Create Debian 9 (Stretch) minimal ISO similar to DebianDog

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

#841 Post by fredx181 »

rcrsn51 wrote:Using the same libinput framework, here is an equivalent control panel for mice/pointers.

But this project isn't gaining any traction, let alone an actual test report, so I don't plan to spend any more time on it.
Not sure what you mean with "But this project isn't gaining any traction, let alone an actual test report", (may be my poor understanding english), anyway I tested pointer and works well to adjust mouse pointer speed, but initially the program wouldn't start on my system set to dutch locale:

Code: Select all

root@live:~# pointer

(gtkdialog4:12632): Gtk-CRITICAL **: IA__gtk_hscale_new_with_range: assertion 'step != 0.0' failed

(gtkdialog4:12632): Gtk-CRITICAL **: IA__gtk_range_set_value: assertion 'GTK_IS_RANGE (range)' failed
**
ERROR:variables.c:189:variables_new_with_widget: assertion failed: (widget != NULL)
/usr/local/bin/pointer: regel 107: 12632 Afgebroken              gtkdialog4 -c -p DIALOG
root@live:~# 
No idea why it didn't start, (we had some conversation about "grepping" commands on PM and the risk when using locale other than english, but looks to me this is different case), anyway this works well for me:

Code: Select all

LANG=C pointer
EDIT: Just remember now having issues in the past with scripts using some sort of "scale" or "range" option, that when other locale set, it has different output, but can't remember how/what exactly.

Fred

dancytron
Posts: 1519
Joined: Wed 18 Jul 2012, 19:20

#842 Post by dancytron »

I just tried the new Firefox.

It comes as a tar.bz file. I just unzipped it to a folder on my ext4 drive and clicked the firefox icon and it opened. No run as root problems. Seems to run fine. I went to netflix and it asked me to enable DRM. I said yes and it downloaded a bunch of stuff. Video played, but no audio. It told me I needed to install pulse audio. I stopped there for now, but I assume that isn't a huge issue.

In my short test, it did seem lighter and faster than Chrome. Might have potential for us.

I'll mess with it some more later.

https://www.mozilla.org/en-US/firefox/

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

#843 Post by fredx181 »

dancytron wrote:I just tried the new Firefox.

It comes as a tar.bz file. I just unzipped it to a folder on my ext4 drive and clicked the firefox icon and it opened. No run as root problems. Seems to run fine. I went to netflix and it asked me to enable DRM. I said yes and it downloaded a bunch of stuff. Video played, but no audio. It told me I needed to install pulse audio. I stopped there for now, but I assume that isn't a huge issue.

In my short test, it did seem lighter and faster than Chrome. Might have potential for us.

I'll mess with it some more later.

https://www.mozilla.org/en-US/firefox/
Thanks!
If you install latest apulse (from rcrsn51 with menu entry "Firefox-apulse") and copy extracted from firefox-57.0.tar.bz2 "firefox" folder to /opt (so you have directory /opt/firefox/), you can have audio working by running from menu 'Firefox-apulse', type "about:config" in firefox adress bar , then search for "security.sandbox.content.write_path_whitelist", double click it and type "/dev/snd/" (without the quotes), then quit firefox , then by running "Firefox-apulse" from menu, the audio works, from what I tested.

See also here:
http://murga-linux.com/puppy/viewtopic. ... 650#971650

Fred

dancytron
Posts: 1519
Joined: Wed 18 Jul 2012, 19:20

#844 Post by dancytron »

Okay.

I just ran a new ./mklive-stretch adding pulse audio and pasting firefox into /opt/ during the script. I'll see how that works and if it doesn't I'll try what you suggest.

I'll have to go to the doctor tomorrow (nothing big) so it might be a day or 2 until I have something to report.

Dan

edit: okay, that didn't work. I'll try what you suggested next.

Dry Falls
Posts: 616
Joined: Tue 16 Dec 2014, 23:37
Location: Upper Columbia

#845 Post by Dry Falls »

fredx181 wrote:Can anyone test ?
(just recently I found when building mplayer there's special configure option for it to make it run on other machines also, but for mpv I couldn't find such option, so not sure it works on other machines)
Hi Fred. Made a slight path change for compatibility and installed your mpv build in lighthouse. needed libuchardet, liblua5 I got from packages.org. Worked right off the bat. Played everything I asked it to. Very impressive. I've adopted it for JL64, if you don't mind.

df

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

#846 Post by fredx181 »

Dry Falls wrote:
fredx181 wrote:Can anyone test ?
(just recently I found when building mplayer there's special configure option for it to make it run on other machines also, but for mpv I couldn't find such option, so not sure it works on other machines)
Hi Fred. Made a slight path change for compatibility and installed your mpv build in lighthouse. needed libuchardet, liblua5 I got from packages.org. Worked right off the bat. Played everything I asked it to. Very impressive. I've adopted it for JL64, if you don't mind.

df
OK, Nice!!

dancytron
Posts: 1519
Joined: Wed 18 Jul 2012, 19:20

#847 Post by dancytron »

fredx181 wrote:
dancytron wrote:I just tried the new Firefox.

It comes as a tar.bz file. I just unzipped it to a folder on my ext4 drive and clicked the firefox icon and it opened. No run as root problems. Seems to run fine. I went to netflix and it asked me to enable DRM. I said yes and it downloaded a bunch of stuff. Video played, but no audio. It told me I needed to install pulse audio. I stopped there for now, but I assume that isn't a huge issue.

In my short test, it did seem lighter and faster than Chrome. Might have potential for us.

I'll mess with it some more later.

https://www.mozilla.org/en-US/firefox/
Thanks!
If you install latest apulse (from rcrsn51 with menu entry "Firefox-apulse") and copy extracted from firefox-57.0.tar.bz2 "firefox" folder to /opt (so you have directory /opt/firefox/), you can have audio working by running from menu 'Firefox-apulse', type "about:config" in firefox adress bar , then search for "security.sandbox.content.write_path_whitelist", double click it and type "/dev/snd/" (without the quotes), then quit firefox , then by running "Firefox-apulse" from menu, the audio works, from what I tested.

See also here:
http://murga-linux.com/puppy/viewtopic. ... 650#971650

Fred
I followed your instructions and everything works now. I installed apulse when I ran ./mklive-stretch and pasted in firefox at the pause in the script.

I'll probably stick with Chrome for daily use, but I think this will work well for a lot of people.

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

#848 Post by fredx181 »

Hi All

*** Some tips for making the system boot faster and/or running the system more "light" *** (e.g. less RAM usage) on a frugal install

The following options will have most effect if you do all together, but just doing only one of them might improve a bit for making the system boot faster and/or running the system more 'light'

1) First: To convert the main '01-filesystem.squashfs', while running the system, (e.g. to LZ4, see below), upgrade the package 'edit-sfs-thunar' or edit-sfs-pcmanfm' to latest version: 2.4
(previous versions gave error -already in use- when trying to convert '01-filesystem.squashfs' from frugal install)
- Convert the main module '01-filesystem.squashfs' to LZ4, if it isn't already (by default it may be "xz" or "gzip"),
Use "Open with Convert-SFS" from right-click menu from Thunar or PcmanFm on live/01-filesystem.squashfs, or from commandline, e.g.:

Code: Select all

conv-sfs /path-to-frugal-install/live/01-filesystem.squashfs
And choose LZ4 compression, when done, replace live/01-filesystem.squashfs (better first make backup) with the new LZ4 compressed module and reboot.
Using LZ4 will make boot time shorter and run the system more 'light'

2) Boot with systemd, add to your kernel boot commandline: init=/bin/systemd
This will make boot time much shorter (and has the systemd configuration enabled, some people may like that, some not)

3) Install 'slim' login-manager (if already installed but not enabled yet, run "Start/Stop Slim display-manager" from Menu, to enable)
If slim isn't already configured to autologin and you want that, edit /etc/slim.conf to e.g:

Code: Select all

default_user        root
...
auto_login          yes
This will make boot time shorter and there's 'true' multi-user support then.

Additional info: When using "porteus-boot" option changes=EXIT:/.... (changes saved only at shutdown, or when running "save2flash" in terminal), the system runs faster, as the changes made are running completely in RAM (which is very fast)
But if you have small amount of RAM it's not recommended (may be not sufficient space, e.g. when installing lots of packages).

Fred

backi
Posts: 1922
Joined: Sun 27 Feb 2011, 22:00
Location: GERMANY

#849 Post by backi »

Hi Fred !

Good Infos .
Additional info: When using "porteus-boot" option changes=EXIT:/.... (changes saved only at shutdown, or when running "save2flash" in terminal), the system runs faster, as the changes made are running completely in RAM (which is very fast)
But if you have small amount of RAM it's not recommended (may be not sufficient space, e.g. when installing lots of packages).
This can be circumvented by booting with a second Boot Menu Entry Option when doing Ram intensive Downloading/Installing.
This second menu.lst Entry looks like this :

title DEBIAN_DOG- Direct changes to /DEBIAN_DOG/live/changes
find --set-root /DEBIAN_DOG/live/vmlinuz
kernel /DEBIAN_DOG/live/vmlinuz noauto from=/DEBIAN_DOG changes=/DEBIAN_DOG/live/
nitrd /DEBIAN_DOG/live/initrd1.xz

So every Change during Session will be saved automatically to changes Folder instantly .......so no overflow of ram during Session .( If i understood this Boot Concept correctly ).
When your Job is done ......you can reboot with your changes=EXIT:/
menu.lst ---- Boot-Entry again .

So always have a Backup of your changes Folder .
In case somethings is not to your liking or something went wrong ,just replace new changes Folder with the old One .

Regards !
Last edited by backi on Mon 20 Nov 2017, 16:39, edited 2 times in total.

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

#850 Post by rcrsn51 »

I have posted touchpad v1.2 above. It now configures touchpad/mouse/touchscreen devices in one app and handles more options like button mapping.

I have also addressed the locales issue by changing the slider controls to integer values. This needs testing.

belham2
Posts: 1715
Joined: Mon 15 Aug 2016, 22:47

#851 Post by belham2 »

Hi Fred & all,

I've got some questions that I was faced with last night, and I previously thought I knew the answers but after thinking about them today, thought it best to ask you guys here to make sure.

A week ago, I set my neighbor up with a DDog build, using the mklive script. He really likes it. So last night, he calls to tell me that on his main computer in the house (ironically running dual boot, Windows 10 and Debian 9 Stretch full-install) that he thought he'd been hit with a bad malware.

So, he unplugged this possibly infected hard drive, fired up the USB 'frugal-installed' DDog I gave him and proceeded to plug that possibly malware Harddrive in while DDdog was booting. He said he then used Gparted to wipe it so as to re-install stuff. Which he did.

My first thought was, yeah, all that's fine & no worries....but then I started wondering: by plugging in that infected harddrive to the Ddog (while it was booting and leaving it in as he used Gparted to wipe it), could that infected drive have infected DDog before it was formatted? He says he did not click on the drive when he plugged it in (which means he didn't "mount" it, right?)...but dang, he plugged it in while DDog was in the midst of booting up.

So, my questions:

1) how does Debian overall even know that the drive, even if it doesn't get officially mounted, is there? The Debian OS had to communicate with it, yes? How? What kind of info was exchanged?? Could malware jump across during this exchange of info??

2) would not this question not only apply to Debian, but possibly all the Puppies too? If you know for sure you've an infected drive (where most of us would probably use a booted CD to DD command wipe it & then format it after), is it safe to plug it in the DDogs and Pups to either DD-wipe it and/or use Gparted to format it & such? Or are we risking infection since at some level Debian/DDogs and Pups are communicating with the drive(s) even though they are not officially "mounting" them.


Thanks for any help in understanding.

zagreb999
Posts: 567
Joined: Fri 11 Apr 2014, 06:39
Location: Yugoslavija

puppyrus boots in 3 sec.

#852 Post by zagreb999 »

HI FRED,

GOOD INFO
DO YOU KNOW
THAT puppyrus boots in 3 sec.

REGARDS.

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

#853 Post by fredx181 »

belham2 wrote:1) how does Debian overall even know that the drive, even if it doesn't get officially mounted, is there? The Debian OS had to communicate with it, yes? How? What kind of info was exchanged?? Could malware jump across during this exchange of info??
Perhaps better I should leave this question answered by someone with more knowledge about infections, but I'd say if the infected drive was never mounted while running (Debian)Dog (or Puppy), there could be no danger.
But it depends on the kernel commandline, if (using porteus-boot) "noauto" was used or not (if not, all drives found would be mounted automatically)

Fred

belham2
Posts: 1715
Joined: Mon 15 Aug 2016, 22:47

#854 Post by belham2 »

fredx181 wrote:
belham2 wrote:1) how does Debian overall even know that the drive, even if it doesn't get officially mounted, is there? The Debian OS had to communicate with it, yes? How? What kind of info was exchanged?? Could malware jump across during this exchange of info??
Perhaps better I should leave this question answered by someone with more knowledge about infections, but I'd say if the infected drive was never mounted while running (Debian)Dog (or Puppy), there could be no danger.
But it depends on the kernel commandline, if (using porteus-boot) "noauto" was used or not (if not, all drives found would be mounted automatically)

Fred

Hi Fred,

Ok, that helps greatly. I am 99.99999% sure when I gave him his USB for his DDog, I did not use porteus-boot. I used the other one, where changes are saved "live". So me thinks maybe I better get his USB (and that drive he re-installed stuff too) and do it right. Wish he would have asked me first....I would have went over & booted one of my DDogs in RAM, and we'd have nothing to worry about. Hope he doesn't get irked or mad that he's got to re-install everything to the harddrive again (how many times does Windows 10 let people do that??) :cry:


P.S. Fred, one question about Debian overall. By default, Debian never mounts anything, right? Unless you first put it in the fstab? Is that correct or am I not thinking of it correctly??

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

#855 Post by fredx181 »

backi wrote:So every Change during Session will be saved automatically to changes Folder instantly .......so no overflow of ram during Session .( If i understood this Boot Concept correctly ).


Thanks backi, indeed no "overflow of ram during Session" when not using EXIT:/...
But can you edit your post ?, your example is not correct and might lead to confusion, better make it "DEBIAN_DOG" (without space in it) and from=/DEBIAN_DOG (only capitals)

Fred

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

#856 Post by fredx181 »

belham wrote:I am 99.99999% sure when I gave him his USB for his DDog, I did not use porteus-boot.
So you used "live-boot" ? (with "boot=live" in kernel command line) if so, then no auto mounting has been done, so even less worries, I'd say. (only porteus-boot automounts if you leave out the "noauto")
P.S. Fred, one question about Debian overall. By default, Debian never mounts anything, right? Unless you first put it in the fstab? Is that correct or am I not thinking of it correctly??
Yes, as far as I know that's correct.

Fred

belham2
Posts: 1715
Joined: Mon 15 Aug 2016, 22:47

#857 Post by belham2 »

fredx181 wrote:
belham wrote:I am 99.99999% sure when I gave him his USB for his DDog, I did not use porteus-boot.
So you used "live-boot" ? (with "boot=live" in kernel command line) if so, then no auto mounting has been done, so even less worries, I'd say. (only porteus-boot automounts if you leave out the "noauto")
P.S. Fred, one question about Debian overall. By default, Debian never mounts anything, right? Unless you first put it in the fstab? Is that correct or am I not thinking of it correctly??
Yes, as far as I know that's correct.


Fred

Thanks, Fred. I think I am going to tell him he's ok then. But I sure would feel better, lol, if he would ask me next time anything like this happens. Hopefully it never happens to him again.

User avatar
rufwoof
Posts: 3690
Joined: Mon 24 Feb 2014, 17:47

#858 Post by rufwoof »

How about adding a twm version to the current set of build choices Fred?

It has after all recently celebrated its 30th Birthday since Tom LaStrange sat down to write it. Whilst the default config looks lame, with a little refinement of the configuration (single file) and it can look much better.

The small size but great power is oft overlooked and for functional setups twm can work really well, leaving you with missing options when you use other windows managers

Here's a animated gif of my twm setup

Under OpenBSD after a base install I just add (pkg_add ... similar to apt-get), firefox-esr, libreoffice, mpv, xfe, mtpaint and osmo ... and that pretty much fills my needs.

twm by rights is a integral part of X11 i.e. conceptually is included with all X11 releases, however some refine X11 to exclude twm.

Not the prettiest of wm's, but easily configured and it does a single job well aka Unix style.

I've attached my .twmrc (gzip compressed)
Attachments
.twmrc.gz
(3.16 KiB) Downloaded 302 times

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

#859 Post by fredx181 »

Hi All,

Added to build options in case of 32bit build, choice for to install pae or no-pae kernel
For no-pae it will install (version number as it is now):
Package: linux-image-4.9.0-4-686 (4.9.51-1)

Linux 4.9 for older PCs

The Linux kernel 4.9 and modules for use on PCs with one or more processors not supporting PAE.
So added to one of the questions at start:

Code: Select all

Do you want to install (default) pae or a no-pae kernel ?
(no-pae is more suitable for older (with not pae capable cpu) PCs)
Install no-pae kernel ? (Enter = pae, y = no-pae) (y/N)?
There was request for this on an issue at github:
https://github.com/DebianDog/MakeLive/issues/1

Also available as a .tar.gz, which can be extracted in the frugal install 'live' folder
Including:
- .squashfs module including Debian non-pae kernel 4.9 (with aufs included)
- initrd1.xz (non-pae)
- initrd.img (non-pae)
- vmlinuz1 (non-pae)
(extracting will replace initrd1.xz, initrd.img and vmlinuz1 with no-pae versions, adviced is to make backups first)
https://debiandog.github.io/MakeLive/ke ... pae.tar.gz

New mklive-stretch:
mklive-stretch script

@rufwoof, Thanks! I'll have a look at twm

Fred

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

#860 Post by fredx181 »

rcrsn51 wrote:I have posted touchpad v1.2 above. It now configures touchpad/mouse/touchscreen devices in one app and handles more options like button mapping.

I have also addressed the locales issue by changing the slider controls to integer values. This needs testing.
Thanks, added to repos, I can confirm that slider controls work fine now with my system set to dutch locale.

Fred

Post Reply