ArtfulPup (& Upup-Bionic-Beaver-RC)

[jrb]

Hi PeeBee,
Thanks again for the great work! I thought I would try Woofing up a custom ArtfulPup of my own so I followed your excellent instructions. The only problem I had was with 1download. I kept getting:

Code: Select all

Downloading busybox-1.25.0-i686_s700.pet
  - from http://smokey01.com/peebee/artfulpup/extras/pet_packages-artfulpup
  failed (file not found on server or could not finish download)

I got around this by using pmirror to download http://smokey01.com/peebee/artfulpup/extras and temporarily ##ing out all packages from that source in DISTRO_PKGS_SPECS-ubuntu-artful. Thought you might like to know.

Cheers, J

[peebee]

Hi PeeBee,
Thanks again for the great work! I thought I would try Woofing up a custom ArtfulPup of my own so I followed your excellent instructions. The only problem I had was with 1download. I kept getting:

Code: Select all

Downloading busybox-1.25.0-i686_s700.pet
  - from http://smokey01.com/peebee/artfulpup/extras/pet_packages-artfulpup
  failed (file not found on server or could not finish download)

I got around this by using pmirror to download http://smokey01.com/peebee/artfulpup/extras and temporarily ##ing out all packages from that source in DISTRO_PKGS_SPECS-ubuntu-artful. Thought you might like to know.

Cheers, J

Wow - many thanks @jrb - you must be the first to try that.....

Apologies for the download failure - I really should have tested that - now fixed by correcting the directory structure on smokey01.com

Should work as advertised now (I hope!)

Cheers!
peebee

[jrb]

Should work as advertised now (I hope!)

Worked perfectly! Some of my mods took a bit of fiddling, but all went well in the end.

Thanks again, J

[ally]

post 'em up

[gyro]

I've just found an issue with the "/etc/init.d/udev" script.
An error about "ps" failing is reported in "/tmp/bootsysinit.log".
I've found that this line of code in "udev" fails:

Code: Select all

if ! ps --no-headers --format args ax | egrep -q '^\['; then

The 'ps' fails because in puppy it's a script that contains this line:

Code: Select all

[ "`which ps-FULL`" != "" ] && exec ps-FULL $@

Which fails because "which" is not in the PATH as defined in the "udev" script.
This could be fixed by changing the PATH definition in "udev", but I suspect that this script is borrowed from elsewhere, so best not to have a local patch.
Suggested fix is to change the offending line in "/bin/ps" to:

Code: Select all

[ "`/usr/bin/which ps-FULL`" != "" ] && exec ps-FULL $@

or

Code: Select all

[ -x /bin/ps-FULL ] && exec /bin/ps-FULL $@

Either of these removes the requirement for "which" to be in the PATH.
The patched "ps" should still work the same as before.

gyro

[peebee]

I've just found an issue with the "/etc/init.d/udev" script.
gyro

Thanks @gyro....

Pull request initiated on github to change woof-code/rootfs-skeleton/bin/ps:
https://github.com/puppylinux-woof-CE/woof-CE/pull/1123

[Marv]

ArtfulPup 17.11 +8 running sucessfully with the 4.14.11 32b PAE kernel from LxPupSc 18.01 +1T. glxgears FPS and hardinfo CPUhash and FPUraytracing benchmarks pretty much unchanged from the stock kernel on this i5. By the release notes and kernel changelog the 4.14.11 should have kpti backported into it. With the newer kernel, on the i5 laptops I do add the following two lines to rc.local to bring the brightness up on boot.

Code: Select all

brightness=`cat /sys/class/backlight/acpi_video0/max_brightness`
echo $brightness > /sys/class/backlight/acpi_video0/brightness

# inxi -bw
System: Host: puppypc1857 Kernel: 4.14.11-lxpup-32-pae i686 (32 bit)
Desktop: JWM 2.3.7 Distro: ArtfulPup 17.11
Machine: Device: laptop System: FUJITSU product: LIFEBOOK S761 serial: R2500459
Mobo: FUJITSU model: FJNB225
UEFI [Legacy]: FUJITSU // Phoenix v: Version 1.17 date: 03/14/2012
Battery CMB1: charge: 56.0 Wh 100.0% condition: 56.0/56.0 Wh (100%)
CPU: Dual core Intel Core i5-2520M (-HT-MCP-) speed: 2501 MHz (max)
Graphics: Card: Intel 2nd Generation Core Processor Family Integrated Graphics Controller
Display Server: X.org 1.19.5 drivers: modesetting (unloaded: fbdev,vesa)
tty size: 80x24 Advanced Data: N/A for root
Network: Card-1: Intel 82579LM Gigabit Network Connection driver: e1000e
Card-2: Qualcomm Atheros AR9287 Wireless Network Adapter (PCI-Express)
driver: ath9k
Drives: HDD Total Size: 64.0GB (25.4% used)
Weather: Conditions: 5 F (-15 C) - Clear Time: January 6, 11:52 AM CST
Info: Processes: 126 Uptime: 2 min Memory: 110.6/4897.3MB
Client: Shell (bash) inxi: 2.3.7

cheers,

[gyro]

I've just noticed another minor issue reported in "/tmp/bootsysinit.log".
It complains about line 80 of "rc.sysinit", but the actual problem is line 79.
The issue can be avoided by changing line 79 of "rc.sysinit" from:

Code: Select all

SWAPSIZEBYTES=`fdisk -l $ONESWAP | grep $ONESWAP | cut -f 5 -d ' '` #bytes

to

Code: Select all

SWAPSIZEBYTES=`fdisk -l $ONESWAP | grep -m1 $ONESWAP | cut -f 5 -d ' '` #bytes

The problem is that "fdisk" in artfulpup is just a link to "busybox fdisk" and wheras the real "fdisk" produces only 1 line containing $ONESWAP, "busybox fdisk" produces 2 lines containing $ONESWAP, resulting in $SWAPSIZEBYTES containing 2 lines of data, instead of just a number.
This causes line 80 to fail, so later code does not increase the size of the "tmpfs" by half the size of the swap partition.
e.g. I have 4GiB of memory so the "tmpfs" starts as 2GiB, but I also have a 4GiB swap partition, so "tmpfs" should end up as 4GiB.
Without the patch my "tmpfs" ends up at 2GiB, with the patch it ends up at 4GiB, as it should.

gyro

[peebee]

The issue can be avoided by changing line 79 of "rc.sysinit"
gyro

Pull request raised on github....

[peebee]

Update +10 - BUILD_FROM_WOOF='testing;358697d4;2018-01-10 22:32:18 +0800' + Ubuntu updates

[hamoudoudou]

Artfulpup is really a nice puppy, but i fail to run Gramps with it. (Dpup Stretch the same). Old version is better than no version at all. as emergency OBprecise allows me to continue and record may relatives, soldiers first world war...
Hobbies as genealogy need Puppy Linux, Keep tis app running please.. How ? Devs will find what is wrong, sure.
Gramps: genealogical research program
Where Python is installed would be to change , said zippytex, an efficient Gramps user (and Linux expert)

[backi]

Hi you guys !

Artful Pup is quite awful .
But........tried to install Audacious Audio-Player..... no chance .
Same for Vlc-player...... no chance .....

Is there any possibility to get them as a sfs-Package maybe ?.....somewhere ?

[hamoudoudou]

mediaplayers VLC Audacious :
If people on tahrpup want mediaplayers needing FFmpeg and with artfulpup mediaplayers needing QT , that is just what will be unavailable..
choose your Puppy looking what apps are default, lot of puppies have VLC.. Those will VLC will miss applications needing FFmpeg..
If puppy still want to be a light OS, it cannot afford to have doublons, or more for a same functionality. VLC will do what Audacious does, Audacious will not play videos.
Artfulpup is the best Puppy since Precise 5.7.1. IMO
Gnome-mplayer wll do as well as VLC, better than VLC. However if you want VLC and Artfulpup, i am sure we can succeed.. But you had better to choose tahrpup.. VLC is included.

[hamoudoudou]

success
Just install python pet provided here
Muggins : "Made from the latest sourceforge debian release. It needs the attached libdb-4.4.so.pet, plus Technosaurus's Python%2B-2.6.4-i486.pet."
Merci Muggins what would we do whitout devs as you, change our hobbies ?

[backi]

Hi hamoudoudou !

What i do appreciate on Vlc-Audacious-Smplayer is , they have some cool gimmicks like Equalizer ,Stereo-Effects for fine tuning sound.... in contrast to Gnome-Mplayer or mplayer ..
One can enhance Sound Quality dramatically . So this is (for me) a real Flaw in Artful Pup not getting Vlc/Audacious .
It seems ,for my demands , Tahrpup -Xenial/Debian Dogs will fit better .

Greetings !

[hamoudoudou]

with so many puppies, i would be strange to not find the one which fits your needs..
To make a fat Puppy with everything, or many Slim Puppies, that is the question.. But you can easily find slim or fat, by taking a glance in the forum.. I Thin PeeBee was in the thin option.
... puppies only have to not freeze after ten minutes. Arfulpup has been running for hours this morning with heavy tasks as ckecking scanned archives for old births, burials and wedding. Processors work hard. Light 49 never failed.
Csipesz likes to make huge Puppies, all included. Try them.. that will make him happy..

[peebee]

For VLC, download and install both:

http://ftp.nluug.nl/ftp/pub/os/Linux/di ... ko-7.0.sfs

and

http://ftp.nluug.nl/ftp/pub/os/Linux/di ... 6_s700.pet

[backi]

Hi peebee !

Thanks for the Tip !
Will give it a trial .

Greetings from Good ole Germany !

[stemsee]

Hi Peebee

Just trying artfulpup out. First impressions are great.

My touchscreen works with pinch to zoom, and two fingers for right click on desktop calling up menu. The cursor and digitizer are correctly aligned, without any further calibration, just as with DebianDog, FatDog and LightHouse64. I have not used a 32bit pup for a couple of years and this was the main reason! Now there is no reason not too. I wonder which 32bit pup first had this capability ... it wasn't slacko32/64, or studio1337 (based on slacko) or quirky. Maybe this is the first one Is it using libinput?

Cheers!
stemsee

[stemsee]

I've just found an issue with the "/etc/init.d/udev" script.
An error about "ps" failing is reported in "/tmp/bootsysinit.log".
I've found that this line of code in "udev" fails:

Code: Select all

if ! ps --no-headers --format args ax | egrep -q '^\['; then

The 'ps' fails because in puppy it's a script that contains this line:

Code: Select all

[ "`which ps-FULL`" != "" ] && exec ps-FULL $@

Which fails because "which" is not in the PATH as defined in the "udev" script.
This could be fixed by changing the PATH definition in "udev", but I suspect that this script is borrowed from elsewhere, so best not to have a local patch.
Suggested fix is to change the offending line in "/bin/ps" to:

Code: Select all

[ "`/usr/bin/which ps-FULL`" != "" ] && exec ps-FULL $@

or

Code: Select all

[ -x /bin/ps-FULL ] && exec /bin/ps-FULL $@

Either of these removes the requirement for "which" to be in the PATH.
The patched "ps" should still work the same as before.

gyro

what about

Code: Select all

busybox which ps-FULL

which in fatdog returns

Code: Select all

ps-FULL not found in `echo $PATH`

more or less; whereas `busybox which` when nothing found returns nothing. Both return one line when finding the exe. Seems to me to aid portability.

stemsee