Re: [gentoo-user] Error with infinality font while emerging sane-backends

[Danny YUE]

On 2018-03-28 14:55, Floyd Anderson <f...@31c0.net> wrote:
> On Wed, 28 Mar 2018 13:58:21 +0800
> Danny YUE <sheepd...@gmail.com> wrote:
>>
>>
>>Thanks for your quick reply :-)
>>
>>I created the symbolic link as you told,
>
> That solves the:
>
>   Fontconfig error: failed reading config file
>   Fontconfig error: Cannot load config file "infinality/conf.d"
>
> errors – they’re gone.
>
>>but it still does not
>>work...same error message.
>
> Now you have to solve the:
>
>   Error: /invalidfont in /findfont
>
>   Last OS error: No such file or directory
>   GPL Ghostscript 9.21: Error: /invalidfont in /findfont
>
> errors. The culprit seems here the Type-1 font selection, see bug
> #620148 [1]. Commenting the mentioned section in:
>
>/etc/fonts/infinality/infinality.conf
>
> works here, bumping Ghostscript to version 9.23 doesn’t.
>
> Hope that helps. If not, you probably can temporarily rename:
>
>/usr/bin/fig2dev (package media-gfx/transfig)
>
> to hide it from the media-gfx/sane-backends build process, so it wont build
> the PDF documentation. Previously sane-backends-1.0.27 compiles fine without
> media-gfx/xfig and therefore media-gfx/transfig.
>
>
> Reference:
>  [1] <https://bugs.gentoo.org/show_bug.cgi?id=620148>

Oops, I did not really see this email last week...my stupidity.

Thank you so much Floyd, you solved my problem perfectly.

After commenting out the section, the font problem never appeared any
more.

However the PDF file still refuses to build.
I knew the solution: disable parallel compilation feature.

After that it worked.

Thanks again!

Danny

Re: [gentoo-user] Error with infinality font while emerging sane-backends

[Danny YUE]

On 2018-03-28 04:34, Floyd Anderson <f...@31c0.net> wrote:
> On Wed, 28 Mar 2018 11:17:59 +0800
> Danny YUE <sheepd...@gmail.com> wrote:
>>Hi folks,
>>
>>I am trying to emerge xsane, which requires sane-backends.
>>However, the sane-backends package fails forever with error message like
>>below:
>>--- BEGIN ---
>>Fontconfig error: Cannot load config file "infinality/conf.d"
>>Error: /invalidfont in /findfont
>>--- END ---
>>
>>I have only "52-infinality.conf" enabled in "eselect fontconfig".
>>I suppose this is a problem related to infinality fontconfig I am using.
>>But how am I supposed to fix this?
>
> That is probably because media-libs/fontconfig seems not to canonicalise
> relative target path for symlinks.
>
> As a quick, but volatile, solution you can re-create symlink:
>
>ln -s /etc/fonts/infinality/styles.conf.avail/linux \
>  /etc/fonts/infinality/conf.d
>
> or you have to patch the module:
>
>/usr/share/eselect/modules/infinality.eselect
>
> of app-eselect/eselect-infinality-1 to use full qualified paths.
>
> Test configuration loading afterwards, e.g. with `fc-match monospace`.
> ---
> infinality.eselect | 5 +++--
> 1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/infinality.eselect b/infinality.eselect
> index 8f397dc..8b069ae 100644
> --- a/infinality.eselect
> +++ b/infinality.eselect
> @@ -1,5 +1,5 @@
> # -*-eselect-*-  vim: ft=eselect
> -# Copyright 2005-2012 Gentoo Foundation
> +# Copyright 2005-2018 Gentoo Foundation
> # Distributed under the terms of the GNU GPL version 2 or later
> #
> # Original author: MeisterP <pon...@spahan.ch>
> @@ -34,7 +34,8 @@ set_symlink() {
> [[ -z ${target} || ! -d 
> ${EROOT}/etc/fonts/infinality/styles.conf.avail/${target} ]] \
> && die -q "Target \"$1\" doesn't appear to be valid!"
>
> -ln -s "styles.conf.avail/${target}" 
> "${EROOT}/etc/fonts/infinality/conf.d"
> +ln -s "${EROOT}/etc/fonts/infinality/styles.conf.avail/${target}" \
> +"${EROOT}/etc/fonts/infinality/conf.d"
> echo "Selected style: ${target}"
> echo "You should set eselect lcdfilter to match your current style"
> }

Thanks for your quick reply :-)

I created the symbolic link as you told, but it still does not
work...same error message.

--- BEGIN ---
~ $ ll /etc/fonts/infinality/
total 20k
drwxr-xr-x   4 root   root 4.0k 2018-03-28 12:43 .
drwxr-xr-x   5 root   root 4.0k 2018-03-24 12:17 ..
lrwxrwxrwx   1 root   root   45 2018-03-28 12:43 conf.d -> 
/etc/fonts/infinality/styles.conf.avail/linux
drwxr-xr-x   3 root   root 4.0k 2017-11-11  2017 conf.src
-rw-r--r--   1 root   root 3.7k 2017-11-11  2017 infinality.conf
drwxr-xr-x  11 root   root 4.0k 2017-11-11  2017 styles.conf.avail
--- END ---

https://pastebin.com/DGuzvmNx

[gentoo-user] Error with infinality font while emerging sane-backends

[Danny YUE]

Hi folks,

I am trying to emerge xsane, which requires sane-backends.
However, the sane-backends package fails forever with error message like
below:
--- BEGIN ---
Fontconfig error: Cannot load config file "infinality/conf.d"
Error: /invalidfont in /findfont
--- END ---

I have only "52-infinality.conf" enabled in "eselect fontconfig".
I suppose this is a problem related to infinality fontconfig I am using.
But how am I supposed to fix this?
I googled a lot but no result. :-(

The full build log is here: https://pastebin.com/8fjnqmHw
emerge --info: https://pastebin.com/aharCut1

Thanks in advance for any potential help. :-)

Danny

Re: [gentoo-user] pdftk - replacement

[Danny YUE]

On 2018-03-14 20:08, Poncho  wrote:
> On 14.03.2018 20:10, the...@sys-concept.com wrote:
>> Is there a suitable replacement package for "pdftk".
>> Currently pdftk is masked and might be dropped from portage tree  :-/
>> 
>> The packages require GCC 5 or earlier and are therefore masked in the
>> 17.0 profiles.
>> 
>> I use pdftk only to merge two pdf file (overlap tax form).
>> eg.
>> pdftk t4-flat-02b.pdf stamp  1.pdf  output out1.pdf
>> 
>
> /usr/bin/pdfunite
>
> it's a provided by app-text/poppler

Besides, app-text/poppler contains some very useful functions.
pdfunite is one of them, which let you concatenate several PDF files
into one.

Another piece I found useful is pdfseparate. It can separate a PDF file
into pages.

The combination of pdfunite and pdfseparate let you replace a single
page in a PDF file.

Command `equery f poppler' shows:
  /usr/bin/pdfdetach
  /usr/bin/pdffonts
  /usr/bin/pdfimages
  /usr/bin/pdfinfo
  /usr/bin/pdfseparate
  /usr/bin/pdftocairo
  /usr/bin/pdftohtml
  /usr/bin/pdftoppm
  /usr/bin/pdftops
  /usr/bin/pdftotext
  /usr/bin/pdfunite

Re: [gentoo-user] What do you think about Firefox 57?

[Danny YUE]

On 2017-09-11 01:19, Daniel Campbell <z...@gentoo.org> wrote:
> On 09/07/2017 05:26 AM, Danny YUE wrote:
>> Hi all,
>> 
>> I have been using FoxyProxy in Firefox for a really long time, until
>> today I found its new version really sucks.
>> 
>> Then I read the comment from author who declared that the old version
>> can *only* be used before (roughly) end of 2017 before Firefox 57 and in
>> new version some features must perish.
>> 
>> Afterwards I found that it seems Firefox 57 will use a new ecosystem for
>> extensions and be more strict for plugin developers.
>> 
>> So Firefox gurus, what do you think about it?
>> 
>> 
>> Danny
>> 
> 
> I switched to Pale Moon a while ago, though I suspect fewer and fewer
> mainstream sites will work with it as devs will begin requiring features
> enabled in newer Firefox and Chrome (e.g. WebRTC, EME, localStorage,
> etc). GitHub has already dropped support for Pale Moon, despite PM
> supporting just about everything GitHub makes use of.
>
> Losing XUL may be great from a security standpoint, but the feature-set
> is lacking, it negatively impacts performance (no cache sharing,
> blockers can't block correctly without a full render prior) and it all
> reeks of a code merge. Why else would Mozilla be putting all this work
> into looking *and* acting like Chrome? This behavior is that of a
> company that is looking to get out of the market. They've already
> abandoned their phone OS and their e-mail/calendar client. Firefox is
> just the final nail in the coffin. Servo isn't up to snuff yet, and the
> power users that gave Firefox its popularity are (like me) disinterested
> in what passes for "modern Web". Many websites are flat-out malicious,
> and more are insecure in general, largely due to feature creep in the
> browser. Without the ability to protect yourself, it becomes a risky
> decision to continue browsing a space filled with surveillance and
> malware. In short, it's a dumpster fire. Like all grim scenarios,
> however, there are sites out there that don't abuse people. But that
> number is dwindling every day.
>
> Aside from that, the hard requirement on PulseAudio is another strike
> against it, and their culture wrt diversity is off-putting. Mozilla
> isn't the Web leader it once was. To its credit, I don't think any
> organization is "leading" the Web well. With the W3C approving DRM as a
> standard in HTTP, it indicates a corporate acquisition of the standards
> body, and it's no longer fit for purpose. We need a browser that is
> opinionated and sticks to the standards that make sense, and hands
> control of media to other programs. That would severely simplify the
> browser, and leverage software that's generally already on a computer.
> Web browsers as they are are fine for netbooks, which have little in the
> way of system software. But for desktop machines, at least, most things
> can be handed to a media player, PDF viewer, etc. The code's already
> there: there are handlers for different protocols like irc:, mailto:,
> torrent:, etc. Adding handlers via MIME-type would be fine.
>
> As it is, I already don't read much on the Web. The experience has
> become crap, even with blocking extensions. More trouble than it's
> worth, most of the time. I have better things to do than endlessly tweak
> my privacy just so sites don't slurp up all the metadata they can on my
> connection. uBO, Privacy Badger, uMatrix, and others are great -- huge
> jumps in quality compared to their predecessors -- but the rampant
> misuse of the medium leaves me disinterested in the Web.
>
> So few websites these days are designed with graceful degradation in
> mind, let alone accessibility. It's all ECMAscript bells and whistles,
> web "apps", etc. to the point where you have two systems: your Gentoo
> system and your Web browser. I try to reduce complexity where possible,
> balanced against safety. That leads me to an upstream who won't screw
> with my interface and disrupt the add-on ecosystem because "this is
> better for you".
>
> Based on what I've read so far, Moonchild is up front about any
> breakage, and warns about unsupported compilers or settings. One of our
> regulars (Walter Dnes) helps maintain PM for us, too, so that's even
> better. :)
>
> But to be fair, I'll try it out when 57 is released so I have a stronger
> opinion. I suspect I will be let down.
> 

Such a long response, thank you Daniel.

I don't know if adding DRM into HTTP protocol is a good idea.
Maybe it does help reduce spreading of pirate, but HTTP then somehow
works beyond "transfer".

Personally speaking, I prefer to be able to pick software in a grand
market, instead of integrate everything into one big monster with
security/privacy holes.

I would like to try 57 also (with old Firefox profile backup).


Danny

Re: [gentoo-user] What do you think about Firefox 57?

[Danny YUE]

On 2017-09-08 00:48, Adam Carter <adamcart...@gmail.com> wrote:
> On Thu, Sep 7, 2017 at 10:26 PM, Danny YUE <sheepd...@gmail.com> wrote:
>
>> Hi all,
>>
>> I have been using FoxyProxy in Firefox for a really long time, until
>> today I found its new version really sucks.
>>
>> Then I read the comment from author who declared that the old version
>> can *only* be used before (roughly) end of 2017 before Firefox 57 and in
>> new version some features must perish.
>>
>
> I am assuming that most actively developed addons will be updated before 57
> drops, so the transition wont be too painful. Is the issue you're having
> with FoxyProxy related to it relying on features that are no longer
> available under the new framework?

Yes. The FoxyProxy author declared that some features are no long
available under the new framework.
I am really depressed about it because I just use that plugin every time
I use Firefox.


> This update will break my favorite extension, Vimperator. I am already
> looking at replacements for Firefox due to this. So far I have found
> qutebrowser which isn't really as featureful.
>
> R0b0t1.

I am using Keysnail, no idea whether or not it will be deprecated.

If the features I want are blocked by the new framework, I'm afraid I
would lock the version (fortunately it's Gentoo) for a long time, until
I find replacement for plugins, or even Firefox...

Danny

Re: [gentoo-user] What do you think about Firefox 57?

[Danny YUE]

On 2017-09-07 14:47, Ralph Seichter <m16+gen...@monksofcool.net> wrote:
> On 07.09.2017 15:20, Danny YUE wrote:
>
>> Well, I know it is not the perfect place but I don't want to spam my
>> inbox with Firefox topics, so that's why I don't subscribe that mail
>> list and am asking here. ;-)
>
> And you believe that spamming the inbox people interested in Gentoo
> (that's why we subscribed to this mailing list) is OK ? ;-) Seriously,
> this is completely off-topic here, so please use the appropriate mailing
> lists instead. Thanks.
>
> -Ralph

Maybe this can be understood as "application software discussion in
Gentoo"? XD

Let's just stop here...sorry for any potential spam.


Danny

Re: [gentoo-user] What do you think about Firefox 57?

[Danny YUE]

On 2017-09-07 13:00, Ralph Seichter <m16+gen...@monksofcool.net> wrote:
> On 07.09.2017 14:26, Danny YUE wrote:
>
>> I have been using FoxyProxy in Firefox for a really long time, until
>> today I found its new version really sucks.
>
> If you're curious about Mozilla's move to the WebExtensions API, see
> https://blog.mozilla.org/addons/2016/11/23/add-ons-in-2017/ . You'll
> probably have more in-depth discussions about this on other mailing
> lists, since it is not Gentoo-specific.
>
> -Ralph

Thanks so much for the link.

Well, I know it is not the perfect place but I don't want to spam my
inbox with Firefox topics, so that's why I don't subscribe that mail
list and am asking here. ;-)

By the way I have not used Firefox 57 yet...
I have set keyword ~amd64 so my version is 55.0.2 currently.


Danny

[gentoo-user] What do you think about Firefox 57?

[Danny YUE]

Hi all,

I have been using FoxyProxy in Firefox for a really long time, until
today I found its new version really sucks.

Then I read the comment from author who declared that the old version
can *only* be used before (roughly) end of 2017 before Firefox 57 and in
new version some features must perish.

Afterwards I found that it seems Firefox 57 will use a new ecosystem for
extensions and be more strict for plugin developers.

So Firefox gurus, what do you think about it?


Danny

Re: [gentoo-user] Questions on Raspberry Pi cross compiling

[Danny YUE]

On 2017-07-20 10:10, Neil Bothwick <n...@digimed.co.uk> wrote:
> On Thu, 20 Jul 2017 13:42:32 +0800, Danny YUE wrote:
>
>> And if some (even very few) packages fail to build on the PC, it is
>> hardly possible to keep PC and RPI 'consistent'.
>> (Yes, I did setup the 'make.profile' symlink to the right place.)
>> 
>> So my question is:
>> 1) If some packages are *doomed* to fail, how do you keep the
>> emerge world environment consistency between PC and RPI?
>> Or is my understanding of this method incorrect?
>
> For the packages that will not build on the PC, you'll have to build them
> on the Pi, create a package then install that on the PC.

Oh man! I am stupid

I did not think about install packages from RPI to PC...

You made my day! 

I will try to setup distcc then...Thank you so much.

Re: [gentoo-user] Questions on Raspberry Pi cross compiling

[Danny YUE]

On 2017-07-20 06:33, Danny YUE <sheepd...@gmail.com> wrote:
> On 2017-07-20 06:25, R0b0t1 <r03...@gmail.com> wrote:
>> On Thu, Jul 20, 2017 at 1:20 AM, Danny YUE <sheepd...@gmail.com> wrote:
>>>
>>> On 2017-07-20 05:59, R0b0t1 <r03...@gmail.com> wrote:
>>>> On Thu, Jul 20, 2017 at 12:42 AM, Danny YUE <sheepd...@gmail.com> wrote:
>>>>> Hi guys,
>>>>>
>>>>> I am setting up cross compiling environment for my newly bought
>>>>> Raspberry Pi 3, under the guide of:
>>>>> https://wiki.gentoo.org/wiki/Raspberry_Pi
>>>>> https://wiki.gentoo.org/wiki/Raspberry_Pi/Quick_Install_Guide
>>>>>
>>>>> My original idea was to use crossdev to cross compile packages on my PC
>>>>> and install binaries on RPI.
>>>>> However I found it really nasty because it kept giving me the error
>>>>> message about "libintl: no such file or directory" during compilation of
>>>>> packages such as attr, python etc.
>>>>>
>>>>> And if some (even very few) packages fail to build on the PC, it is
>>>>> hardly possible to keep PC and RPI 'consistent'.
>>>>> (Yes, I did setup the 'make.profile' symlink to the right place.)
>>>>>
>>>>> So my question is:
>>>>> 1) If some packages are *doomed* to fail, how do you keep the
>>>>> emerge world environment consistency between PC and RPI?
>>>>> Or is my understanding of this method incorrect?
>>>>>
>>>>
>>>> Match compilation and USE flags.
>>>>
>>>>> 2) If it is not really a good idea to use crossdev, which one do you
>>>>> recommend between distcc and chroot method?
>>>>> (I googled but did not really get one answer about compilation speed.)
>>>>>
>>>>
>>>> You should be able to mix crossdev compiled packages freely with
>>>> device compiled ones. Incompatible packages will be ignored and it
>>>> will try to recompile a package with matching flags, so pay attention.
>>>> If you need to compile something on device then I suspect you want to
>>>> use distcc if at all possible.
>>>>
>>> Thanks for your reply. :-)
>>>
>>> Well, yes. But do you know how is distcc compared with chroot referring
>>> to compilation speed?
>>>
>>
>> Using qemu-user to emulate the target architecture and hosting the
>> system in a chroot is generally slower than compiling on device, if
>> that is what you are referring to. I've read of people who tested this
>> with the RPi3 and some Hardkernel devices.
>
> Oops, really?
> I have read of this point of view, too.
> But the Wiki page says that it is faster than native compilation on RPI.
>
> To be honest, I prefer to compile using crossdev...
>
> Ok then, I will try re-setup crossdev on my PC and see if everything
> compiles well. If not, I will post the error message here.
>
> Thanks.

Ok, I am back.

This time I purged cross toolchain for Raspberry Pi and started it over
again.

I ran:
$ crossdev -S -v -t armv7a-hardfloat-linux-gnueabi
$ cd /usr/aarch64-unknown-linux-gnu/etc/portage
$ rm make.profile
$ ln -s /usr/portage/profiles/default/linux/arm64/13.0/armv7a make.profile

Then I did `armv7a-hardfloat-linux-gnueabi-emerge -auDNU @world`.
Most packages succedded to build, but 'util-linux-2.30' failed and gave
following messages:

--- BEGIN ---
/usr/libexec/gcc/armv7a-hardfloat-linux-gnueabi/ld: skipping incompatible 
/usr/lib64/libncursesw.so when searching for -lncursesw
/usr/libexec/gcc/armv7a-hardfloat-linux-gnueabi/ld: skipping incompatible 
/usr/lib64/libc.so when searching for -lc
/usr/lib64/libc.a: error adding symbols: File format not recognized
collect2: error: ld returned 1 exit status
make[2]: *** [Makefile:7070: ul] Error 1
make[2]: *** Waiting for unfinished jobs
libtool: link: armv7a-hardfloat-linux-gnueabi-gcc -fsigned-char -fno-common 
-Wall -Werror=sequence-point -Wextra -Wmissing-declarations 
-Wmissing-parameter-type -Wmissing-prototypes -Wno-missing-field-initializers 
-Wredundant-decls -Wsign-compare -Wtype-limits -Wuninitialized 
-Wunused-but-set-parameter -Wunused-but-set-variable -Wunused-parameter 
-Wunused-result -Wunused-variable -Wnested-externs -Wpointer-arith 
-Wstrict-prototypes -Wformat-security -Wimplicit-function-declaration -O2 
-march=armv7-a -mfpu=neon-vfpv4 -mfloat-abi=hard -Wl,-O1 -o setsid 
sys-utils/setsid.o  -Wl,--as-needed
libtool: link: armv7a-hardfloat-linux-gnueabi-gcc -fsigned-char -fno-common 
-Wall -Werror=sequence-point -Wextra -Wmissi

Re: [gentoo-user] Questions on Raspberry Pi cross compiling

[Danny YUE]

On 2017-07-20 06:25, R0b0t1 <r03...@gmail.com> wrote:
> On Thu, Jul 20, 2017 at 1:20 AM, Danny YUE <sheepd...@gmail.com> wrote:
>>
>> On 2017-07-20 05:59, R0b0t1 <r03...@gmail.com> wrote:
>>> On Thu, Jul 20, 2017 at 12:42 AM, Danny YUE <sheepd...@gmail.com> wrote:
>>>> Hi guys,
>>>>
>>>> I am setting up cross compiling environment for my newly bought
>>>> Raspberry Pi 3, under the guide of:
>>>> https://wiki.gentoo.org/wiki/Raspberry_Pi
>>>> https://wiki.gentoo.org/wiki/Raspberry_Pi/Quick_Install_Guide
>>>>
>>>> My original idea was to use crossdev to cross compile packages on my PC
>>>> and install binaries on RPI.
>>>> However I found it really nasty because it kept giving me the error
>>>> message about "libintl: no such file or directory" during compilation of
>>>> packages such as attr, python etc.
>>>>
>>>> And if some (even very few) packages fail to build on the PC, it is
>>>> hardly possible to keep PC and RPI 'consistent'.
>>>> (Yes, I did setup the 'make.profile' symlink to the right place.)
>>>>
>>>> So my question is:
>>>> 1) If some packages are *doomed* to fail, how do you keep the
>>>> emerge world environment consistency between PC and RPI?
>>>> Or is my understanding of this method incorrect?
>>>>
>>>
>>> Match compilation and USE flags.
>>>
>>>> 2) If it is not really a good idea to use crossdev, which one do you
>>>> recommend between distcc and chroot method?
>>>> (I googled but did not really get one answer about compilation speed.)
>>>>
>>>
>>> You should be able to mix crossdev compiled packages freely with
>>> device compiled ones. Incompatible packages will be ignored and it
>>> will try to recompile a package with matching flags, so pay attention.
>>> If you need to compile something on device then I suspect you want to
>>> use distcc if at all possible.
>>>
>> Thanks for your reply. :-)
>>
>> Well, yes. But do you know how is distcc compared with chroot referring
>> to compilation speed?
>>
>
> Using qemu-user to emulate the target architecture and hosting the
> system in a chroot is generally slower than compiling on device, if
> that is what you are referring to. I've read of people who tested this
> with the RPi3 and some Hardkernel devices.

Oops, really?
I have read of this point of view, too.
But the Wiki page says that it is faster than native compilation on RPI.

To be honest, I prefer to compile using crossdev...

Ok then, I will try re-setup crossdev on my PC and see if everything
compiles well. If not, I will post the error message here.

Thanks.

Re: [gentoo-user] Questions on Raspberry Pi cross compiling

[Danny YUE]

On 2017-07-20 05:59, R0b0t1 <r03...@gmail.com> wrote:
> On Thu, Jul 20, 2017 at 12:42 AM, Danny YUE <sheepd...@gmail.com> wrote:
>> Hi guys,
>>
>> I am setting up cross compiling environment for my newly bought
>> Raspberry Pi 3, under the guide of:
>> https://wiki.gentoo.org/wiki/Raspberry_Pi
>> https://wiki.gentoo.org/wiki/Raspberry_Pi/Quick_Install_Guide
>>
>> My original idea was to use crossdev to cross compile packages on my PC
>> and install binaries on RPI.
>> However I found it really nasty because it kept giving me the error
>> message about "libintl: no such file or directory" during compilation of
>> packages such as attr, python etc.
>>
>> And if some (even very few) packages fail to build on the PC, it is
>> hardly possible to keep PC and RPI 'consistent'.
>> (Yes, I did setup the 'make.profile' symlink to the right place.)
>>
>> So my question is:
>> 1) If some packages are *doomed* to fail, how do you keep the
>> emerge world environment consistency between PC and RPI?
>> Or is my understanding of this method incorrect?
>>
>
> Match compilation and USE flags.
>
>> 2) If it is not really a good idea to use crossdev, which one do you
>> recommend between distcc and chroot method?
>> (I googled but did not really get one answer about compilation speed.)
>>
>
> You should be able to mix crossdev compiled packages freely with
> device compiled ones. Incompatible packages will be ignored and it
> will try to recompile a package with matching flags, so pay attention.
> If you need to compile something on device then I suspect you want to
> use distcc if at all possible.
>
Thanks for your reply. :-)

Well, yes. But do you know how is distcc compared with chroot referring
to compilation speed?

> Please submit a bug report for packages that don't compile. You
> probably want to do it on the Gentoo tracker. These tend to be
> problems with autoconf, but developers seem to be reluctant to
> learning about the autoconf tests they make use of or providing fixes
> for them even though they're likely the people most capable creating
> fixes. The autoconf developers do not seem to have any idea about how
> to deal with cross compilation failures in a centralized way.
>
>> Thanks in advance for any potential help. ;-)
>>
>> P.S. Does anyone have any idea about the 'libintl' error? I googled a
>> lot but it seems that it should be part of glibc...
>
> Can you post the full output of the error you're receiving?

Sorry, I am not using my PC currently.

I will re-install cross toolchain later on today and post the result
here.

Thanks again.

Danny

[gentoo-user] Questions on Raspberry Pi cross compiling

[Danny YUE]

Hi guys,

I am setting up cross compiling environment for my newly bought
Raspberry Pi 3, under the guide of:
https://wiki.gentoo.org/wiki/Raspberry_Pi
https://wiki.gentoo.org/wiki/Raspberry_Pi/Quick_Install_Guide

My original idea was to use crossdev to cross compile packages on my PC
and install binaries on RPI.
However I found it really nasty because it kept giving me the error
message about "libintl: no such file or directory" during compilation of
packages such as attr, python etc.

And if some (even very few) packages fail to build on the PC, it is
hardly possible to keep PC and RPI 'consistent'.
(Yes, I did setup the 'make.profile' symlink to the right place.)

So my question is:
1) If some packages are *doomed* to fail, how do you keep the
emerge world environment consistency between PC and RPI?
Or is my understanding of this method incorrect?

2) If it is not really a good idea to use crossdev, which one do you
recommend between distcc and chroot method?
(I googled but did not really get one answer about compilation speed.)

Thanks in advance for any potential help. ;-)

P.S. Does anyone have any idea about the 'libintl' error? I googled a
lot but it seems that it should be part of glibc...

Danny

[gentoo-user] emerge problem after sync today

[Danny YUE]

Hi all,

I encountered a problem after `emerge --sync` today.

When I ran `emerge -auDNU @world`, it told me something strange
following the list of upgradable software:

Would you like to merge these packages? [Yes/No] 
>>> Verifying ebuild manifests
!!! A file is not listed in the Manifest: 
'/usr/portage/app-arch/unrar/unrar-5.4.5.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/sys-apps/less/less-494.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/dev-python/sip/sip-4..ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/dev-python/pillow/pillow-4.2.1.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/dev-python/sphinx/sphinx-1.3.1-r2.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/www-servers/nginx/nginx-1.13.0.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/app-text/podofo/podofo-0.9.6_pre20170508-r1.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/dev-lang/vala/vala-0.32.1.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/sys-libs/db/db-4.8.30-r2.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/sys-libs/db/db-4.8.30-r2.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/net-dns/bind-tools/bind-tools-9.11.0_p3.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/dev-lang/php/php-5.6.30.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/app-admin/apache-tools/apache-tools-2.4.25.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/app-text/dos2unix/dos2unix-7.3.5.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/media-libs/harfbuzz/harfbuzz-1.4.6-r1.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/dev-cpp/gtkmm/gtkmm-2.24.5.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/app-text/calibre/calibre-3.1.1-r2.ebuild'
!!! A file is not listed in the Manifest: 
'/usr/portage/gnome-base/gnome-keyring/gnome-keyring-3.20.1.ebuild'

Then I checked unrar ebuild directory. I found that Manifest file is
empty.

The problem did not disappear after I purged /usr/portage and redo
`emerge --sync`. So I suspect this is an error in the upstream
repository.


This is really strange. Any idea?


Danny

Re: [gentoo-user] Gentoo vs Raspbian on Raspberry Pi 3?

[Danny YUE]

On 2017-07-07 16:29, james <gar...@verizon.net> wrote:
> On 06/26/17 04:41, Danny YUE wrote:
>> Hi guys,
>> 
>> I am planning to buy a Raspberry Pi 3 to setup a local network storage
>> and try some IoT stuff. Now I am searching for some background things.
>> 
>> Which OS do you install on your Raspberry Pi?
>> 
>> Gentoo? Raspbian? Or even...arch?
>> 
>> This might seems to be the wrong place to ask, but I consider Gentoo as
>> my first-place option, so I would like to know your experiences.
>> 
>> Thanks.
>> 
>> Danny
>
> Hello Danny,
>
> There was another gentoo wiki page, that I just ran across with many of
> your (cross compiling) issues addressed, albeit for the Banana Rasp. Pi::
>
> https://wiki.gentoo.org/wiki/Banana_Pi_the_Gentoo_Way
>
>
> hth,
> James

Hi James,

Thanks for that.

Actually I got to know that it is possible to use chroot to compile.

I will try to compare different ways after my holiday ;-)


Danny

Re: [gentoo-user] Gentoo vs Raspbian on Raspberry Pi 3?

[Danny YUE]

On 2017-06-27 22:21, Bill Kenworthy  wrote:
> On 28/06/17 02:29, Herminio Hernandez, Jr. wrote:
>> Where they using the original RPI? I built LXQT on an RPI2 and it did
>> not take me a day.
>> 
>> On Tue, Jun 27, 2017 at 11:26 AM, R0b0t1 > > wrote:
>> 
>> On Tue, Jun 27, 2017 at 3:44 AM, konsolebox > > wrote:
>> > On Tue, Jun 27, 2017 at 11:37 AM, R0b0t1 > > wrote:
>> >> it's
>> >> not experience.
>> >
>> > Ok.
>> >
>> 
>> Why do you think I made it up? I was quoting people who had done it.
>> 
>> 
>
> rpi1 ~ # genlop -t gcc
>  * sys-devel/gcc
>
>  Fri Jul  1 10:01:57 2016 >>> sys-devel/gcc-4.9.3
>merge time: 1 day, 4 hours, 32 minutes and 26 seconds.
>
>  Fri Nov  4 10:49:45 2016 >>> sys-devel/gcc-4.9.4
>merge time: 1 day, 4 hours, 43 minutes and 51 seconds.
>
> rpi1 ~ #
>
> rpi1 ~ # genlop -t glibc
>  * sys-libs/glibc
>
>  Fri May 20 02:48:34 2016 >>> sys-libs/glibc-2.22-r4
>merge time: 5 hours, 30 minutes and 52 seconds.
>
>  Thu Sep  1 20:07:48 2016 >>> sys-libs/glibc-2.22-r4
>merge time: 4 hours, 58 minutes and 44 seconds.
>
>  Fri Jan 13 15:11:12 2017 >>> sys-libs/glibc-2.23-r3
>merge time: 5 hours, 19 minutes and 51 seconds.
>
> rpi1 ~ #
>
> pi model B (512M ram), ccache and distcc to a single vm with crossdev
>
> Fact! :)
>
> BillK

I do not really understand, since building on RPI takes so long compared
with PC, why do you still compile natively?
Is it because cross compiling will cause any bugs or failure?

I mean, even some packages will fail with cross compiler, it is still
possible to cross compile some *big* packages...

Danny

Re: [gentoo-user] Gentoo vs Raspbian on Raspberry Pi 3?

[Danny YUE]

On 2017-06-27 02:59, taii...@gmx.com  wrote:
> I would advise to buy an open source device such as beaglebone not a 
> closed source RPI, bb also has higher performance options and is a 
> better company.
>
> I was not at all pleased with the transfer speed of an RPI I tried out, 
> the low end arm stuff is garbage (high end like appliedmicro is decent tho)
>
> If you want a decent fileserver I would advise getting a KCMA-D8 with a 
> 35W opteron and installing the libre version of coreboot on it, dual 
> onboard gigabit ethernet will satisfy you for sure.

Well, file server is only one of my use cases.

I would probably try to run a little server, or some little tricks I
write myself.

I will have a look at BeagleBone.

Thanks!

Danny

Re: [gentoo-user] Gentoo vs Raspbian on Raspberry Pi 3?

[Danny YUE]

On 2017-06-26 16:53, james <gar...@verizon.net> wrote:
> On 06/26/17 04:41, Danny YUE wrote:
>> Hi guys,
>> 
>> I am planning to buy a Raspberry Pi 3 to setup a local network storage
>> and try some IoT stuff. Now I am searching for some background things.
>> 
>> Which OS do you install on your Raspberry Pi?
>> 
>> Gentoo? Raspbian? Or even...arch?
>> 
>> This might seems to be the wrong place to ask, but I consider Gentoo as
>> my first-place option, so I would like to know your experiences.
>> 
>> Thanks.
>> 
>> Danny
>> 
>> 
>
> I'm assuming you've seen these pages?
>
>
> https://wiki.gentoo.org/wiki/Raspberry_Pi
>
> https://wiki.gentoo.org/wiki/Raspberry_Pi#Stage_4
>
> https://wiki.gentoo.org/wiki/Raspberry_Pi/Quick_Install_Guide
>
>
>  Why not run a Rpi-3 cluster? (lots of folks experimenting and posting):
>
> https://wiki.gentoo.org/wiki/Corosync
>
> https://wiki.osuosl.org/ganeti/cluster_gentoo.html
>
> https://forums.gentoo.org/viewtopic-t-1058886.html
>
> https://forums.gentoo.org/viewtopic-t-1058886.html
>
>
>
> Note 'distcc' purports to run on arm::
>
> https://packages.gentoo.org/packages/sys-devel/distcc
>
> hth,
> James

Yes.

I briefly know how to install Gentoo on RPI, also I have used crossdev
before.

But...distcc was described as "not recommended" in the Gentoo
installation guide.
(Do not remember where exactly.)

So I did not really used that though.

Danny

Re: [gentoo-user] Gentoo vs Raspbian on Raspberry Pi 3?

[Danny YUE]

Gentoo 64 bit was made working on Raspberry Pi 3 not long ago.

I tried Arch for a short time last year. I was really missing those USE
flags and ability to lock/select software version.

Also when I was using Arch, I always felt that it was pushing me hard to
upgrade my system, which really feels uncomfortable.

I think that Gentoo runs more stable than Arch (might be wrong, just my
feeling), and with cross compiling it should not be a big problem to
upgrade whole system.

I do agree that using Arch on Raspberry Pi would be much simpler than
Gentoo. Anyway I will try to compile natively after I receive my Pi ;-)

But till now nobody ever mentioned *official* raspbian OS. Funny.


On 2017-06-26 09:00, Herminio Hernandez, Jr. <herminio.hernande...@gmail.com> 
wrote:
> The next time I use Gentoo I am planning have my build directoy on seperate
> storage. I know I can cross compile, but building natively did not take as
> long as you might think. If you are looking for similicty then try Arch.
>
> On Mon, Jun 26, 2017 at 1:58 AM, Herminio Hernandez, Jr. <
> herminio.hernande...@gmail.com> wrote:
>
>> Well right now I am running Arch on my Pi mostly b/c when I was running
>> Gentoo I was compiling natiavly.  Arch has the benefit of running a minimal
>> install and you can basically build from there. I had some issues running
>> Gentoo, but that was mostly b/c I was building nativaly on the SD card.
>>
>> On Mon, Jun 26, 2017 at 1:49 AM, Danny YUE <sheepd...@gmail.com> wrote:
>>
>>> Any reasons? I want to hear some ideas to make the decision ;-)
>>>
>>> Seriously, I do not think it is a good idea to compile using Raspberry
>>> Pi because of the weak CPU.
>>>
>>> I prefer to cross compile on my PC when using Gentoo.
>>>
>>>
>>> On 2017-06-26 08:47, Herminio Hernandez, Jr. <
>>> herminio.hernande...@gmail.com> wrote:
>>> > I prefer Gentoo or Arch. With Gentoo consider making /var/tmp/portage on
>>> > USB drive. This way you have your build directory not on the SD card.
>>> >
>>> > On Mon, Jun 26, 2017 at 1:41 AM, Danny YUE <sheepd...@gmail.com> wrote:
>>> >
>>> >> Hi guys,
>>> >>
>>> >> I am planning to buy a Raspberry Pi 3 to setup a local network storage
>>> >> and try some IoT stuff. Now I am searching for some background things.
>>> >>
>>> >> Which OS do you install on your Raspberry Pi?
>>> >>
>>> >> Gentoo? Raspbian? Or even...arch?
>>> >>
>>> >> This might seems to be the wrong place to ask, but I consider Gentoo as
>>> >> my first-place option, so I would like to know your experiences.
>>> >>
>>> >> Thanks.
>>> >>
>>> >> Danny
>>> >>
>>> >>
>>>
>>>
>>>
>>

Re: [gentoo-user] Gentoo vs Raspbian on Raspberry Pi 3?

[Danny YUE]

Any reasons? I want to hear some ideas to make the decision ;-)

Seriously, I do not think it is a good idea to compile using Raspberry
Pi because of the weak CPU.

I prefer to cross compile on my PC when using Gentoo.


On 2017-06-26 08:47, Herminio Hernandez, Jr. <herminio.hernande...@gmail.com> 
wrote:
> I prefer Gentoo or Arch. With Gentoo consider making /var/tmp/portage on
> USB drive. This way you have your build directory not on the SD card.
>
> On Mon, Jun 26, 2017 at 1:41 AM, Danny YUE <sheepd...@gmail.com> wrote:
>
>> Hi guys,
>>
>> I am planning to buy a Raspberry Pi 3 to setup a local network storage
>> and try some IoT stuff. Now I am searching for some background things.
>>
>> Which OS do you install on your Raspberry Pi?
>>
>> Gentoo? Raspbian? Or even...arch?
>>
>> This might seems to be the wrong place to ask, but I consider Gentoo as
>> my first-place option, so I would like to know your experiences.
>>
>> Thanks.
>>
>> Danny
>>
>>

[gentoo-user] Gentoo vs Raspbian on Raspberry Pi 3?

[Danny YUE]

Hi guys,

I am planning to buy a Raspberry Pi 3 to setup a local network storage
and try some IoT stuff. Now I am searching for some background things.

Which OS do you install on your Raspberry Pi?

Gentoo? Raspbian? Or even...arch?

This might seems to be the wrong place to ask, but I consider Gentoo as
my first-place option, so I would like to know your experiences.

Thanks.

Danny

Re: [gentoo-user] Firefox depends on rust??

[Danny YUE]

On 2017-06-22 11:31, Vadim A. Misbakh-Soloviov  wrote:
>> I just installed firefox-bin and did not feel any difference.
>> 
>> I will try to compile firefox just for curiosity.
>
> Well. Only flags that can give you any difference is bindist (logos and so 
> on), 
> pgo (hard-way optimization), custom-*, and system-* ones.
>
> So, taking default "fx-bin" vs "fx" doesn't give you much difference.
>
> Although, I myself distrust any "*-bin" packages, and I'd never run them 
> outside of `firejail` with blacklisted access to anywhere on the filesystem.
>
> Although, I running even built-from-source firefox under firejail anyway (to 
> prevent java, js, drm and other crap to do anything on my system. even in 
> home 
> dir.

Well, I finally compiled firefox and rust and used them.

Just...keep it the Gentoo way ;)

Re: [gentoo-user] Firefox depends on rust??

[Danny YUE]

On 2017-06-22 08:27, Alexey Eschenko <skobkin...@ya.ru> wrote:
> For me it was a *slightly* less quicker start. Which I think don't means 
> much when speaking about browser.
>
>
> On 06/22/2017 11:09 AM, Danny YUE wrote:
>> On 2017-06-22 07:43, Rasmus Thomsen <rasmus.thom...@protonmail.com> wrote:
>>> I'm using firefox-bin ( and libreoffice-bin ) on my laptop and I didn't 
>>> have problems with either of them
>>>
>>> Regards,
>>> Rasmus
>>>  Original Message 
>>> On 22 Jun 2017, 09:34, Danny YUE wrote:
>>>
>>> On 2017-06-22 07:23, Vadim A. Misbakh-Soloviov  wrote:
>>>>> Does anyone knows why? Any idea?
>>>> The reason is in that fact, that many of it's components are in rust.
>>>> And since it was possible to dodge it until now, maintainers considered it 
>>>> is
>>>> not a way since now.
>>>>
>>>> And, by the way, it is not that many time to build rust, as you think:
>>>>> Thu Jun 22 12:34:00 2017 >>> dev-lang/rust-1.16.0
>>>>> merge time: 1 hour, 48 minutes.
>>>> Than was on 1.9GHz with hardly limited portage (MAKEOPTS="-j5 -l2",
>>>> NICENESS=18, ionice -c3, and cgroupped on cpu shares and ram).
>>>>
>>>> So, ~20 mins would be enough on non-limited portage and full power of that 
>>>> i7.
>>> Thank you all for replying.
>>>
>>> So it can be around 30~40 minutes or so on my i5 machine.
>>> Just it feels strange to install something large that I would probably
>>> never use myself.
>>>
>>> I am considering using binary package instead of compiling it myself.
>>>
>>> But I am afraid that using firefox-bin package would cause some
>>> dependency problem. I once tried libreoffice-bin, but found it really
>>> painful to resolve dependency issues during system upgrading.
>>>
>>> Anyone tried firefox-bin smoothly?
>>>
>>> Danny
>>> @mva.name>
>> Well, I ran into the same problem with libreoffice-bin *last time*, as
>> Alexey.
>>
>> It seems that version number of libreoffice-bin is always smaller than
>> libreoffice. So dependency issue is always a problem with it.
>>
>> I noticed firefox-bin only has *usual* packages as dependencies...
>> By the way what is the difference between compiled and binary firefox
>> from a user's perspective?
>>
>> P.S. Someone told me that people in this list do not like top-posting.
>>
>> Thanks.
>> Danny
>>

I agree.

Anyway I just start my Emacs and Firefox and keep them open until system
shutdown.

I just installed firefox-bin and did not feel any difference.

I will try to compile firefox just for curiosity.

Thank you all.

Danny

Re: [gentoo-user] Firefox depends on rust??

[Danny YUE]

On 2017-06-22 07:43, Rasmus Thomsen <rasmus.thom...@protonmail.com> wrote:
> I'm using firefox-bin ( and libreoffice-bin ) on my laptop and I didn't have 
> problems with either of them
>
> Regards,
> Rasmus
>  Original Message 
> On 22 Jun 2017, 09:34, Danny YUE wrote:
>
> On 2017-06-22 07:23, Vadim A. Misbakh-Soloviov  wrote:
>>> Does anyone knows why? Any idea?
>> The reason is in that fact, that many of it's components are in rust.
>> And since it was possible to dodge it until now, maintainers considered it is
>> not a way since now.
>>
>> And, by the way, it is not that many time to build rust, as you think:
>>> Thu Jun 22 12:34:00 2017 >>> dev-lang/rust-1.16.0
>>> merge time: 1 hour, 48 minutes.
>> Than was on 1.9GHz with hardly limited portage (MAKEOPTS="-j5 -l2",
>> NICENESS=18, ionice -c3, and cgroupped on cpu shares and ram).
>>
>> So, ~20 mins would be enough on non-limited portage and full power of that 
>> i7.
>
> Thank you all for replying.
>
> So it can be around 30~40 minutes or so on my i5 machine.
> Just it feels strange to install something large that I would probably
> never use myself.
>
> I am considering using binary package instead of compiling it myself.
>
> But I am afraid that using firefox-bin package would cause some
> dependency problem. I once tried libreoffice-bin, but found it really
> painful to resolve dependency issues during system upgrading.
>
> Anyone tried firefox-bin smoothly?
>
> Danny
> @mva.name>

Well, I ran into the same problem with libreoffice-bin *last time*, as
Alexey.

It seems that version number of libreoffice-bin is always smaller than
libreoffice. So dependency issue is always a problem with it.

I noticed firefox-bin only has *usual* packages as dependencies...
By the way what is the difference between compiled and binary firefox
from a user's perspective?

P.S. Someone told me that people in this list do not like top-posting.

Thanks.
Danny

Re: [gentoo-user] Firefox depends on rust??

[Danny YUE]

On 2017-06-22 07:23, Vadim A. Misbakh-Soloviov  wrote:
>> Does anyone knows why? Any idea?
> The reason is in that fact, that many of it's components are in rust.
> And since it was possible to dodge it until now, maintainers considered it is
> not a way since now.
>
> And, by the way, it is not that many time to build rust, as you think:
>> Thu Jun 22 12:34:00 2017 >>> dev-lang/rust-1.16.0
>>   merge time: 1 hour, 48 minutes.
> Than was on 1.9GHz with hardly limited portage (MAKEOPTS="-j5 -l2",
> NICENESS=18, ionice -c3, and cgroupped on cpu shares and ram).
>
> So, ~20 mins would be enough on non-limited portage and full power of that i7.

Thank you all for replying.

So it can be around 30~40 minutes or so on my i5 machine.
Just it feels strange to install something large that I would probably
never use myself.

I am considering using binary package instead of compiling it myself.

But I am afraid that using firefox-bin package would cause some
dependency problem. I once tried libreoffice-bin, but found it really
painful to resolve dependency issues during system upgrading.

Anyone tried firefox-bin smoothly?

Danny

[gentoo-user] Firefox depends on rust??

[Danny YUE]

Hi guys,

I just found during upgrading my system that Firefox 54.0 depends on
dev-lang/rust and cannot be disabled?!

It really takes a long time to compile rust, and I do not want to add
such a burden to my system compilation time.

Does anyone knows why? Any idea?

Thanks.

Danny

Re: [gentoo-user] Problem while writing ebuild

[Danny YUE]

On 2017-05-04 10:53, Nils Freydank <nils.freyd...@posteo.de> wrote:
> Am Donnerstag, 4. Mai 2017, 11:59:37 CEST schrieb Danny YUE:
>> Hi,
>> 
>> As a noob in ebuild files, I got some problems while created my own
>> ebuild. I googled a lot but did not really find any clue.
>> 
>> Let's say package A has features B and C. They are Github repositories
>> and A is the entry, while B and C are submodules.
>> 
>> Now I want to make it three different packages respectively, and use USE
>> flags to control which to install just like texlive.
>> 
>> 1) I know I can add USE `B' and `C' into `IUSE' variable, but if I run
>> `equery u A', those flags are listed but with an ""
>> description. Where can I add some description information?
>
> In general, these descriptions are in XML files. Globally used flags (e.g. 
> 'X' 
> or 'python') are set in the portage profiles (search for ".desc" in the 
> portage tree[1]). Flags used only by certain packages are in the ebuild dir 
> in 
> metadata.xml[2]. I suggest you take a look into www-client/firefox’s 
> metadata.xml, because it uses even descriptions with line breaks. (Avoid the 
> look into firefox’s ebuild, though :D)

That's really helpful. Now it shows the message I want.

>
>> 2) Only A has Github release URL, B and C by default should be fetched
>> via `git submodule...'. How am I supposed to fetch these packages in my
>> ebuild file? Download inside `src_unpack'?
>
> Sorry, I don’t know much about git submodule. Maybe there’s something useful 
> in git-r3.eclass (eclass/git-r3.eclass).
>
>> Any help is appreciated. Thanks in advance.
>> 
>> 
>> Danny
>
> [1] find /usr/portage/ -iname "*desc"
> [2] find /usr/portage/ -iname "*metadata.xml"
>
> Nils
>
> PS: Great you write ebuilds - keep going on and suggest bugfixes for broken 
> ebuilds later! :)

Thank you so much :-D

P.S. Well, riscv-tools is out there for too long..I am really surprised
that nobody ever wrote an ebuild for it...

[gentoo-user] Problem while writing ebuild

[Danny YUE]

Hi,

As a noob in ebuild files, I got some problems while created my own
ebuild. I googled a lot but did not really find any clue.

Let's say package A has features B and C. They are Github repositories
and A is the entry, while B and C are submodules.

Now I want to make it three different packages respectively, and use USE
flags to control which to install just like texlive.

1) I know I can add USE `B' and `C' into `IUSE' variable, but if I run
`equery u A', those flags are listed but with an ""
description. Where can I add some description information?

2) Only A has Github release URL, B and C by default should be fetched
via `git submodule...'. How am I supposed to fetch these packages in my
ebuild file? Download inside `src_unpack'?

Any help is appreciated. Thanks in advance.


Danny

[gentoo-user] How do you get newer version of Arduino?

[Danny YUE]

Hi folks,

I am playing around with my arduino board...
I know there is an official package arduino, but that is way too old.
So...where do you get the latest version? For example 1.8?

Thanks in advance.

Danny

Re: [gentoo-user] How do you manage manually compiled software?

[Danny YUE]

On 2017-04-28 02:19, Michael Orlitzky <m...@gentoo.org> wrote:
> On 04/27/2017 09:33 PM, Danny YUE wrote:
>> Hi guys,
>> 
>> I am compiling RISC-V tools...I am just curious how do you manage your
>> manually compiled software?
>
> Don't, write an ebuild for it.

Well, since everyone is recommending the "correct Gentoo way"..
I will give it a try.

Thanks.

[gentoo-user] How do you manage manually compiled software?

[Danny YUE]

Hi guys,

I am compiling RISC-V tools...I am just curious how do you manage your
manually compiled software? Here are some options:

1) Just "make install" everything into /usr/local. But if the Makefile
does not have removing option, how do you remove it then?

2) Put compiled software into /usr/local/some-dir, then create symbolic
links to /usr/local/some-path. But this does not work if the software
relies on some environment variables (well, you can set it in .bashrc
but it then loses flexibility).

3) Just install it anywhere and change PATH. To be honest, I don't
really like this way because changing the PATH requires extra actions,
e.g. open another terminal.

Any idea?

Danny

Re: [gentoo-user] replacement for ftp?

[Danny YUE]

On 2017-04-26 20:25, R0b0t1 <r03...@gmail.com> wrote:
> On Tue, Apr 25, 2017 at 8:26 PM, Danny YUE <sheepd...@gmail.com> wrote:
>>
>> On 2017-04-25 19:59, R0b0t1 <r03...@gmail.com> wrote:
>>> On Tue, Apr 25, 2017 at 10:05 AM, Danny YUE <sheepd...@gmail.com> wrote:
>>>>
>>>> On 2017-04-25 14:29, lee <l...@yagibdah.de> wrote:
>>>>> Hi,
>>>>>
>>>>> since the usage of FTP seems to be declining, what is a replacement
>>>>> which is at least as good as FTP?
>>>>>
>>>>> I'm aware that there's webdav, but that's very awkward to use and
>>>>> missing features.
>>>>
>>>> What about sshfs? It allows you to mount a location that can be accessed
>>>> via ssh to your local file system, as if you are using ssh.
>>>>
>>>
>>> In a similar vein, scp.
>>
>> And considering something still robust but a little more smart, rsync.
>>
>
> I was actually going to come back and suggest rscync over ssh. I
> didn't originally mention it because I typically associate rsync with
> backups.

I recently found that rsnapshot (based on rsync) is a good and solid
tool for backup...You may try it out ;-)

Re: [gentoo-user] replacement for ftp?

[Danny YUE]

On 2017-04-25 19:59, R0b0t1 <r03...@gmail.com> wrote:
> On Tue, Apr 25, 2017 at 10:05 AM, Danny YUE <sheepd...@gmail.com> wrote:
>>
>> On 2017-04-25 14:29, lee <l...@yagibdah.de> wrote:
>>> Hi,
>>>
>>> since the usage of FTP seems to be declining, what is a replacement
>>> which is at least as good as FTP?
>>>
>>> I'm aware that there's webdav, but that's very awkward to use and
>>> missing features.
>>
>> What about sshfs? It allows you to mount a location that can be accessed
>> via ssh to your local file system, as if you are using ssh.
>>
>
> In a similar vein, scp.

And considering something still robust but a little more smart, rsync.

Re: [gentoo-user] replacement for ftp?

[Danny YUE]

On 2017-04-25 14:29, lee  wrote:
> Hi,
>
> since the usage of FTP seems to be declining, what is a replacement
> which is at least as good as FTP?
>
> I'm aware that there's webdav, but that's very awkward to use and
> missing features.

What about sshfs? It allows you to mount a location that can be accessed
via ssh to your local file system, as if you are using ssh.

Also samba can be a replacement. I have a samba server on my OpenWRT
router and use mount.cifs to mount it...

May these be helpful.

Danny

Re: [gentoo-user] How do you emerge emacs 25 with xwidgets support?

[Danny YUE]

On 2017-04-20 20:09, Neil Bothwick  wrote:
> On Thu, 20 Apr 2017 18:09:48 +0200, Raffaele Belardi wrote:
>
>> Calculating dependencies... done!
>> [ebuild  N ] app-editors/emacs-25.2_rc2:25::gentoo  USE="X alsa
>> dbus gif gpm gtk gtk3 inotify jpeg png ssl svg tiff xpm zlib -Xaw3d
>> -acl (-aqua) -athena (-cairo) -dynamic-loading -games -gconf -gfile
>> -gsettings -gzip-el -hesiod -imagemagick -kerberos -libxml2 -livecd
>> -m17n-lib -motif -pax_kernel (-selinux) -sound -source
>> -toolkit-scroll-bars -wide-int -xft (-xwidgets)" 0 KiB
>> 
>> The parentheses indicate that the flag is 'forced, masked, or
>> removed' (from man emerge) so it is ignored on purpose. I have no idea
>> why.
>
> It is masked in the profile
>
> % grep -r xwidgets /var/portage/profiles
> ...
> /var/portage/profiles/base/package.use.mask:app-editors/emacs:25 xwidgets

Well, this is really surprising...I never thought of this. You made my day!

I checked the file "/usr/portage/profiles/base/package.use.mask", it tells:
# Ulrich Müller  (4 Feb 2017)
# Uses old and vulnerable net-libs/webkit-gtk:3, bug #584156.
app-editors/emacs:25 xwidgets
app-editors/emacs-vcs:25 xwidgets

For those who don't care about it and just want to use xwidgets:
Edit your use.mask file, mine is:
"/etc/portage/profile/package.use.mask".

Add line below:
app-editors/emacs -xwidgets

Danny

Re: [gentoo-user] How do you emerge emacs 25 with xwidgets support?

[Danny YUE]

On 2017-04-20 14:48, Raffaele Belardi <raffaele.bela...@st.com> wrote:
> Danny YUE wrote:
>> Hi guys,
>>
>> Maybe I am too stupid, but how can you emerge emacs-25 with xwidgets?
>>
>> Running `equey use emacs' shows:
>> ~ $ equery u emacs
>> [ Legend : U - final flag setting for installation]
>> [: I - package is installed with flag ]
>> [ Colors : set, unset ]
>>  * Found these USE flags for app-editors/emacs-25.2_rc2:
> ...
>>
>> Where is the xwidgets USE flag?
>>
>> I can see it by running `eix -e emacs' but setting that flag does not
>> work.
>>
>> `USE=xwidgets emerge -a emacs' does not work.
>>
>> Any idea? Thanks in advance for any clue.
>>
>> Danny
>>
>> .
>>
>
> $ euse -i -l xwidgets
> local use flags (searching: xwidgets)
> 
> [-  ] xwidgets
>  app-editors/emacs: Enable use of GTK widgets in Emacs buffers
>  (requires GTK3)
>(18) 18.59-r11 [gentoo]
>(23) 23.4-r14 [gentoo]
>(23) 23.4-r15 [gentoo]
>(24) 24.4-r4 [gentoo]
>(24) 24.5-r3 [gentoo]
>  [-  ] (25) 25.1 [gentoo]
>  [-  ] (25) 25.2_rc1 [gentoo]
>  [-  ] (25) 25.2_rc2 [gentoo]
>
>
> Maybe you need to set gtk3 also?
>
> raffaele

Thanks for the quick reply.

Well, my emacs was already built with gtk3. I wanted to try xwidget
browser out just now, but found no corresponding USE flag.

Corresponding line in package.use:
app-editors/emacs xft imagemagick libxml2 gtk3 xwidgets

Still not working.

BTW I got:
~ $ euse -i -l xwidgets
local use flags (searching: xwidgets)

[-  ] xwidgets
app-editors/emacs: Enable use of GTK widgets in Emacs buffers 
(requires GTK3)
  (18) 18.59-r11 [gentoo]
  (23) 23.4-r14 [gentoo]
  (23) 23.4-r15 [gentoo]
  (24) 24.4-r4 [gentoo]
  (24) 24.5-r3 [gentoo]
[+P ] (25) 25.1 [gentoo]
[+P ] (25) 25.2_rc1 [gentoo]
[+P ] (25) 25.2_rc2 [gentoo]

[-  ] xwidgets
app-editors/emacs-vcs: Enable use of GTK widgets in Emacs buffers 
(requires GTK3)
[-  ] (25) 25.1.91 [gentoo]
[-  ] (25) 25.2. [gentoo]
[-  ] (26) 26.0. [gentoo]

It seems strange...euse tells me there is a xwidgets USE for emacs, but
equery says something different.

Weird...

Danny

[gentoo-user] How do you emerge emacs 25 with xwidgets support?

[Danny YUE]

Hi guys,

Maybe I am too stupid, but how can you emerge emacs-25 with xwidgets?

Running `equey use emacs' shows:
~ $ equery u emacs
[ Legend : U - final flag setting for installation]
[: I - package is installed with flag ]
[ Colors : set, unset ]
 * Found these USE flags for app-editors/emacs-25.2_rc2:
 U I
 + + X   : Add support for X11
 - - Xaw3d   : Add support for the 3d athena widget set
 + + acl : Add support for Access Control Lists
 + + alsa: Add support for media-libs/alsa-lib (Advanced Linux 
Sound Architecture)
 - - athena  : Enable the MIT Athena widget set (x11-libs/libXaw)
 + + dbus: Enable dbus support for anything that needs it 
(gpsd, gnomemeeting, etc)
 - - dynamic-loading : Enable loading of dynamic libraries at runtime
 - - games   : Support shared score files for games
 - - gconf   : Use gnome-base/gconf to read the system font name
 - - gfile   : Use gfile (dev-libs/glib) for file notification
 + + gif : Add GIF image support
 + + gpm : Add support for sys-libs/gpm (Console-based mouse 
driver)
 - - gsettings   : Use gsettings (dev-libs/glib) to read the system 
font name
 - - gtk : Add support for x11-libs/gtk+ (The GIMP Toolkit)
 + + gtk3: Prefer version 3 of the GIMP Toolkit to version 2 
(x11-libs/gtk+)
 - - gzip-el : Compress bundled Emacs Lisp source
 - - hesiod  : Enable support for net-dns/hesiod
 + + imagemagick : Use media-gfx/imagemagick for image processing
 + + inotify : Enable inotify filesystem monitoring support
 + + jpeg: Add JPEG image support
 - - kerberos: Add kerberos support
 + + libxml2 : Use dev-libs/libxml2 to parse XML instead of the 
internal Lisp implementations
 - - livecd  : !!internal use only!! DO NOT SET THIS FLAG 
YOURSELF!, used during livecd building
 - - m17n-lib: Enable m17n-lib support
 - - motif   : Add support for the Motif toolkit
 - - pax_kernel  : Enable building under a PaX enabled kernel
 + + png : Add support for libpng (PNG images)
 - - sound   : Enable sound support
 - - source  : Install C source files and make them available for 
find-function
 + + ssl : Add support for Secure Socket Layer connections
 + + svg : Add support for SVG (Scalable Vector Graphics)
 + + tiff: Add support for the TIFF image format
 - - toolkit-scroll-bars : Use the selected toolkit's scrollbars in preference 
to Emacs' own scrollbars
 - - wide-int: Prefer wide Emacs integers (typically 62-bit). This 
option has an effect only on architectures where "long" and "long long" types 
have different size.
 + + xft : Build with support for XFT font renderer 
(x11-libs/libXft)
 + + xpm : Add support for XPM graphics format
 + + zlib: Add support for zlib (de)compression

Where is the xwidgets USE flag?

I can see it by running `eix -e emacs' but setting that flag does not
work.

`USE=xwidgets emerge -a emacs' does not work.

Any idea? Thanks in advance for any clue.

Danny

Re: [gentoo-user] Booting from USB

[Danny YUE]

On 2017-03-28 05:32, J. Roeleveld <jo...@antarean.org> wrote:
> On March 28, 2017 6:41:30 AM GMT+02:00, Danny YUE <sheepd...@gmail.com> wrote:
>>Not sure about mini-PC, but you may try another USB stick.
>>Seriously, Kingston isI never succeeded with it.
>>
>>Danny
>>
>>On 2017-03-28 04:02, the...@sys-concept.com wrote:
>>> On 03/27/2017 09:52 PM, tu...@posteo.de wrote:
>>>> On 03/27 09:46, the...@sys-concept.com wrote:
>>>>> I just got some Mini-PC from China with Windows 10 on it and trying
>>to boot from USB to install Gentoo on it.  
>>>>> But it will not boot from USB, I've tried front and back USB ports.
>>Windows start regardless what I do.
>>>>>
>>>>> Yes, I set in Bios 1st boot is USB. I've tried pressing F12. Tried
>>to follow some instruction from google:
>>>>>
>>https://www.tenforums.com/tutorials/21756-boot-usb-drive-windows-10-pc.html
>>>>>
>>>>> Doesn't help.
>>>>>
>>>>> I used rufus to make bootable USB sticks (have two of them). 
>>>>> -- 
>>>>> Thelma
>>>>>
>>>> 
>>>> Hi,
>>>> 
>>>> ...from my experience with SoC and such:
>>>> Some boards have a button on the board, which lets the board
>>>> boot from USB/sdcard while pressed...
>>>> 
>>>> Only a shot in the dark...
>>>> 
>>>> HTH!
>>>> Cheers
>>>> Meino
>>>
>>> I hope there is an easier option? I booted OK from USB my main server
>>> two months ago, but this Mini PC from China is different.
>
> Please stop top-posting.
> It is not accepted on this list.

Just curious: why is top-posting forbidden here?
I have never found any reference for that.

I would appreciate it if you could tell me something about that.

Danny

Re: [gentoo-user] Booting from USB

[Danny YUE]

Not sure about mini-PC, but you may try another USB stick.
Seriously, Kingston isI never succeeded with it.

Danny

On 2017-03-28 04:02, the...@sys-concept.com wrote:
> On 03/27/2017 09:52 PM, tu...@posteo.de wrote:
>> On 03/27 09:46, the...@sys-concept.com wrote:
>>> I just got some Mini-PC from China with Windows 10 on it and trying to boot 
>>> from USB to install Gentoo on it.  
>>> But it will not boot from USB, I've tried front and back USB ports. Windows 
>>> start regardless what I do.
>>>
>>> Yes, I set in Bios 1st boot is USB. I've tried pressing F12. Tried to 
>>> follow some instruction from google:
>>> https://www.tenforums.com/tutorials/21756-boot-usb-drive-windows-10-pc.html
>>>
>>> Doesn't help.
>>>
>>> I used rufus to make bootable USB sticks (have two of them). 
>>> -- 
>>> Thelma
>>>
>> 
>> Hi,
>> 
>> ...from my experience with SoC and such:
>> Some boards have a button on the board, which lets the board
>> boot from USB/sdcard while pressed...
>> 
>> Only a shot in the dark...
>> 
>> HTH!
>> Cheers
>> Meino
>
> I hope there is an easier option? I booted OK from USB my main server
> two months ago, but this Mini PC from China is different.

Re: [gentoo-user] Booting from USB

[Danny YUE]

Have you ever turned off "secure boot" or something similar in your BIOS
settings?

Also sometimes the USB stick itself has some problem. For example I can
never boot any system from my Kingston USB...

Danny

On 2017-03-28 03:46, the...@sys-concept.com wrote:
> I just got some Mini-PC from China with Windows 10 on it and trying to boot 
> from USB to install Gentoo on it.  
> But it will not boot from USB, I've tried front and back USB ports. Windows 
> start regardless what I do.
>
> Yes, I set in Bios 1st boot is USB. I've tried pressing F12. Tried to follow 
> some instruction from google:
> https://www.tenforums.com/tutorials/21756-boot-usb-drive-windows-10-pc.html
>
> Doesn't help.
>
> I used rufus to make bootable USB sticks (have two of them). 

Re: [gentoo-user] Re: Steam downloading extremely

[Danny YUE]

Hi Kai,

Oops, fixed?

I just updated my system and tried to randomly download another game
and it looks fine...

Weird.

Anyway, thanks for you guys.
May the good code be with you.

Danny


On 2017-03-19 19:11, Kai Krakow <hurikha...@gmail.com> wrote:
> Am Thu, 16 Mar 2017 14:04:14 +0800
> schrieb Danny YUE <sheepd...@gmail.com>:
>
>> Hi Kai,
>> 
>> Thanks for your help :-)
>> 
>> Code here:
>> /usr/share/info $ cat /proc/sys/net/ipv4/tcp_fastopen
>> 1
>> /usr/share/info $ cat /proc/sys/net/core/default_qdisc
>> pfifo_fast
>> /usr/share/info $ cat /proc/sys/net/ipv4/tcp_congestion_control
>> cubic
>> 
>> dnsmasq may help because...if my understanding is correct, Steam Linux
>> client has a bug that it tries to query the DNS too often during
>> downloading, then its request got throttled. Please see
>> https://wiki.gentoo.org/wiki/Steam/Client_troubleshooting
>> Section "Slow download speeds".
>
> This has been fixed with the March 9th 2017 Update. It's in the current
> stable client.
>
>> For disk, I don't think it fits my case because for a downloading
>> speed of 100KB/s, disk write should not be a bottleneck.
>
> Well, it still can be if there's a lot of data backlogged and the
> writeback cache of the kernel is saturated.
>
>> I suspect it is related to Linux client because Steam Windows client
>> on my machine downloads at the normal speed...
>
> This makes sense... However, here the linux client is downloading at 48
> mbytes/s which is pretty much the maximum of my 400 mbit link.
>
> So, if it is still slow for you, there seem to be other issues.
>
>> Well, I am not that familiar with network tuning things...so I will
>> definitely check the methods you mentioned.
>
> Feel free to ask.
>
>> On 2017-03-15 21:07, Kai Krakow <hurikha...@gmail.com> wrote:
>> > Am Wed, 15 Mar 2017 21:53:44 +0100
>> > schrieb Kai Krakow <hurikha...@gmail.com>:
>> >  
>> >> Am Wed, 15 Mar 2017 21:24:10 +0800
>> >> schrieb Danny YUE <sheepd...@gmail.com>:
>> >>  
>>  [...]  
>> >>
>> >> Here, it's downloading with peak bandwidths of 48 mbytes/s but
>> >> usually it's around 38 mbytes/s. However, I sometimes see
>> >> slowdowns, too. I guess that games are downloaded file by file,
>> >> and when a lot of small files are left in the queue, it just
>> >> cannot get good bandwidth.
>> >>
>> >> But I only see such slowdowns mostly short before the last few
>> >> megabytes of a game.
>> >>
>> >> Could you check if your downloaded games consist of many smallish
>> >> files? Then that could be the explanation.
>> >>
>> >> You could try activating fq_codel and tcp fastopen:
>> >>
>> >> In /proc/sys/net/ipv4/tcp_fastopen it should say 1.
>> >> In /proc/sys/net/core/default_qdisc it should say fq_codel.
>> >>
>> >> Also, you may want to try out bbr congestion control:
>> >>
>> >> In /proc/sys/net/ipv4/tcp_congestion_control it should say bbr.
>> >>
>> >> By recompiling the kernel, you can reconfigure the defaults for
>> >> this (and enable support). Some of these need modern kernels.
>> >>
>> >> Additionally, many small tcp request need a good portion of your
>> >> upload bandwidth and are very dependent on good round trip times.
>> >> Traditional DSL lines with ping times of 50-60ms can really slow
>> >> down requests of small files a lot due to three-way tcp
>> >> handshaking, that is, you could request only one smallish file per
>> >> 100-120ms. This can totally destroy the usable bandwidth. Maybe
>> >> watch a running ping while the downloads are running. If the ping
>> >> times increase while the download slows down, your bottleneck is
>> >> the upload.
>> >>
>> >> But also keep in mind that traditional spinning disks may not keep
>> >> up with the bandwidth if confronted with many small files. If
>> >> you're using SSD all should be fine. I'm running on bcache with
>> >> writeback caching which gives a really good performance boost by
>> >> adding a small SSD to one or more big HDDs.  
>> >
>> > BTW: I don't see how dnsmasq could help you here... It can do
>> > nothing about bandwidth. It only acts as a DNS cache which helps
>> > keeping latency of the DNS resolver down. But this doesn't matter
>> > here because during download, steam won't do many DNS requests.
>> >
>> > As already stated, part of the problem may be the upload, and/or bad
>> > queue handling within your broadband router. You can work around it
>> > with a traffic shaper and throttling upload below what's physically
>> > possible with your internet line, thus keeping the queue in front
>> > of the broadband router. That way, a traffic shaper could handle it
>> > and work around bad queue handling.
>> >
>> > To resolve the issue it is important to sophistically test the
>> > suggestions one step at a time, starting with the easy ones, and do
>> > your measurements.  
>> 
>> 

Re: [gentoo-user] Has someone managed to compile nvidia-drivers against a 4.10.*-Linux-kernel?

[Danny YUE]

Hi,

I got the same problem here.
And in the error message of nvidia-drivers, you would see something
like below:
currently only supports  wrote:
> On 18 March 2017 at 13:13,  wrote:
>
>> Hi,
>>
>> if someone has managed to compile the nividia-drivers
>> against one of the linux-4.10.* kernels I would
>> be glad fpr the information what version are compatible
>> with each other... :)
>>
>
> There are some patches for both the latest drivers and the older branch on
> the forums. I have not tried them, but look here:
> https://forums.gentoo.org/viewtopic-t-1059660-highlight-nvidia.html
>
> Cheers,
> Arve

Re: [gentoo-user] Re: Steam downloading extremely

[Danny YUE]

Hi Kai,

Thanks for your help :-)

Code here:
/usr/share/info $ cat /proc/sys/net/ipv4/tcp_fastopen
1
/usr/share/info $ cat /proc/sys/net/core/default_qdisc
pfifo_fast
/usr/share/info $ cat /proc/sys/net/ipv4/tcp_congestion_control
cubic

dnsmasq may help because...if my understanding is correct, Steam Linux
client has a bug that it tries to query the DNS too often during
downloading, then its request got throttled. Please see
https://wiki.gentoo.org/wiki/Steam/Client_troubleshooting
Section "Slow download speeds".

For disk, I don't think it fits my case because for a downloading speed
of 100KB/s, disk write should not be a bottleneck.

I suspect it is related to Linux client because Steam Windows client on
my machine downloads at the normal speed...

Well, I am not that familiar with network tuning things...so I will
definitely check the methods you mentioned.

Thanks,
Danny


On 2017-03-15 21:07, Kai Krakow <hurikha...@gmail.com> wrote:
> Am Wed, 15 Mar 2017 21:53:44 +0100
> schrieb Kai Krakow <hurikha...@gmail.com>:
>
>> Am Wed, 15 Mar 2017 21:24:10 +0800
>> schrieb Danny YUE <sheepd...@gmail.com>:
>>
>> > Hi guys,
>> >
>> > I just got Steam installed and running successfully on my machine,
>> > and tried to get CS:GO running smoothly, which made me really
>> > happy :-D
>> >
>> > However when Steam is downloading games, the speed is extremely
>> > slow, down to several KB/s, even some bytes/s.
>> > I have already installed dnsmasq and it *was* good during
>> > downloading CS:GO (~4MB/s), but became slow again with Civilization
>> > V.
>> >
>> > I googled a lot but all point to installing dnsmasq, which I don't
>> > think is really helpful since I already have done that...
>> >
>> > Also I'm sure downloading region is correct.
>> >
>> > Anybody experienced the same issue with dnsmasq installed?
>> > Any clue is welcome and thanks in advance.
>>
>> Here, it's downloading with peak bandwidths of 48 mbytes/s but usually
>> it's around 38 mbytes/s. However, I sometimes see slowdowns, too. I
>> guess that games are downloaded file by file, and when a lot of small
>> files are left in the queue, it just cannot get good bandwidth.
>>
>> But I only see such slowdowns mostly short before the last few
>> megabytes of a game.
>>
>> Could you check if your downloaded games consist of many smallish
>> files? Then that could be the explanation.
>>
>> You could try activating fq_codel and tcp fastopen:
>>
>> In /proc/sys/net/ipv4/tcp_fastopen it should say 1.
>> In /proc/sys/net/core/default_qdisc it should say fq_codel.
>>
>> Also, you may want to try out bbr congestion control:
>>
>> In /proc/sys/net/ipv4/tcp_congestion_control it should say bbr.
>>
>> By recompiling the kernel, you can reconfigure the defaults for this
>> (and enable support). Some of these need modern kernels.
>>
>> Additionally, many small tcp request need a good portion of your
>> upload bandwidth and are very dependent on good round trip times.
>> Traditional DSL lines with ping times of 50-60ms can really slow down
>> requests of small files a lot due to three-way tcp handshaking, that
>> is, you could request only one smallish file per 100-120ms. This can
>> totally destroy the usable bandwidth. Maybe watch a running ping
>> while the downloads are running. If the ping times increase while the
>> download slows down, your bottleneck is the upload.
>>
>> But also keep in mind that traditional spinning disks may not keep up
>> with the bandwidth if confronted with many small files. If you're
>> using SSD all should be fine. I'm running on bcache with writeback
>> caching which gives a really good performance boost by adding a small
>> SSD to one or more big HDDs.
>
> BTW: I don't see how dnsmasq could help you here... It can do nothing
> about bandwidth. It only acts as a DNS cache which helps keeping
> latency of the DNS resolver down. But this doesn't matter here because
> during download, steam won't do many DNS requests.
>
> As already stated, part of the problem may be the upload, and/or bad
> queue handling within your broadband router. You can work around it
> with a traffic shaper and throttling upload below what's physically
> possible with your internet line, thus keeping the queue in front of the
> broadband router. That way, a traffic shaper could handle it and work
> around bad queue handling.
>
> To resolve the issue it is important to sophistically test the
> suggestions one step at a time, starting with the easy ones, and do
> your measurements.

Re: [gentoo-user] Steam downloading extremely

[Danny YUE]

Thanks for your fast reply, and sorry for my late response.

The original is described as: Steam starts with a fast download speed,
and eventually goes down, even to 0 (which is probably caused by a bug
in Steam Linux client).

>From this link:
http://steamcommunity.com/app/221410/discussions/2/616189106498372437/
Just installing dnsmasq *does* solve the problem. And it did on my side.

Also this link https://wiki.gentoo.org/wiki/Steam/Client_troubleshooting
mentioned the same method. (See section "Slow download speeds").

However, my current problem is that when downloading Civilization V the
problem comes back (unstable speed, sometimes down to 0 KB/s), while
downloading CS:GO is acceptable high speed 4MB/s.
(Yes, ISP here is rather slow...My bandwidth is only 30 or 50 Mbit I
don't remember.)

dnsmasq is up, and downloading from other sources (browser, emerge etc)
is definitely fast.

I suspect it is a problem of Steam client...


Danny


On 2017-03-15 13:48, Nils Freydank <nils.freyd...@posteo.de> wrote:
> Hi guy,
>
> Am Mittwoch, 15. März 2017, 14:24:10 CET schrieb Danny YUE:
>> Hi guys,
>>
>> I just got Steam installed and running successfully on my machine,
>> and tried to get CS:GO running smoothly, which made me really happy :-D
> nice to hear, have fun with that!
>
>> However when Steam is downloading games, the speed is extremely slow,
>> down to several KB/s, even some bytes/s.
>> I have already installed dnsmasq and it *was* good during downloading
>> CS:GO (~4MB/s), but became slow again with Civilization V.
>>
>> I googled a lot but all point to installing dnsmasq, which I don't think
>> is really helpful since I already have done that...
> Just installing dnsmasq doesn’t change anything, and just starting it does
> neither. I *assume* you did set it up as a local DNS cache, but please provide
> some information about it:
> - the source of the information, i.e. link to the page
> - your setup:
>   * dnsmasq config
>   * /etc/resolv.conf
>   * other configs you think that matter here
>
>> Also I'm sure downloading region is correct.
>>
>> Anybody experienced the same issue with dnsmasq installed?
>> Any clue is welcome and thanks in advance.
>>
>>
>> Danny
>
> Nils

[gentoo-user] Steam downloading extremely

[Danny YUE]

Hi guys,

I just got Steam installed and running successfully on my machine,
and tried to get CS:GO running smoothly, which made me really happy :-D

However when Steam is downloading games, the speed is extremely slow,
down to several KB/s, even some bytes/s.
I have already installed dnsmasq and it *was* good during downloading
CS:GO (~4MB/s), but became slow again with Civilization V.

I googled a lot but all point to installing dnsmasq, which I don't think
is really helpful since I already have done that...

Also I'm sure downloading region is correct.

Anybody experienced the same issue with dnsmasq installed?
Any clue is welcome and thanks in advance.


Danny