Fatdog64-620 Final (17 April 2013) and 621 (9 May 2013)

[Ted Dog]

hi all.
I found the readme file in the initrd.

I intend to compile all of them, for the 32bit version. static linked.
is there anything that can give you the correct way how to use the aboriginal cross compiler or uclibc buildroot. because I've tried them but always failed.

Code: Select all

Contains:
- busybox 1.19.4 (most applet) (uclibc 0.9.33, buildroot 2012.02)
- busybox init 1.19.4 (uclibc, br 2012.02), compiled separately for memory reasons
- dropbear (uclibc, br 2012.02)
- guess_fstypes 20130215 (musl, patched for 64-bit, SFS and LUKS detection, from barryk/jesse)
- losetup-klibc (losetup from klibc 2.0.1)
- ntfs-3g 2012.1.15 (uclibc, br 2012.02)
- wpa_supplicant (uclibc, br 2012.02)
- iwconfig (uclibc, br 2012.02 not used at the moment)
- sdparm 1.07 (can't remember I used musl or uclibc or klibc)
- lvm 2.02.98 w/o udev - (uclibc, aboriginal cross compiler 1.2.0)
- cryptsetup 1.6.0 (uclibc, aboriginal cross compiler 1.2.0)
- e2fsck 1.42.6 (uclibc, aboriginal cross compiler 1.2.0)
- dosfsck 3.0.13 (uclibc, aboriginal cross compiler 1.2.0)
- mount.cifs 1.10-apple187 (uclibc, aboriginal cross compiler 1.2.0)
- mdadm 3.2.6 (uclibc, aboriginal cross compiler 1.2.0)
- full fledged init, supports coldplug and hotplug
- barebone rc.sysinit & rc.cleanup

Thanks in advance.

I also would like to use same type of setup. If anyone in the 64bit world can translate/ guide us to the aboriginal methods within puppy would be richly thanked. I just can't make sense of aboriginal directions.

[Gobbi]

Thank's for answering kirk & Q5sys ! Yes , I think it's a closed loop for me ... At least Fatdog runs fine without a drive anyway , and if I shall ever need to look inside the drive , testing Red Hat , but not buying it , would be an option ...

Billtoo : reading a bit more the posts in forum I found a command that works for my case :
vblank_mode=0 glxgears

[Gnuxo]

I have a quick question.

I plugged in my generic eink ebook reader and Fatdog doesn't recognize it.

How do I get it mounted?

...And also is Calibre available for Fatdog?

[WillM]

Hi Gnuxo,

Calibre has an automatic installer script on their website. It requires python from the Fatdog64 repository.
Leave the sudo part out when copying the script. Default install is to /opt.
http://calibre-ebook.com/download_linux

[Gobbi]

Thank's for the tip WillM . I did try this before but it didn't work at that time ... ( It was without python's dependencies )
I installed python , with all the dependencies , and then the automatic script did everything . I was pleased to see the program starting ...

[Ted Dog]

Nice, I'm a ebook junky, just got the paperwhite its awesome, to bad they do not have a kindle bookreader for linux, even with almost all versions of the hardware use linux kernels.

[jamesbond]

Again, the primary change, this time, was to NOT install all of the PETs required for this version to replace my older FATDOG520

What were the pets you installed that caused the startup to fail?

I have set the login graphical and as root with a new password,

I still get the same message. Can only install as root...

That, even if I do log in as root.

Don't do it from the browser. In Fatdog the browser runs as unprivileged user "spot" and doesn't have the rights to install pets.
Download it, use Rox to open the folder that contains it, then (double-)click it.

To elaborate on my mouse issues, which for some reason have now transferred to 6.0.1. I am unsure if this is even related to fatdog anymore. Anyway, when the OS boots my mouse and keyboard lights go off.

For whatever it is worth, I had specific problems with a particular mouse and a particular laptop. Fatdog+that laptop+that mouse = not working. "dmesg" would show messages as if the mouse as constantly plugged and unplugged. Change any of the combination - works (ie, Fatdog+mouse = works on another laptop, or Fatdog+laptop = works with another mouse, or Windows+laptop+mouse = works)... Go figure

missing cdrtools source used in FD64v620 in http://distro.ibiblio.org/fatdog/source/c/

Ted, I think James has that, but he's out of town for a while.

Thanks kirk.
Yes, sorry I forgot. The source is here: ftp://ftp.berlios.de/pub/cdrecord/alpha ... 13.tar.bz2 and I have uploaded it to /source/c too just now.

There are two separate questions here: 1) desired behavior, in terms of searching subdirectories;

I remember we discussed this before but google refused to help me anyway, Fatdog by default only searches the root directory of every drive. If you want it to search sub-directories as well, add the "search=n" parameter (root directory is "level 1", so to search one directory depth, you use search=2). Remember that searching is slow and that's the main reason why Fatdog only searches at the root directory. When you install Fatdog you're to know where the savefile is, and it is best to tell Fatdog about it. And if you don't know how to create the obscure boot parameter - as rcrsn51 point out, there is a useful tool in the control panel.

2) consistency between separate user interactions. Having a GUI for creating a save file doesn't help much if you need to write obscure bootloader code to use it. Shouldn't the default action be to look in the directory where the kernel or initrd resides? If the user explicitly chooses to place the savefile elsewhere couldn't a flag be passed to indicate this?

No, because there isn't any logical place to keep this information. In fact, as far as I know, there is no way for a running kernel to know the original location of its own "vmlinuz" binary ... (other than by hacks and by searching *all* disks - but then it can still fail to find it ...)

If a flag file already exists next to the kernel and initrd, the new location could be added to a list for the user to choose from at boot time. There would be no searching for locations not used by some savefile.

It actually involves searching as well, that is, searching for that flag file. In addition to be unreliable - searching is, as discussed above, slow.

This would be particularly useful for testing multiple versions, as with the experimentation going on now with proprietary drivers.

With grub, grub4dos or syslinux, one can always specify the default boot parameters and edit it / modify it at every boot.
Yes it is probably more troublesome than if the handholding is done by Fatdog itself (by a menu etc), but I would prefer to focus on fast everyday performance rather than ease of experimentation - after all (for me at least) I use Fatdog as my *primary* operating system.

First off, I discovered that I can't choose any of the boot options - it doesn't yet respond to my USB keyboard at that stage of the boot process...

Flash, what's the brand of your keyboard? Fatdog has a few drivers for generic HID devices built-in, but they keep inventing new ones...

Also, I can no longer copy by highlighting text in Geany, then paste the highlighted text into a console window (either rxvt or xterm) by simultaneously clicking the left and right mouse buttons.

There is a way. You need to install xinput, available from x_misc_tools from the package manager. Basically you can use xinput to enable the 3button-emulation on the fly. I can guide you with the rest of the steps later on if you're interested.

is there anything that can give you the correct way how to use the aboriginal cross compiler or uclibc buildroot. because I've tried them but always failed.

This should be discussed in the Programming forum, but anyway ... the aboriginal cross compiler is here: http://www.landley.net/aboriginal/downloads/binaries/ and how to use them is explained here: http://www.landley.net/aboriginal/downl ... ies/README. How to use Buildroot is explained here: http://buildroot.busybox.net/downloads/ ... anual.html. The binaries were built with unpatched original sources.

cheers!

[Billtoo]

@kirk

There is a new proprietary nvidia driver:
NVIDIA-Linux-x86_64-319.17.run that I downloaded, then ran the
getnvidia utility but it stops with an error saying it needs the
kernel source sfs which is already loaded, see screenshot.
I got the kernel sources sfs that's on ibiblio and the checksum is good, it's dated apr. 13th I think.

I have the nvidia-310.44-3.8.7.pet installed already.

I tried compiling the new driver but it gives an error that it can't
create a symlink, I stopped it at that point and uninstalled it which
got the previous driver working again.

[scsijon]

I have created a 64bit JWM Package for Fatdog64 and Slackbones64 at Version 750 for people to test on my jwm thread.
http://www.murga-linux.com/puppy/viewto ... 002#680002

Please read the note at the top if the first message (dated 7 may) for some important information.

There are quite a few changes and improvements between what your current one is and the new one, so it could be well worth scanning the few pages of the thread before installing.

Any reports relating to JWM and the build should go in the JWM thread with a title including '64bit' please, so I know which build I'm dealing with.

Also Joe monitors and ocassionally answers in the thread,

Hopefully a jwmconfig2 will appear soon to match your systems.

regards
scsijon

[kirk]

BillToo,

I haven't used the getnvidia utility, maybe James has some ideas with that. I uploaded a 319.17 pet if you want to try it:

http://distro.ibiblio.org/fatdog/pets/6 ... -3.8.7.pet

[dancytron]

Nice, I'm a ebook junky, just got the paperwhite its awesome, to bad they do not have a kindle bookreader for linux, even with almost all versions of the hardware use linux kernels.

There is a Kindle app for Chrome which works in puppy (at least in Lucid and Precise).

[Billtoo]

BillToo,

I haven't used the getnvidia utility, maybe James has some ideas with that. I uploaded a 319.17 pet if you want to try it:

http://distro.ibiblio.org/fatdog/pets/6 ... -3.8.7.pet

The new pet is working great, I find that the proprietary driver works
better than the nouveau driver when using two monitors.
Thanks very much for the pet.

[gcmartin]

Edited: Sorry for this question to forum. It appears as dm-0 on the desktop. I mis-understood that icon's reference. The system accurately references the LVM

A boot time to desktop question
I booted by selecting the "...LVM and MD...", then hitting tab, and adding savefile=direct:multi.
The system boots fine to desktop. On this system, I list the HDD and start the SWAP as follows:

Code: Select all

# fdisk -l

Disk /dev/sda: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders, total 312581808 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x2afbc966

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1              63    20466809    10233373+  12  Compaq diagnostics
/dev/sda2   *    20466810    64002959    21768075    7  HPFS/NTFS
/dev/sda3        74220300   312576704   119178202+   f  W95 Ext'd (LBA)
/dev/sda4        64002960    64412559      204800   83  Linux
/dev/sda5       104952645   210162329    52604842+   7  HPFS/NTFS
/dev/sda6       210162393   230645204    10241406   83  Linux
/dev/sda7       230645760   261357567    15355904    7  HPFS/NTFS
/dev/sda8       261361548   304190774    21414613+   7  HPFS/NTFS
/dev/sda9       304190838   312576704     4192933+  82  Linux swap / Solaris
/dev/sda10       74220363   104936266    15357952   8e  Linux LVM

Partition table entries are not in disk order

Disk /dev/dm-0: 9684 MB, 9684647936 bytes
255 heads, 63 sectors/track, 1177 cylinders, total 18915328 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

Disk /dev/dm-0 doesn't contain a valid partition table

Disk /dev/dm-1: 6039 MB, 6039797760 bytes
255 heads, 63 sectors/track, 734 cylinders, total 11796480 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

Disk /dev/dm-1 doesn't contain a valid partition table
# swapon /dev/sda9

But, the desktop only list the following partitions, below, for mounting.

Question

• Should I see the LVM as an available for mounting?

Answer: YES! its the dm-0 icon reference in desktop.

[prehistoric]

...There is a Kindle app for Chrome which works in puppy (at least in Lucid and Precise).

Do we have Chrome for Fatdog 6.20? If so I missed it.

There is a Kindle Cloud reader app for Firefox which should work, though I have not tried it.

Myself, I don't feel like reading while sitting at a keyboard. What I've done is to put Kindle for Android on my B&N Nook Color, after upgrading to Android 4.1.2 with CyanogenMod.

[Gobbi]

Trying FD620 on an Acer Aspire 5738ZG .It has an ATI Mobility Radeon 4650 . Booted at first try . Zarfy show only 1024/768 and 800/600 resolutions available.
I attach the hard-info...

[kirk]

Trying FD620 on an Acer Aspire 5738ZG .It has an ATI Mobility Radeon 4650 . Booted at first try . Zarfy show only 1024/768 and 800/600 resolutions available.
I attach the hard-info...

That report doesn't show a radeon module loaded, which is needed for you video card, or you would need the proprietary fglrx module. I think the radeon 4650 is well supported by radeon, check that you don't have something in Grub, or whatever boot loader, that would cause radeon not to load and try booting with savefile=none in case you have something configured wrong.

[nooby]

http://www.murga-linux.com/puppy/viewtopic.php?t=86053
Especially this IP 198.101.241.44

most likely something innocent? Forget I mention it here

[Gobbi]

That report doesn't show a radeon module loaded, which is needed for you video card, or you would need the proprietary fglrx module. I think the radeon 4650 is well supported by radeon, check that you don't have something in Grub, or whatever boot loader, that would cause radeon not to load and try booting with savefile=none in case you have something configured wrong.

That report was not of downloaded FD620.iso , but from a remaster made on an HD 5770 adding ati_13.4 pet . I realized the error late , but I went to bed and today I post the right report for the downloaded version . I didn't have the time to try to add later the correct driver because I had to give it back to it's owner . Sorry for missleading...

[gcmartin]

In running the Hardinfo tool for local use on the PC, I found couple issues. I share them here. And, in doing so I say/suggest that these are very very minor items as all information is contained in the reports (cli/desktop).

cli messages

Code: Select all

# hardinfo -r > /tmp/hardinfo_report.txt
Computer
 Summary
 Operating System
 Kernel Modules
 Boots
 Languages
 Filesystems
 Display
 Environment Variables
 Development
 Users
 Groups
Devices
 Processor
 Memory
 PCI Devices
 USB Devices
 Printers
 Battery
 Sensors
 Input Devices
 Storage
 DMI
 Memory SPD
 Resources
Network
 Interfaces
 IP Connections
 Routing Table
 ARP Table
 DNS Servers
 Statistics
error parsing /proc/net/snmp: Success
 Shared Directories
sh: net: command not founds...
Benchmarks
 CPU Blowfish
 CPU CryptoHash
 CPU Fibonacci
 CPU N-Queens
 FPU FFT
 FPU Raytracing
 GPU Drawing
# 

Thus, its messages in Shared Directories section (above) and the info in the RAM section (below) not appearing.

Again, this is minor.

Here to help

[gcmartin]

Again, the primary change, this time, was to NOT install all of the PETs required for this version to replace my older FATDOG520

What were the pets you installed that caused the startup to fail?

The attached Picture belows shows the FATDOG PETs installed and the servers listings. Remember: The servers were enabled and started before reboot with a save-session to the Live media.

This is the behavior I have using the initial use of the 620-Final ISO with a Live media. ALSO, remember that in an earlier post I share that when ONLY SAMBA was addressed with ONLY the SAMBA Extras PET, then, after enabling and starting before shutdown with save-session to media, ONLY then, did the server(s) behave as expected. From that extra step at the beginning with another Live media disc forward, did the servers maintain expected behavior on all subsequest save-sessions and reboots.

The ISO's md5 matches and the base system files, on each of the 2 Live media disc, also match with each other.

I will continue to save the "bad" LiveDVD for another week or so before retiring it.

Any additional information or tasks I can provide to assist problem understanding/discovery, I will do per request.

Here to help