BionicDog (updated: 2018-06-04)

A home for all kinds of Puppy related projects
Message
Author
User avatar
fredx181
Posts: 4448
Joined: Wed 11 Dec 2013, 12:37
Location: holland

#346 Post by fredx181 »

ITSMERSH wrote:Directory /etc/rc.d only includes rc.network, So when I'm creating /etc/rc.d/rc.local its content will be executed at boot?
No, /etc/rc.local is the one. The rc.network script is called from /etc/init.d/rc.network-start (is for peasywifi)
Pcmanfm doesn't provide the wallpaper - even after reboot.
I was wrong earlier, sorry, It's set in ~/.config/pcmanfm/default/desktop-items-0.conf

Fred

ITSMERSH

#347 Post by ITSMERSH »

fredx181 wrote:
ITSMERSH wrote:Directory /etc/rc.d only includes rc.network, So when I'm creating /etc/rc.d/rc.local its content will be executed at boot?
No, /etc/rc.local is the one. The rc.network script is called from /etc/init.d/rc.network-start (is for peasywifi)
Thanks, meanwhile I'd stumbled upon it. :wink:
fredx181 wrote:
Pcmanfm doesn't provide the wallpaper - even after reboot.
I was wrong earlier, sorry, It's set in ~/.config/pcmanfm/default/desktop-items-0.conf

Fred
No, this doesn't have the wallpaper either.

Code: Select all

[*]
wallpaper_mode=screen
wallpaper_common=1
wallpaper=/opt/docs/pictures/uk_glacial_lake_wales_gwynedd_county_snowdonia_region_100866_1920x1080.jpg
desktop_bg=#000000
desktop_fg=#ffffff
desktop_shadow=#000000
desktop_font=Sans 10
show_wm_menu=0
sort=mtime;ascending;
show_documents=0
show_trash=1
show_mounts=0

[abiword.desktop]
x=204
y=102

[DogRadio.desktop]
x=105
y=100

[gnome-mplayer.desktop]
x=103
y=179

[pcmanfm.desktop]
x=207
y=3

[firefox.desktop]
x=103
y=7

[PicDrop.desktop]
x=304
y=3
My wallpaper is /opt/docs/pictures/development.jpg and it is displayed after boot. And it is still in nitrogen's config.

Code: Select all

[:0.0]
file=/opt/docs/pictures/08.jpg
mode=4
bgcolor=#000000

[xin_-1]
file=/opt/docs/pictures/development.jpg
mode=0
bgcolor=#000000
Never mind.

Since nitrogens config stores it correctly and it is displayed correctly the file /root/.config/pcmanfm/default/desktop-items-0.conf doesn't seem to affect it.

I'm almost sure to recall that /opt/docs/pictures/uk_glacial_lake_wales_gwynedd_county_snowdonia_region_100866_1920x1080.jpg is the default wallpaper after a fresh install / at first boot.

Ok, two questions. Is there a boot option to boot without

- additional squashfs modules loaded
- to have the changes in 'casper/changes' loaded

??

stemsee

#348 Post by stemsee »

fredx181 wrote: @stemsee
I tried Wifi-TrayNet, sorry to say, but it didn't work for me, no connection could be made for wlan0.
But maybe it's me, I found some things very confusing and complicated, also the timeout dialogs gave me sort of a stress.
Here's output from terminal (I did several things after creating the profile, e.g. reconnect etc..):
Fred
It's definetely not you! It's the gremlins ... AGAIN! The only error you see in term that I could not explain is the yad, suggesting ampersands be escaped with '&' which made no difference!

Please see previous post 'EDIT'! Thanks for your patience!

User avatar
fredx181
Posts: 4448
Joined: Wed 11 Dec 2013, 12:37
Location: holland

#349 Post by fredx181 »

ITSMERSH wrote:
fredx181 wrote:
Pcmanfm doesn't provide the wallpaper - even after reboot.
I was wrong earlier, sorry, It's set in ~/.config/pcmanfm/default/desktop-items-0.conf

Fred
No, this doesn't have the wallpaper either.

Never mind.

Since nitrogens config stores it correctly and it is displayed correctly the file /root/.config/pcmanfm/default/desktop-items-0.conf doesn't seem to affect it.

I'm almost sure to recall that /opt/docs/pictures/uk_glacial_lake_wales_gwynedd_county_snowdonia_region_100866_1920x1080.jpg is the default wallpaper after a fresh install / at first boot.
There's small misunderstanding, pcmanfm has it's own wallpaper chooser, when selected Desktop mode is "pcmanfm + lxpanel" (through "Desktop Manager") and right-click on the Desktop "Desktop preferences" (separate from nitrogen).
Then the chosen wallpaper will be in /root/.config/pcmanfm/default/desktop-items-0.conf
Ok, two questions. Is there a boot option to boot without

- additional squashfs modules loaded
Yes, sort of, see here:
http://www.porteus.org/tutorials/26-gen ... -them.html
Put them in the optional folder and only specified modules will be loaded
cheatcodes wrote:load=module
load=module[1];module[n]

... Load optional modules from /optional/ directory on the booting
media. Additional modules can be listed, separated
by semicolons, without spaces.
Exampe: 'load=gimp;inkscape' would search for and load any
modules in the /porteus/optional folder that contain the strings
'gimp' or 'inkscape'.
- to have the changes in 'casper/changes' loaded
Just remove "changes=....." from kernel commandline for to boot without changes.

@stemsee, I'll test Wifi-TrayNet later and reply in your Wifi-TrayNet thread

Fred

ITSMERSH

#350 Post by ITSMERSH »

@fredx181

That's cool, thanks.

This will do some tasks a lot easier... :)

stemsee

#351 Post by stemsee »

I notice that neither /etc/xdg/Startup and /root/Startup are in path, nor referenced in rc.local, so how are the scripts in /root/Startup triggered?

can they be added safely with

Code: Select all

sed -i 's|:/usr/games"|:/usr/games:/etc/xdg/Startup:/root/Startup"|' /etc/profile
EDIT: found /usr/local/bin/start-up
would this be ok for xdg and backwards compatibility with puppy

Code: Select all

#!/bin/bash
# items are listed numbers first then alphabetical order
sleep 4
ls $HOME/Startup/* | while read J
do
   "$J" &
done
[[ -d /etc/xdg/Startup ]] && ls /etc/xdg/Startup/* | while read k
do
    "$k" &
done
[[ -d $HOME/my-applications/bin ]] && ls $HOME/my-applications/bin/* | while read l
do
     "$l" &
done

User avatar
fredx181
Posts: 4448
Joined: Wed 11 Dec 2013, 12:37
Location: holland

#352 Post by fredx181 »

stemsee wrote:would this be ok for xdg and backwards compatibility with puppy
Code:
#!/bin/bash
# items are listed numbers first then alphabetical order
sleep 4
ls $HOME/Startup/* | while read J
do
"$J" &
done
[[ -d /etc/xdg/Startup ]] && ls /etc/xdg/Startup/* | while read k
do
"$k" &
done
[[ -d $HOME/my-applications/bin ]] && ls $HOME/my-applications/bin/* | while read l
do
"$l" &
done
Startup from /etc/xdg/autostart (I assume you mean that, /etc/xdg/Startup does not exist ) only works with .desktop launchers, not for scripts or binaries.
Adding $HOME/my-applications/bin is good idea for backwards compatibility with puppy but just to be added to PATH, not to auto-start from IMO.

Fred

stemsee

#353 Post by stemsee »

I just researched the xdg-user-dirs specs and didn' find /etc/xdg/Startup, that seems to be a customisation in FatDog, i thought it was standard to xdg. /etc/xdg/autostart only takes .desktop files, so I didn't mean that.

stem

ITSMERSH

#354 Post by ITSMERSH »

Hi.

There seems to be an issue at reboot/shutdown in BionicDog after booting with 'changes' activated. The boot drive isn't cleanly unmounted. Need to do a e2fsck in a Puppy since I can't unmount the boot drive in BionicDog to do a e2fsck.

If I don't do a e2fsck (e.g. by overseen it before next boot), I get results similar to this:

Code: Select all

sh-4.3# e2fsck -y /dev/sda1
e2fsck 1.42.9 (4-Feb-2014)
Intern_1 was not cleanly unmounted, check forced.
Pass 1: Checking inodes, blocks, and sizes
Deleted inode 6291475 has zero dtime.  Fix? yes

Deleted inode 6291481 has zero dtime.  Fix? yes

Deleted inode 6291568 has zero dtime.  Fix? yes

Deleted inode 6291613 has zero dtime.  Fix? yes

Deleted inode 6291733 has zero dtime.  Fix? yes

Deleted inode 6291734 has zero dtime.  Fix? yes

Deleted inode 6291987 has zero dtime.  Fix? yes

Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
Pass 4: Checking reference counts
Pass 5: Checking group summary information
Block bitmap differences:  -(2832--2847) -(2912--2943) -(4416--4435) -(7456--7519) -(10112--10239) -(20176--20191) -(20240--20255) -(20320--20351) -(20768--20831) -(20864--20991) -(1999490--1999745) -(2017280--2018303) -(2018816--2019327) -(2282896--2282911) -(2282976--2283007) -(2283040--2283103) -(2283392--2283763) -(2284544--2285567) -(2285940--2286195) -(2288128--2288639) -(2289664--2289907) -(2359810--2359811) -(2359936--2359937) -(2360841--2360856) -(2492603--2492858) -(2493440--2494463) -(2497024--2497535) -(2697216--2703359) -(2721792--2745668) -(3015216--3015263) -(3015296--3015359) -(3015680--3016191) -(3022208--3022335) -(3022464--3022719) -(3022848--3023871) -(3024896--3047423) -(3049472--3080191) -(3082240--3112959) -(3125248--3126579) -(25167452--25167455) -(25176080--25176082) -(25186833--25186848) -(25255988--25255999) -25256449 -(25256544--25256564) -25258496 -(25259008--25259022) -(25259520--25259522) -(25262592--25262607) -25265231 -(25265272--25265279) -(25265288--25265295) -(25265792--25265800) -25265811 -(25265816--25265823) -(25265913--25265916) -(25265922--25265937) -25266013 -(25266168--25266171) -25266211 -(25266330--25266332) -(25266398--25266399) -25266408 -25266410 -(25266420--25266427) -(25266461--25266463) -(25266472--25266476) -25266488 -(25266566--25266567) -(25281025--25281040)
Fix? yes

Free blocks count wrong for group #0 (10844, counted=11360).
Fix? yes

Free blocks count wrong for group #61 (3581, counted=5373).
Fix? yes

Free blocks count wrong for group #69 (7906, counted=10426).
Fix? yes

Free blocks count wrong for group #72 (10441, counted=10461).
Fix? yes

Free blocks count wrong for group #76 (6015, counted=7807).
Fix? yes

Free blocks count wrong for group #82 (25961, counted=32105).
Fix? yes

Free blocks count wrong for group #83 (8377, counted=32254).
Fix? yes

Free blocks count wrong for group #92 (7694, counted=32254).
Fix? yes

Free blocks count wrong for group #93 (1534, counted=32254).
Fix? yes

Free blocks count wrong for group #94 (1534, counted=32254).
Fix? yes

Free blocks count wrong for group #95 (8108, counted=9440).
Fix? yes

Free blocks count wrong for group #768 (27147, counted=27170).
Fix? yes

Free blocks count wrong for group #770 (32099, counted=32168).
Fix? yes

Free blocks count wrong for group #771 (30800, counted=30903).
Fix? yes

Free blocks count wrong (14925749, counted=15049937).
Fix? yes

Inode bitmap differences:  -6291475 -6291481 -6291568 -6291613 -(6291733--6291734) -6291987
Fix? yes

Free inodes count wrong for group #768 (6617, counted=6624).
Fix? yes

Free inodes count wrong (9484941, counted=9484948).
Fix? yes


Intern_1: ***** FILE SYSTEM WAS MODIFIED *****
Intern_1: 132460/9617408 files (2.1% non-contiguous), 23413551/38463488 blocks

sh-4.3# 
sh-4.3# 

User avatar
rcrsn51
Posts: 13096
Joined: Tue 05 Sep 2006, 13:50
Location: Stratford, Ontario

#355 Post by rcrsn51 »

I cannot reproduce this.

I ran BionicDog with changes in play.

I then ran Stretch-Live in "clean mode" and did an e2fsck.

There were no problems.

ITSMERSH

#356 Post by ITSMERSH »

Yes, as I said: when I'm NOT doing (just overseen) a e2fsck immediately at next boot I get those results similar. If I do it immediately there's no problem, though the drive still remains not cleanly unmounted. So booting and rebooting from a drive not cleanly unmounted causes those errors.

But the main cause is: the drive isn't cleanly unmounted with changes used. If the drive were cleanly unmounted there would be no need to do a e2fsck and then there were no errors after doing reboots from that drive.

User avatar
fredx181
Posts: 4448
Joined: Wed 11 Dec 2013, 12:37
Location: holland

#357 Post by fredx181 »

Hi RSH, thanks for reporting, however I cannot reproduce either, also I'm not sure that I understand your last message.
Can you explain the steps in detail how to reproduce ?

Fred

wiak
Posts: 2040
Joined: Tue 11 Dec 2007, 05:12
Location: not Bulgaria

#358 Post by wiak »

Hi Fred,

Just noticed you've included IUPLua programming capability in the iso. That's great; I will have to move from XenialDog to BionicDog now and get back to IUPLua programming soonish. It is supposed to be easy calling C routines from Lua; if so, now that I've brushed up on my C again, I might be able to offer some nice tricks later.

As things stand, there are actually additional bolt-on module/batteries available for IUPLua which I have compiled but not yet packaged, which will be a nice addition for the repo - allow drawing graphs and much more.

But the core IUPLua stuff is enough for most simple gui building. Main thing will be to write some more exemplars on how to interface with underlying utilities (including shell). You may well find it proves more useful than YAD or gtkdialog/gtkwialog eventually; certainly more powerful and nice to encourage users to learn another programming language other than just bash.

I'm busy on gtkwialog still - but fixing up the <input> side of things to allow bash -c is alot of work and involves many source files; been trying various alternative means. I have to keep at it though or the C code required will evaporate from my head leaving too much rust again.

wiak

User avatar
fredx181
Posts: 4448
Joined: Wed 11 Dec 2013, 12:37
Location: holland

#359 Post by fredx181 »

Hi wiak,

Yes, IUPLua programming looks nice, I may give a shot at it later as it seems easier to learn than gtkdialog, but this period I'm feeling not very energetic to take on new stuff (sort of summer vacation laziness... :wink: )

Fred

hamoudoudou

Bonnes vacances !

#360 Post by hamoudoudou »

Bonnes vacances ! For all devs.. :)

ITSMERSH

#361 Post by ITSMERSH »

fredx181 wrote:Hi RSH, thanks for reporting, however I cannot reproduce either, also I'm not sure that I understand your last message.
Can you explain the steps in detail how to reproduce ?

Fred
I'm working on a program trying to get it to work in a Puppy as well as in BionicDog. Therefor I need to do frequently reboots between Puppy and BionicDog.

I'm booting BionicDog with some .sqashfs files loaded (needed for the program to test):

02-additional-fonts.squashfs
02-modifications-BionicDog.squashfs
03-usb-modeswitch.squashfs
10-gimp-2.8.22.squashfs
10-gimp-plugins-2.8.squashfs
15-additional-audio.squashfs
16-additional-audio-updates.squashfs
28-SunVox-1.7.5.squashfs
29-OcenAudio.squashfs

After the 'i386-linux-gnu' issue I built all of those .squashfs files manually in BionicDog, except:

28-SunVox-1.7.5.squashfs
29-OcenAudio.squashfs

I just renamed them from .sfs to squashfs. OcenAudio is placed in /opt, SunVox is placed in /usr/local/sunvox.

This is the boot menu entry:

title BionicDog clean SFS Testings, save on exit
find --set-root --ignore-floppies /BioDogClean/casper/initrd1.xz
kernel /BioDogClean/casper/vmlinuz noauto from=/BioDogClean changes=EXIT:/BioDogClean/casper/
initrd /BioDogClean/casper/initrd1.xz

When testing the program usually I do not save on exit, though 'changes' is used.
Last edited by ITSMERSH on Sat 23 Jun 2018, 01:39, edited 2 times in total.

ITSMERSH

#362 Post by ITSMERSH »

Ok, regarding my previous post above...

I've made two reboots to make sure it's not caused by the .sqashfs files built manually. So I disabled them all (moved them to another directory) to not to load at boot. Only loaded 99-bionic_linux-firmware.squashfs and 99-locales-BionicDog-2018-04-21.squashfs.

First boot I made with .sqashfs files disabled but 'changes' in place/use. Did not save on exit. Rebooted into Puppy and made a e2fsck.

Drive remains not cleanly unmounted.

I disabled directory 'changes' (renamed it and created a new empty one), after finishing e2fsck.

Rebooted BionicDog and saved on exit.

Rebooted Puppy and e2fsck still complains about a not cleanly unmounted drive.

I should have mentioned that I'm using the version of BionicDog pre the 2018-06-04 update.

ITSMERSH

#363 Post by ITSMERSH »

Ok, regarding my previous two posts above...

I created a new empty directory 'changes' and made again a reboot. Did not save on exit. Booted into Puppy and made a e2fsck.

Drive still remains not cleanly unmounted.

I don't know if I would need the 99-bionic_linux-firmware.squashfs to boot and use BionicDog for my needs. Though I would need the 99-locales-BionicDog-2018-04-21.squashfs, so trying to boot without any .squashfs loaded doesn't seem to make much sense to me.

Btw: how do they load? Are they 'mounted/loaded' from HD or are they copied into ram and mounted/loaded from ram?

I'm asking since they might 'lock' the drive from unmounting, if they are mounted/loaded from HD and if any of those .squashfs did not properly unmount/unload at shutdown/reboot.

This happens on a very few .sfs files in Puppy, that needs to have X restarted before unloading properly.

User avatar
fredx181
Posts: 4448
Joined: Wed 11 Dec 2013, 12:37
Location: holland

#364 Post by fredx181 »

Hi RSH, thanks for the infos.

I did a lot of investigating today, really like to solve this.
At first I couldn't reproduce (Bionicdog install on ext4) but then I thought trying to install BD on a ext2 formatted filesystem.
Then I got the same as you, so on ext3 and ext4 there's no "unclean unmounted" problem for me, on ext2 there is. Is your Bionicdog install on ext2 ?
Tried several things to solve, but I guess it's above my knowledge.
Any tips ? How Puppy does it to unmount everything cleanly at shutdown ?
As it is on Bionicdog the "umount -l ... " commands are in /usr/bin/wmreboot and /usr/bin/wmpoweroff.
Tried changing umount -l /mnt/live/memory/images/01-filesystem.squashfs to umount -l /mnt/live/memory/images/* (for to unmount all loaded modules), but made no difference.
Btw: how do they load? Are they 'mounted/loaded' from HD or are they copied into ram and mounted/loaded from ram?
If you add copy2ram parameter to the kernel boot commandline, then all modules will be copied to RAM.

For now I'm giving up on this, but I may get a EUREKA later :wink:

P.S. for what it's worth: With a frugal install of an official Debian Live (on ext2) I had the same "unclean unmounted" problem.
(thought it might be a specific problem with the BD porteus boot method, apparently not)

Fred

ITSMERSH

#365 Post by ITSMERSH »

Hi Fred.

Sorry to not to mention it, but yes, it is installed on ext2 partition. Usually all my bootable partitions are formatted to ext2. So I may change this in the future...

Older Puppies did not unmount automatically at shutdown/reboot, but newer Puppies seem to do.

They probably adopted this from my LazY Puppy ideas, which checked for mounted drives, isos, squash files and file systems on reboot/shutdown. It unmounts squash files and file systems first, since they could have been mounted from within a mounted iso. ISOs won't unmount otherwise. And drives won't unmount if there's still a iso mounted.

So, at least tahr puppy has this feature adopted. It also gives a message of 'unmounting stray file systems' if there's still anything mounted. If I'm faced to this message I know, I have to do a e2fsck directly after next boot.

Usually it is the Kdenlive SFS that is causing this problem. It unloads properly only after a restart of X. Sometimes I do oversee this...

Need to dive into the files to find the related code. Will report, when found...

Ok, parameter copy2ram could be a part of a solution on ext2. Will check this out and report.

Thanks

Post Reply