required by
(virtual/perl-NEXT-0.650.0-r4:0/0::perl-experimental, installed)
^  ^
dev-lang/perl:0/5.22=[-build(-)] required by
(perl-gcpan/GnuPG-0.19:0/0::splog-musicbrainz-mirror, installed)
 
(and 435 more with the same problems)

  (dev-lang/perl-5.24.0:0/5.24::gentoo, ebuild scheduled for merge)
pulled in by
=dev-lang/perl-5.24* required by
(virtual/perl-Exporter-5.720.0-r1:0/0::gentoo, installed)
^  ^
(and 59 more with the same problem)

and portage bails out.
BUT IT'S NOT A HARD BLOCKER.

I've given up trying to figure this out. Masking perl-5.24 and running
emerge with "--backtrack=99" gets portage back into a state where it's
willing to continue.

Perhaps someone else with more patience can figure this one out.


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

Re: [gentoo-user] kde-base/baloo is blocking an upgrade

[Mick]

On Sunday 29 May 2016 08:27:50 Alan McKinnon wrote:

> Mick, have you re-synced the tree and updated just portage in the last
> day? I'd do that first as all indications are that your setup isn't
> wrong and it should work.

I've resync'ed again this morning, just in case this box was caught on the 
hop.


> You might have to go through your original emerge output and set
> USE="minimal" for all kde4 packages that support it and hwere there's an
> equivalent frameworks package.

Ha!  What do you know!  I've re-emerged kde-base/baloo with USE="minimal" and 
then tried again to emerge world.  It is now emerging 6 out 144 packages 
without a Blocker!

Thank you all for your help.  :-)

-- 
Regards,
Mick

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

[gentoo-user] Re: undefined symbol "...[abi:cxx11]" (was: Do we have any Blender build experts in the house?)

[David Haller]

Hello,

First of all, this has nothing to do with blender "per se" ...

On Sat, 28 May 2016, J. García wrote:
>El sáb, 28-05-2016 a las 17:26 -0600, J. García escribió:
>> BTW, eix won't help you here, I tried it, so the next step is to find
>> what provides the symbol 'Imf_2_1::Header::view', after some searches
>> with find and then google, I arrived at opencv(though not 100% sure),
>> so you might need to rebuild that and make sure(read the log) g++ is
>> being called with std=c++11. tracking this sort of stuff might be
>> tricky.
>I was left with the doubt if opencv was the package causing trouble,
>and it is not, so I searched a bit more in depth, and I have found the
>library you need to rebuild, it is libIlmImf.so, and this time I'm
>sure.
>
>Looking for the package:
>
>$ qfile /usr/lib64/libIlmImf.so 
>media-libs/openexr (/usr/lib64/libIlmImf.so)

*Hah* Learned something new, I've overlooked qfile so far :)

>Looking for the symbol: 
>
>$ readelf -s /usr/lib64/libIlmImf.so  | grep   '.*Header.*view.*'
>  1617: 0004a35025 FUNCGLOBAL DEFAULT   10
>_ZNK7Imf_2_16Header4viewB
>  2248: 00049e6025 FUNCGLOBAL DEFAULT   10
>_ZN7Imf_2_16Header4viewB5
>
>So you need to: emerge --oneshot media-libs/openexr

Yes, BUT...

a) use c++filt / nm -C (see below)

b) Won't help unless you actually recompile/emerge openexr with
   -std=c++11 or -std=gnu++11 (or -std={c,gnu}++14 or greater).
   reemerging as per default with -std=c++98 or less (what's the
   default std for c++ in g++?) won't help.

Generally: for any symbol containing '[abi:cxxNN]' you need that lib
compiled with the CXX flag '-std=c++NN'. As far as I can see, the
symbol 'foo[abi:cxxNN]' also provides the plain symbol 'foo', so it's
downwards compatible.

Ok, you got a symbol + [abi:cxxNN]. Usually, the first part of the
symbol gives you a hint as to what lib could be involved. In this case
some lib with *Imf* (Symbols Q* are usually Qt stuff ;). Searching
online usually helps if you've no idea, but use c++filt too if the
symbol seems weird (see below). But in this case with "Imf" ...:

$ ls /usr/lib64/*Imf*
/usr/lib64/libIlmImf-Imf_2_1.so.21  /usr/lib64/libIlmImf.so
/usr/lib64/libIlmImf-Imf_2_1.so.21.0.0

$ equery belongs /usr/lib64/libIlmImf.so 
 * Searching for /usr/lib64/libIlmImf.so ... 
media-libs/openexr-2.1.0 (/usr/lib64/libIlmImf.so -> 
libIlmImf-Imf_2_1.so.21.0.0)
media-libs/openexr-2.1.0 (/usr/lib64/libIlmImf-Imf_2_1.so.21.0.0)

$ nm /usr/lib64/libIlmImf.so | grep 'Imf_2_1.*Header.*view'
00049a40 T _ZN7Imf_2_16Header12previewImageEv
0003d5dc t 
_ZN7Imf_2_16Header14typedAttributeINS_14TypedAttributeINS_12PreviewImageERT_PKc.part.46
00049980 T _ZN7Imf_2_16Header15setPreviewImageERKNS_12PreviewImageE
000492c0 T _ZN7Imf_2_16Header4viewB5cxx11Ev
00049a80 T _ZNK7Imf_2_16Header12previewImageEv
0003d5dc t 
_ZNK7Imf_2_16Header14typedAttributeINS_14TypedAttributeINS_12PreviewImageERKT_PKc.part.51
00049ac0 T _ZNK7Imf_2_16Header15hasPreviewImageEv
00049300 T _ZNK7Imf_2_16Header4viewB5cxx11Ev
$ nm -C /usr/lib64/libIlmImf.so | grep 'Imf_2_1::Header::view'
000492c0 T Imf_2_1::Header::view[abi:cxx11]()
00049300 T Imf_2_1::Header::view[abi:cxx11]() const
$ nm /usr/lib64/libIlmImf.so | c++filt | grep 'Imf_2_1::Header::view'
000492c0 T Imf_2_1::Header::view[abi:cxx11]()
00049300 T Imf_2_1::Header::view[abi:cxx11]() const

Ok, nm does not work on stripped stuff. But strings does:

$ strip /usr/lib64/libIlmImf.so
$ nm /usr/lib64/libIlmImf.so 
nm: /usr/lib64/libIlmImf.so: no symbols
$ strings /usr/lib64/libIlmImf.so | c++filt | grep 'Imf_2_1::Header::view'
Imf_2_1::Header::view[abi:cxx11]()
Imf_2_1::Header::view[abi:cxx11]() const

And so does readelf[4]:

$ readelf -sW /usr/lib64/libIlmImf.so | c++filt | grep 'Imf_2_1::Header::view'
  1565: 0004930064 FUNCGLOBAL DEFAULT   10 
Imf_2_1::Header::view[abi:cxx11]() const
  2178: 000492c064 FUNCGLOBAL DEFAULT   10 
Imf_2_1::Header::view[abi:cxx11]()

(so, I already _DO_ have libIlmImf compiled with -std=c++11 or
-std=gnu++11, so I'm fine, but you're missing the '[abi:cxx11]', so
you need to recompile with -std=c++11/-std=gnu++11).

That's because I "default" by now to c++11 (and the extra CXXSTD var
is due to me starting to switch that (back and forth) before realising
the below).

$ grep CXX /etc/portage/make.conf
CXXSTD=" -std=c++11"
CXXFLAGS="$CFLAGS ${CXXSTD} "

BUT!

Some packages do not compile (yet) with c++11, and still need
c++98/gnu++98.

And some even seem to already need c++14 (mkvtoolnix). *Woah* After
realizing this and finding the stuff about the env-files (I love
gentoo for stuff like this!), I thought a moment... And ...

So, I've got a couple of env-files and use them for the packages that
need it.

First of all: I set the CPU-dependend stuff (-march, -O* etc. _AND
ONLY THAT_ (see below)) in CFLAGS in make.conf, then combine CFLAGS