BusterDog + build system (no-systemd)

A home for all kinds of Puppy related projects
Message
Author
User avatar
rcrsn51
Posts: 13096
Joined: Tue 05 Sep 2006, 13:50
Location: Stratford, Ontario

#141 Post by rcrsn51 »

Force meaning Connect even when there is no wlan0.
Where are you getting the information that there is no wlan0?

To know for sure, run: iwconfig

Edit It may be that this wifi adapter is initially in a "blocked" state and is hidden from the kernel. So the firmware doesn't get loaded.

When you start PeasyWiFi, it runs an "rfkill unblock" operation which would enable the adapter. Then the kernel could load the firmware.

Then the wlan0 interface would appear.

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

#142 Post by fredx181 »

@all, added to Busterdog repository (EDIT: and also to StretchDog repo): package yalarm , 'Yalarm Alarm scheduler'
Install deb:
https://github.com/doglinux/busterdog/r ... .1_all.deb
Or install 'yalarm' with apt-get or synaptic
Run from Menu > Accessories

See also 'Yad - Tips' thread, here:
http://murga-linux.com/puppy/viewtopic. ... 94#1045094
(and earlier posts about it)

EDIT: @enrique, I made this specially for you, so... if the alarm goes off, you must go for a walk, or: go to sleep, or: whatever different from sitting at the computer !!
(no, just kidding :D (mostly talking to myself :wink: )

Fred
Attachments
2019-12-19-171733_765x404_scrot.png
yalarm
(82.98 KiB) Downloaded 444 times

enrique
Posts: 595
Joined: Sun 10 Nov 2019, 00:10
Location: Planet Earth

#143 Post by enrique »

@fredx181 I was reporting what I originally thought was an error of BusterDog Full. A driver load failure that prevented wlan0 to be "Active". What ever that means. I have limited to none experience with IPv6 as I did run my previous system with IPv6 disabled. Now It seems to me from the fact that you guys are not surprise that this may be normal. From what I notice the driver would not even load its firmware until IPv6 gives the ok. That is what it seems to me.

@rcrsn51 I am pretty sure there is more to this procedure that it is new to me. "wifi adapter is initially in a "blocked" state and is hidden from the kernel." or if ""rfkill unblock" operation which would enable the adapter." I do not know. But I will study your info provided. It is very appreciated. But I will post no more related to this as it is seems it is a false error. You guys seems to be fine with its operation.

I will be testing this alarm. And yes, your analysis that I spent a lot of time in PC is correct. There are many projects I do run using PC. It is my hobby and way to relax. I am fine, my comment on the other thread was an attempt to stop the negative Read Between the Lines type of comment I was receiving. It will not affect my future behavior on the forum or this tread. Well it will make me more passive. Thanks my friends.

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

#144 Post by fredx181 »

enrique wrote:@rcrsn51 I am pretty sure there is more to this procedure that it is new to me. "wifi adapter is initially in a "blocked" state and is hidden from the kernel." or if ""rfkill unblock" operation which would enable the adapter." I do not know. But I will study your info provided. It is very appreciated. But I will post no more related to this as it is seems it is a false error. You guys seems to be fine with its operation.
You can check with "rfkill list", for me it's unblocked:

Code: Select all

root@live:~# rfkill list
0: phy0: Wireless LAN
	Soft blocked: no
	Hard blocked: no
1: hp-wifi: Wireless LAN
	Soft blocked: no
	Hard blocked: no
If it'd be hard blocked I need to unblock with the hardware switch on my laptop.
(maybe peasywifi can take care of that too, not sure)

Fred

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

#145 Post by rcrsn51 »

fredx181 wrote:(maybe peasywifi can take care of that too, not sure)
Not unless it can activate a robot arm to reach out and flip the WiFi keyboard switch for you.

enrique
Posts: 595
Joined: Sun 10 Nov 2019, 00:10
Location: Planet Earth

#146 Post by enrique »

Ok this time will be last post. Just to clarify. I have a experience in Android Arm kernel, TL-WR703N Openwrt / AR9331 mips, various satellites boxes and Lubuntu with IPv6 disabled. In any of those I had experience with IPv6 nor have seen that a kernel Driver will load half its firmware. Just to see it finish loading firmware at user space time. So this is a surprise to me.

Then all this continue change of things by kernel.org and distros. Here in particular ifconfig. At the moment ifconfig will not show an adapter that is NOT Active. This and the fact that the drivers where not loaded confuse me. But this is all a misunderstanding.

if we use ip a instead then I can see all adapters and their status. So just doing ifconfig wlan0 up will allow brcmsmac to load its to missing drivers and then it show up in ifconfig. Finally to me it is funny see Ipv6 announcing status.

Code: Select all

[  196.507734] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready
***
[  197.507303] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
That is all. Sorry for the confusion. I wonder if we can still disable iPv6?

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

#147 Post by rcrsn51 »

rcrsn51 wrote:To know for sure, run: iwconfig

User avatar
mauros
Posts: 32
Joined: Mon 02 Dec 2019, 19:15
Contact:

#148 Post by mauros »

enrique wrote: I wonder if we can still disable iPv6?
You can disable IPV6 but in some cases (i.e. using VPN service which masks IPV6) it will block your connection. Check it and make your choice.

1. Open with a text editor as root the file /etc/sysctl.conf (or create a new .conf file in /etc/sysctl.d directory and open it).

2. Put the configuration lines to disable it globally:
net.ipv6.conf.all.disable_ipv6 = 1
net.ipv6.conf.default.disable_ipv6 = 1
net.ipv6.conf.lo.disable_ipv6 = 1
or to disable it for wlan0 only:
net.ipv6.conf.wlan0.disable_ipv6 = 1
and save your flat file.

3. Run as root:

Code: Select all

sysctl -p
To ensure it is disabled run:
globally:

Code: Select all

cat /proc/sys/net/ipv6/conf/all/disable_ipv6
for wlan0 only:

Code: Select all

cat /proc/sys/net/ipv6/conf/wlan0/disable_ipv6
returns "1" if you disabled it successfully or "0" if you don't.

If you have connection issues with VPN or other services, remove the configuration lines to re-enable IPV6.

Diogenes08
Posts: 4
Joined: Mon 18 Sep 2017, 21:17

BPO Kernel

#149 Post by Diogenes08 »

Hey, I am just wondering if there is any way/plans to install the BPO kernel, like there was in StretchDog/DevuanDog, ie upgrade-kernel2?
I recently got a new computer, a 2 in 1, and I am determined to use BusterDog on it. All works well, except the touchscreen, which from what I can gather, works on 4.19.8 and up.
Edit: Nevermind, once I installed to HD, the touchscreen worked perfectly out of the box. Thanks once again for making the distro that stopped me from nearly a decade of distro hopping.
Last edited by Diogenes08 on Sun 22 Dec 2019, 18:43, edited 1 time in total.

User avatar
mauros
Posts: 32
Joined: Mon 02 Dec 2019, 19:15
Contact:

#150 Post by mauros »

fredx181 wrote: Here's also a "package" for the 5.3 backports kernel with aufs included (built with aufs-dkms_5.3), contains 5.3.0-0.bpo.2-amd64.squashfs, vmlinuz1, initrd1.xz, initrd.img .
https://github.com/DebianDog/BusterDog/ ... d64.tar.gz
Fred

enrique
Posts: 595
Joined: Sun 10 Nov 2019, 00:10
Location: Planet Earth

#151 Post by enrique »

mauros thanks. I only want to see how the driver for wlan0 behaves with no ipv6. Understood VPN may fail, understandable as the use your IPv6 and https information to identify your uniqueness in the web. You can try to hide but will never do. I remember old times when we did reset our routers so that we could obtain a new IP address. No https at the time mean we were a new user. Not really this days.

Diogenes08
Posts: 4
Joined: Mon 18 Sep 2017, 21:17

#152 Post by Diogenes08 »

I am not sure if I am doing something wrong, but when I try to build using the mklive build script with a config file, it fails to download the desktop specific configs, and from what I can tell always downloads the default with pcman, lxpanel, etc regardless of the preset I used to make the config. (In my case, if it matters, the ddog preset.)

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

#153 Post by fredx181 »

Diogenes08 wrote:I am not sure if I am doing something wrong, but when I try to build using the mklive build script with a config file, it fails to download the desktop specific configs, and from what I can tell always downloads the default with pcman, lxpanel, etc regardless of the preset I used to make the config. (In my case, if it matters, the ddog preset.)
Yes, looks like you found a bug, I guess you did similar as this ??:

Code: Select all

./mklive-buster -gui ddog.conf
I could reproduce that it doesn't create a full DDog as it should, but not that it's downloading pcmanfm, lxpanel, so not sure what you did exactly, please tell me.
To explain about the bug: The thing is that when running just:

Code: Select all

./mklive-buster -gui 
Then there's choice of Desktop environments and the variable DE_CONFIG will be set.
In case what you probably did, the variable DE_CONFIG is not set and it fails to download and extract the module https://debiandog.github.io/MakeLive/mo ... m.squashfs (in case of ddog.conf (and DE_CONFIG="DDog")
I fixed that now by adding DE_CONFIG= to most configs (where required) here:
https://debiandog.github.io/MakeLive/configs-buster/
And from what I tested it works as expected now.

So... what you can do is download ddog.conf from above url and rebuild using that, or:
Add: DE_CONFIG="DDog" to your existing ddog.conf and rebuild, or:
Download "zz_openbox_xfce-jwm.squashfs" and put it in the live folder of your current (incomplete) build .

Hope this helps.

Fred

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

#154 Post by fredx181 »

New BusterDog-openbox_jwm ISO'S version: 2019-12-28

Changes and Fixes:
- Fixed the annoying delay of the openbox menu display.
- Some other small changes/fixes for menu(s)
- Upgraded Palemoon to latest and upgraded other software (apt upgrade)
- Added some more applications.

Download here:
https://github.com/DebianDog/BusterDog/ ... /README.md

Fred

keniv
Posts: 583
Joined: Tue 06 Oct 2009, 21:00
Location: Scotland

#155 Post by keniv »

Hi Fred,
Have installed new version with clean save folder. All seems to be working. The menu appearance is now much faster. Have not had much time to play but have noticed the addition of you transtray program.

Regards,

Ken.

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

#156 Post by fredx181 »

BusterDog-openbox_jwm-2019-12-29 version:
Made a small mistake yesterday, one change:
- Set timezone to (default) UTC (forgot earlier)
No big deal really, IMHO the default should be UTC.
But when there's an internet connection at boot, it will be set to your location anyway (automatically detected).

Download:
https://github.com/DebianDog/BusterDog/ ... /README.md

Fred

jbv
Posts: 179
Joined: Sat 01 Jan 2011, 00:22

#157 Post by jbv »

Hi,

Does anyone know why the package gnome-disk-utility (gnome-disks) does not start, and is there a simple fix?

Another one that has me stumped, when booting into:
"Devuan-live-boot-3 in RAM persistence"
Keyboard and Mouse don't work (Logitech wireless kb/mouse)
same kb/mouse work fine in the "(no save)" boot menu option, and all other boot variants

Advance thanks.

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

#158 Post by fredx181 »

jbv wrote:Does anyone know why the package gnome-disk-utility (gnome-disks) does not start, and is there a simple fix?
For some mysterious reason the package "libpolkit-gobject-1-0" is not properly installed (on 64-bit Busterdog), so, I've found that gnome-disk-utility should work when reinstalling libpolkit-gobject-1-0 :

Code: Select all

apt update # if required
apt install --reinstall libpolkit-gobject-1-0
Re-uploaded latest 64-bit ISO version including that fix:
https://github.com/DebianDog/BusterDog/ ... 64-bit.iso

Regarding your other question, I have no clue.

Fred

jbv
Posts: 179
Joined: Sat 01 Jan 2011, 00:22

#159 Post by jbv »

Hi Fred, that fixed it thanks.

I have almost worked out what is locking up when I try to boot "Devuan-live-boot-3 in RAM persistence".

It is something to do with the copy to ram option. I'm not sure what it is and I'm not going to bother any more as it seems that the persistence works differently than I am used to anyway, as the *cow directory is always empty, so all of my scripts are useless. It is definately not a memory thing, as the machine has 16Gb of RAM, but it also has 26 discs and that may be giving it grief.

Interestingly none of the Devuan Boot options seem to get a dynamic IP and the only way to get a static IP is to put it on the kernel append line (after you've setup everything else.
This may be a side-effect of having Frizbee and the other network modules, although I am pretty sure I've got most of them turned off.
Once again, not something I intend to persue.

There was one little glitch with the new ISO you made.
It requires a "changes.dat" file in /live, without it, the system won't load and hangs. Fortunately an empty file is enough to get it going.

I have one more issue that I can't work out, and would appreciate some assistance with if you don't mind. It seems that the aufs filesystem is not working in the Porteus Boot(s) - either one.

No matter what I've tried, I can't get an aufs overlay to work.
It works fine with the Devuan live boots.

I have tried adding "union=aufs" to the boot line as follows, without any joy.

Code: Select all

label My Buster (save changes /live/changes.dat)
kernel /live/vmlinuz1
append initrd=/live/initrd1.xz noauto from=/ copy2ram changes=/live/changes.dat union=aufs
I have created a simple script that can quickly and easily show the issue:

Code: Select all

#!/bin/bash
Pause ()
{ 
  read -n1 -r -p "Press any key to Continue ..."
}

mkdir /tmp/test1
mkdir /tmp/test2
mkdir /tmp/merged
mount -t aufs -o dirs=/tmp/test1:/tmp/test2 none /tmp/merged

echo "***"
echo "*** in a moment /tmp/merged will be empty"
echo "***"
sleep 5
ls -al /tmp/merged
Pause

date > /tmp/test1/file1
date > /tmp/test2/file2
echo "***"
echo "*** in a moment /tmp/merged will contain files from /tmp/test1 and /tmp/test2"
echo "***"
sleep 5
ls -al /tmp/merged
Pause

umount /tmp/merged
rm -f -r /tmp/test1
rm -f -r /tmp/test2
rm -f -r /tmp/merged
sync
sync
sync
sleep 5
The above works perfectly with either of the Devuan-live boots.
What am I missing?
How do I get aufs to work properly with the Porteus boot?

Thanks and Regards, Brenton

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

#160 Post by fredx181 »

Hello Brenton,
What am I missing?
How do I get aufs to work properly with the Porteus boot?
As far as I know, on the 'virtual' filesystem it doesn't work, you need to do it on an 'actual' filesystem or on tmpfs, this should work ( you may want to change the "MNTDIR" variable):

Code: Select all

#!/bin/bash

#  change as desired, must be an 'actual' filesystem (e.g. /mnt/sda2) or tmpfs e.g. /mnt/live/tmp
MNTDIR=/mnt/live/tmp        

Pause ()
{
  read -n1 -r -p "Press any key to Continue ..."

}

mkdir -p $MNTDIR/test1
mkdir -p $MNTDIR/test2
mkdir -p $MNTDIR/merged

mount -t aufs -o dirs=$MNTDIR/test1:$MNTDIR/test2 none $MNTDIR/merged
echo "***"
echo "*** in a moment $MNTDIR/merged will be empty"
echo "***"
sleep 5
ls -al $MNTDIR/merged
echo "***"
Pause

date > $MNTDIR/test1/file1
date > $MNTDIR/test2/file2
echo "***"
echo "*** in a moment $MNTDIR/merged will contain files from $MNTDIR/test1 and $MNTDIR/test2"
echo "***"

sleep 10   # increased to 10
ls -al $MNTDIR/merged
echo "***"
Pause

umount $MNTDIR/merged
rm -f -r $MNTDIR/test1
rm -f -r $MNTDIR/test2
rm -f -r $MNTDIR/merged
sync
sync
sync
sleep 5 
The directory /tmp is not a tmpfs filesystem (it used to be in earlier Debian versions but not anymore)
I had to increase the second sleep to 10, otherwise (on my system) the "ls -al $MNTDIR/merged" command doesn't show the files.

EDIT: Strange that your code works with live-boot and not with porteus-boot, the reason may be that with live-boot you're booting with "overlayfs" rather than aufs, then it's a different case and aufs may work mounting in /tmp (not sure, I didn't test)

Fred

Post Reply