Fatdog64 700 alpha1

[mavrothal]

It does not auto-detect and use swap partitions.

By design.

Interesting.
Could you elaborate on the rational?

Changing SFSs loaded results in loosing user-altered screen resolution.

How could this happen?

I have no idea but,
change resolution, reboot. OK.
Add some SFSs, reboot, default resolution.
Change resolution again, reboot. OK.
Remove some SFS, same story again.
In general I find that openbox in various "puppy-like" systems is problematic in keeping screen resolution (at least in my VMs)

Latter: It would appear that is the qt sfs that generates the screen resolution problem.

[jamesbond]

Interesting. Could you elaborate on the rational?

For the typical user, there is a "Create swapfile" in control panel -> systems menu.
For a power user, they can always add the location of swap partition to /etc/fstab and it *will* be mounted at boot.
Using a swap partition on a bad harddisk can make the system unresponsive, so if you're trying to recover a bad disk, and the OS auto-mounts the swap partition on that bad disk - then you're toast.
If you *really* need this - a few lines in rc.local to iterate over /sys/class/block, with guess_fstype and swapon will do the trick.

Changing SFSs loaded results in loosing user-altered screen resolution.

How could this happen?

I have no idea but,
change resolution, reboot. OK.
Add some SFSs, reboot, default resolution.
Change resolution again, reboot. OK.
Remove some SFS, same story again.
In general I find that openbox in various "puppy-like" systems is problematic in keeping screen resolution (at least in my VMs)

Intriguing. How do you change the resolution? Do you use xrandr, or do you use the control panel -> desktop -> display properties?

[mavrothal]

Interesting. Could you elaborate on the rational?

For the typical user, there is a "Create swapfile" in control panel -> systems menu.
For a power user, they can always add the location of swap partition to /etc/fstab and it *will* be mounted at boot.
Using a swap partition on a bad harddisk can make the system unresponsive, so if you're trying to recover a bad disk, and the OS auto-mounts the swap partition on that bad disk - then you're toast.
If you *really* need this - a few lines in rc.local to iterate over /sys/class/block, with guess_fstype and swapon will do the trick.

Thanks.
But I was mostly wandering about the rational of *not* auto-detecting a linux swap partition that apparently the user created to be used with the given hardware.

How could this happen?

I have no idea but,
change resolution, reboot. OK.
Add some SFSs, reboot, default resolution.
Change resolution again, reboot. OK.
Remove some SFS, same story again.
In general I find that openbox in various "puppy-like" systems is problematic in keeping screen resolution (at least in my VMs)

Intriguing. How do you change the resolution? Do you use xrandr, or do you use the control panel -> desktop -> display properties?

The control panel.

You may have missed the (latter) point about the qt sfs doing that.
Also tested kernel source sfs and iron sfs do not affect the resolution.

BTW iron needs libnss3 so trying to start iron from sfs will fail silently.
Installing libnss with gslapt does the trick.

[Bindee]

Does Fatdog have samba package included ?

[jamesbond]

Thanks.
But I was mostly wandering about the rational of *not* auto-detecting a linux swap partition that apparently the user created to be used with the given hardware.

Because the system can't read the user's mind. In a typical Linux installation (Debian, Centos, etc) - you create (or assign an existing) swap partition when you install the OS. The OS just picks up the settings you specified during installation - it doesn't auto-detect, because a presence of a swap partition doesn't always mean that the user wants to use it.

You may have missed the (latter) point about the qt sfs doing that.

Yeah, I missed it. But how could you remove the qt-sfs, because it is locked? You can only remove it by exiting to terminal, run "load_sfs.sh /path/to/qt.sfs unload" and then restart X. Anyway, I'm going to do exactly as you say - and see if I can reproduce the problem.

Also tested kernel source sfs and iron sfs do not affect the resolution.

That's a relief

BTW iron needs libnss3 so trying to start iron from sfs will fail silently.
Installing libnss with gslapt does the trick.

That works, but the bug is in rc.sysinit. libnss3 is present in either firefox or seamonkey; if there is seamonkey but no firefox LD_LIBRARY_PATH should be pointed there (which isn't, at the moment). Thanks for pointing this out.

@Bindee - 700 comes with a super-old Samba (3.0.37). The reason is that it is small, and it fits the bill for doing simple file sharing. If you need the recent samba4, you can get it from the repo.

cheers!

[Bindee]

@Bindee - 700 comes with a super-old Samba (3.0.37). The reason is that it is small, and it fits the bill for doing simple file sharing. If you need the recent samba4, you can get it from the repo.

cheers!

" Critical code execution bug in Samba gives attackers superuser powers "

http://arstechnica.com/security/2014/08 ... er-powers/

This was posted today.

[Gobbi]

Running LiveCD of Fatdog700a1 on HP laptop :

Processor 4x Intel(R) Core(TM) i3-3120M CPU @ 2.50GHz
Memory 6009MB (244MB used)
Machine Type Physical machine
Operating System Fatdog64 [700]
User Name root (root)
Date/Time Tue Aug 5 15:47:33 2014
Display
Resolution 1366x768 pixels
OpenGL Renderer Unknown
X11 Vendor The X.Org Foundation
Audio Devices
Audio Adapter HDA-Intel - HDA Intel PCH

Kernel Linux 3.15.5 (x86_64)
Version #1 SMP Thu Jul 10 09:07:23 EDT 2014
C Library GNU C Library version 2.19 (stable)
Distribution Fatdog64 [700]

It uses an Intel HD4000 graphics .

The wireless button is blocked -OFF - in his native OS ( W8 ) since last night , it's Ralink 5390R adapter has not a proper driver even after The original driver from HP was reinstalled .
Here in FD700a1 the module for a Ralink2800 adapter works great. Also video and sound are working OTB .

The laptop is 1 year old and it has Secure Boot disabled .

[jamesbond]

" Critical code execution bug in Samba gives attackers superuser powers "

http://arstechnica.com/security/2014/08 ... er-powers/

This was posted today.

That's just great
Anyway, the built-in samba isn't affected (the vulnerability is only for Samba 4.x). I'll get the one in the repo updated.

[mavrothal]

You may have missed the (latter) point about the qt sfs doing that.

Yeah, I missed it. But how could you remove the qt-sfs, because it is locked? You can only remove it by exiting to terminal, run "load_sfs.sh /path/to/qt.sfs unload" and then restart X. Anyway, I'm going to do exactly as you say - and see if I can reproduce the problem.

The qt sfs is unloaded (ie is not loaded) after reboot, and is then when the screen resolution setting is lost

[gcmartin]

Thanks goes out the FATDOG team for the progress you've undertaken.

I saw this:

1. Install dnsmasq package.
2. Edit /etc/dnsmasq.conf.netboot as needed
3. Populate the TFTP root directory with the files you want to serve (e.g. pxelinux, etc) (the old guide still applies).
4. Start dnsmasq like this "dnsmasq -C /etc/dnsmasq.conf.netboot"

And the reference to the guide: Step 4, above, is a slight update that does NOT exist in the guide, but, is a valid step, nonetheless. (Assuming you are referencing the original "5-step PXE guide" put together awhile, ago)

Tested by adapting the above recommendation. FATDOG is a special, advanced PUP.

I would, respectfully, like to request that DNSMASQ be included as a part of the OOTB base of the FATDOG distro because of the several areas it brings great value to use. Not just in this, but in other areas of use of the FATDOG distro productively in local, LAN, and VM hosting. Its cost in the ISO size is so small while its value is great in home use of this distro. Assuming its presence, nothing needs installation for the user to begin taking advantage of the benefit it allows. Just some procedural steps.

Thanks, in advance, for any consideration as you move forward.

[Jim1911]

Kirk and James,

Thank you for this great new update of Fatdog64. It booted fine right off and has been working great. The only problem that I've run into is that Xsane does not recognize my HP Photosmart Printer. It does print fine.

I've used a few sfs from 600 that are working super. Libreoffice-4.3.0_64_en-US_xz.sfs (prepared with get libreoffice) and Wine-1.5.4-i486.sfs with 32bit-slacko-1.sfs which works well with some 32bit applications.

When you have time please consider two additional sfs for those of us that have problems building their own.

1. Latest Bibletime
2. Latest Firefox

Thanks,
Jim

[Bindee]

" Critical code execution bug in Samba gives attackers superuser powers "

http://arstechnica.com/security/2014/08 ... er-powers/

This was posted today.

That's just great
Anyway, the built-in samba isn't affected (the vulnerability is only for Samba 4.x). I'll get the one in the repo updated.

Thanks James.

[neerajkolte]

Hello everyone,

1] @mavrothal, I rebooted many times with the qt sfs loaded and unloaded, set to load at boot also with set to not to load at next boot.
But didn't have any problem with resolution of any of my two displays.
Can anyone else confirm that.
Can it be something to do with vmware.
I boot from usb with savefile and base sfs on harddrive.

Also I can unload qt sfs live, just had to make sure vlc was not running.

2] Can any one tell me how to make timeout for entering kernel parameters at boot Zero.
My syslinux.cfg reads

Code: Select all

say Type "fatdog" with kernel parameters, or just press Enter for default boot.
default fatdog
timeout 0
prompt 1
label fatdog
kernel vmlinuz
initrd initrd
append waitdev=0 savefile=ram:device:sda3 basesfs=device:sda3

Timeout was 50 first, I changed it to 0, but it still waits for me to press Enter.
Edit: I interpret the syslinux wiki incorrectly...
It said timeout 0 will desable timeout.
I thought that would mean it would boot directly, it actually waits indefinitely....
So timeout 1 it is.
I get to working desktop on 10th second after I press Power ON.


3] Can it be done such as, unless I kept some key pressed, it won't ask me for kernel parameters.
Edit: I found if I kept key pressed before the boot prompt comes, it doesn't register. I have to press it after boot prompt to disable timeout.
May be I could use timeout 5, half a second should be enough.

Thanks.

- Neeraj.

[James C]

Another frugal install on an older box. Working well ootb.

Finally able to boot without the nouveau.noaccel=1 parameter.

Code: Select all

-Computer-
Processor		: 2x AMD Athlon(tm) 64 X2 Dual Core Processor 3800+
Memory		: 3791MB (334MB used)
Machine Type		: Physical machine
Operating System		: Fatdog64 [700]
User Name		: root (root)
Date/Time		: Wed Aug  6 01:02:33 2014
-Display-
Resolution		: 1440x900 pixels
OpenGL Renderer		: Unknown
X11 Vendor		: The X.Org Foundation
-Audio Devices-
Audio Adapter		: HDA-Intel - HDA NVidia

Looking really good.

[Billtoo]

I used the Fatdog installer to install to a 32gb flash drive.

My older mac keyboard is working fine on this laptop.
Fatdog 700a runs great on this single core laptop.

[eowens2]

I have been having a heckava problem keeping my computer wireless connected to the router, under both Fatdog64_630/631 and 700a1. Sometimes the wpa_gui will cycle repeatedly from connect => disconnect => scan and back.

I have the Atheros wifi chipset with the ath9k driver, and after googling awhile, I conclude this hardware has a history of problems.

I found kirk's post from 7 Nov 2010 (murga-linux topic 61619) regarding the use of /usr/sbin/WIFI-Keepalive. I am not sure of the correct usage of this, or if it needs any arguments. Anyway when I enter it in the terminal it returns: eth0 no wireless extensions, lo no wireless extensions. And it doesn't seem to make much difference in the stability of the connection.

Interestingly Slacko5.7.0 and Ubuntu 14.04 on the same hardware have given me no problem thus far.

Any help would be appreciated.

[avarghese]

I believe 'dnsmasq' is an essential thing that should be included with FatDog Live CD.

We're using FatDog as a DHCP server, in LAN and hence 'dnsmasq' is a must. Otherwise we've to rely on the heavy weights like 'ISC-DHCP-Server'. Setting up it itself is a difficult task, as you can see in my below blog.

http://my-techno-arena.blogspot.in/2014 ... dog64.html

[jamesbond]

@mavrothal - I can't reproduce the resolution screwed-up problem.

@neeraj - set "prompt 0" and it will not stop to prompt you - it will only prompt you if you press shift/alt-key.

@james c/@billtoo: sounds great!

@jim1911 - will look into that. Remind me if you don't hear from me after a while, please.

@gcmartin & @avarghese - ok, due to popular demand, dnsmasq will be added to base sfs.

@eowens2: try turning off power management first: "iwconfig wlan0 power off".
To use /usr/sbin/WIFI-Keepalive, just launch it like this: "nohup /usr/sbin/WIFI-Keepalive &" (let it run in the background).

[Billtoo]

Still on my laptop.

Just to see if it might work I downloaded shinbar's ffconvert-1.4.2.pet
and unpacked it in the Download directory, navigated to the bin
directory and clicked on ffconvert.
After doing the setup it works!
I did a test screen capture of Firefox.

[mavrothal]

@mavrothal - I can't reproduce the resolution screwed-up problem.

Could be only in the VMware VM, but I did verified again that this is the case.

actually when it happens (loading or unloading the qt sfs) and you change the resolution with zarfy and then restart X the new resolution does not stick. However upon reboot the new resolution takes effect
Looking at the contents of the savefolders could not see anything that would justify the resolution change

BTW where does zarfy writes the changes of the resolution?