Fatdog64-700/701 [April 22 2015] [CLOSED]

[quirkian2new]

I have tried to use a savefile on Fatdog631 with this 2nd hand laptop , when a savefile is used, the result is the same as using all other puppy linux or MS Windows, i.e. Fatdog631 CANNOT block the autoshutdown, and so each session can only last 3 to 8 minutes. I think Fatdog701 with savefile could NOT block the autoshutdown also (even though i have not tested this). Other puppy linux , like precise5.7.1 , with "create savefile dialogue box " cannot prevent autoshutdown.

Thus my rubbish-like laptop cannot run any OS more than 10 minutes. It can ONLY run FD631 or FD701. (FD631 each session can last 3-4 hrs) (FD701 each session seems to be less than 3 hrs).

As I bought this 2nd hand at very low price, I do not want to spend any extra money for dealing with whatever overheating issues (which in my humble opinion, it doesn't seem like actual overheating. I would more be convince it may be a hardware bug).

By the way, this 2nd hand laptop has NO battery when I bought it, and I am not going to buy battery for it. So now my kids will use the desktop with windowXP, and then I will use FD631 and FD701 with this laptop.

[quirkian2new]

dear Kirk & Jamesbond, recently i read the quirky7 thread and notice that rufwoof had managed to make all physical ram into zram and then load the q.sfs file all into ram and running from ram.

http://murga-linux.com/puppy/viewtopic. ... 7&start=45


I am now using fatdog631 and fatdog701. My second hand laptop has 2gb ram, after boot up with FD631 or FD701(boot with small initrd), the available ram is about 980mb or so. It is hoped that you will add the above zram feature into fatdog701 and later versions of fatdog.
(i.e. boot with zram loaded, and then make all physical ram into zram, and then copying the fd64.sfs into ram, and then running totally from ram ) . I have a feeling ( not knowing correct or not) that this seems increase computer’s performance on old hardware ? thanks

[rcrsn51]

@step: Nice work with your Samsung printer problem. But I am still wondering why the driver works in Fatdog 6xx and fails in Fatdog 7xx with your particular model. Here's what I think is happening.

1. Your PPD file has some wonky option.

2. The CUPS print chain in Fatdog 6xx uses Ghostscript. It copes with the option and keeps working.

3. Fatdog 7xx has replaced Ghostscript with some Poppler tools. They are choking on the option and causing the print process to fail.

[smokey01]

step, have you tried my package:
http://smokey01.com/fd700/packages/sams ... 6_64-1.txz

It works here on my Samsung C1860FW.

This was made from rcrsn51 original package. I remember having some problems getting it to work in the beginning.

Cheers

[Ted Dog]

@quirkian2new Quirky is named that way since its the title of anything new and untested, some ideas never make it to a next release. As far as I can tell BK made it after we complained... and complained ... and complained .... and (you got the point) that the unpack to flash drive was keeping most of us out of testing new quirky releases. Haven't seen it discussed since, good idea seemed solid but not the direction he was heading.
I also do not think its in the direction of FatDog based on past ideas floated. Closest in fatdog would be base2ram=expand on the kernel boot line.. like this example

Code: Select all

menuentry "Start Fatdog64-700dm EXPAND" {
linux (hd0,msdos3)/fd700dm/vmlinuz base2ram=expand basesfs=local:/fd700dm/fd64.sfs waitdev=1  radeon.dpm=1
initrd (hd0,msdos3)/fd700dm/initrd
}

in this example I am pulling the fd64.sfs from the folder fd700dm on the third dos part, on a MBR based system using grub2
the base2ram=expand will expand to take almost 800M so you will not have much of any RAM left, however if you are just surfing around its enough.

I also use a slow lowend laptop with a heat issue I got for cheap if its a AMD based unit check some hints a few pages back to help.

[Ted Dog]

Actually if Kirk ( where has he been, should we be concerned ?) and/or JB would add ZRAM that might help the other issue I am wanting to improve, shouldn't make it worse.
Which also your other post, gives me the idea to move ALL RAM based save methods before killing X and ending most other process in the savefile.sh area, Since you proved it can block any other shutdown attempt ( till hardware faults ) and would be able to save another way if the multisession saves fail...

[step]

@rcrsn51, I admire your explanation. I was clueless why my printer did work in FD631 but didn't in FD701... to the point that I started doubting my own recollection of being able to print in FD631. Anyway, I'm glad I got it sorted out. In the process I have learned quite a bit about CUPS printing. Now I'm tinkering with an n-up printer driver. I've tested boomaga and some others, none to my satisfaction. Poppler is another story. I compiled version 0.37.0 and installed in FD701 only to discover that evince breaks with an out of memory error when the new poppler version is loaded.

@smokey01, yes, I did try your samsung 0.21 driver package. It too includes a defective PPD file for my printer model, SCX-3400 Series. It can be fixed exactly as I fixed the upgraded driver package.

[rcrsn51]

@step: Here's an easy way to get n-up printing (although you may need something fancier). Use Print > Print to file to generate a PDF. Then use PeasyPDF's Print tools.

[Ted Dog]

OK I am almost sure this is a bug.. since I have not recalled a pop-screen for a long time.

/etc/rc.d/rc.shutdown

Code: Select all

### last minute work - while X is still running
get_DISPLAY
if [ "$DISPLAY" ]; then
	[ "$MULTI_MOUNT" ] && insert_disc_reminder
	if [ -z "$SAVEFILE_MOUNT" ]; then
		savefile.sh	
		. $BOOTSTATE_PATH	# reload bootstate - savefile.sh may have inserted temporary mounts
	fi
fi

line #106 missing the if
only [ "$MULTI_MOUNT" ] && insert_disc_reminder

[Ted Dog]

Ahhhhhhh! double calls, why JUST why...

rc.shutdown lines 123 - 125

Code: Select all

while pidof snapmergepuppy > /dev/null; do sleep 1; done
while pidof fatdog-merge-layers.sh > /dev/null; do sleep 1; done
while pidof savesession-dvd > /dev/null; do sleep 1; done

then inside /usr/sbin/savesession-dvd lines 37 & 38

Code: Select all

		while pidof snapmergepuppy > /dev/null; do sleep 1; done
		while pidof fatdog-merge-layers.sh > /dev/null; do sleep 1; done

then the scary.... line 42

Code: Select all

kill $XPID

which I hope it stops the xdialog [is it even seen?? X has been dead for a while] would'nt that cause an error that could bubble to skip actual save??

[jamesbond]

Actually if Kirk ( where has he been, should we be concerned ?)

Kirk's been busy. He's okay, though, so thanks for the concern.

and/or JB would add ZRAM that might help the other issue I am wanting to improve, shouldn't make it worse.

I don't know how ZRAM will help. ZRAM is RAM-based r/w compressed filesystem; functionality wise it is almost the same as SFS in a regular RAM-based filesystem (the only difference being SFS is read-only and as tighter compression than ZRAM).

To answer quirkian2new ; if you use small initrd, you can also use "baseram=yes" and that will load the SFS into RAM as well; and I'm quite positive that it will take less space than expanding the SFS into ZRAM.

Which also your other post, gives me the idea to move ALL RAM based save methods before killing X and ending most other process in the savefile.sh area, Since you proved it can block any other shutdown attempt ( till hardware faults ) and would be able to save another way if the multisession saves fail...

Save failing on the last minute when other parts of the system have been shutdown is an issue. I've updated it so that if mksquashfs fail, it doesn't get written to the disc. Not much improvement I know but at least you don't get false sense of security - if save fails you simply don't see the SFS on disc.

X is shutdown first because with X running there are other processes running too; this implies that there are open files. Save can't be fully complete when files are still open.

OK I am almost sure this is a bug.. since I have not recalled a pop-screen for a long time.

It could also be because get_DISPLAY failed. Launch terminal, unset DISPLAY, run the get_DISPLAY routine and see if you can get a DISPLAY.

Ahhhhhhh! double calls, why JUST why...

rc.shutdown lines 123 - 125

Code: Select all

while pidof snapmergepuppy > /dev/null; do sleep 1; done
while pidof fatdog-merge-layers.sh > /dev/null; do sleep 1; done
while pidof savesession-dvd > /dev/null; do sleep 1; done

Not sure why you keep saying that this is double calls? What that piece of code is supposed to do is to wait until all those programs (snapmergepuppy, fatdog-merge-layers.sh, and savesession-dvd) to finish before continuing. It does not launch those programs.

then inside /usr/sbin/savesession-dvd lines 37 & 38

Code: Select all

		while pidof snapmergepuppy > /dev/null; do sleep 1; done
		while pidof fatdog-merge-layers.sh > /dev/null; do sleep 1; done

Same as above

then the scary.... line 42

Code: Select all

kill $XPID

which I hope it stops the xdialog [is it even seen?? X has been dead for a while] would'nt that cause an error that could bubble to skip actual save??

Yes, it is supposed to stop Xdialog. savesssion-dvd is supposed to be called when the GUI is still running (the button on the desktop). rc.shutdown doesn't use it.

[step]

@step: Here's an easy way to get n-up printing (although you may need something fancier). Use Print > Print to file to generate a PDF. Then use PeasyPDF's Print tools.

Thanks. I played with PeasyPDF; I'm going to write my comments in the PeasyPDF thread.

[Ted Dog]

Please actually review those calls and lines listed above.. it is actually calls 4X back to back and the dvd call is AFTER shutdown killed X.

[jamesbond]

Please actually review those calls and lines listed above..

I explained to you, piece-by-piece what those lines are for, in my previous post.

it is actually calls 4X back to back

who calls what 4X?

and the dvd call is AFTER shutdown killed X.

Indeed, and I explained the reason why. Savesession is done on line 149 in rc.shutdown, using a shell function called "save_multisession shutdown", not by calling savesession-dvd; as I said in my previous post, savesession-dvd is supposed to be used when the GUI is running. If you read what's inside /usr/sbin/savessession-dvd you will see that eventually it makes the call to the same shell function but with a different parameter ("save_multisession no-shutdown").

If you think those lines are buggy, then what should be the correct one look like? I'm open to patches

[quirkian2new]

Maybe the wording of my previous post is not too clear. I boot Fatdog701 with my desktop (which has 8 gb ram) and found that the right bottom corner showed that about 4 gb ram is available.

I then boot Quirky7.0.4.1 with my desktop (use rufwoof’s method of first making changes to init so that all physical ram were made into ZRAM ) and found that it showed that over 7.5 gb ram is available. (i.e. 4gb vs 7.5gb available). (The original iso of Fatdog701 is below 300mb while original iso for Quirky 7.0.4.1 is below 200mb.)

I am an average user and new to linux. Does this means that I have more ram to use when ZRAM is introduced and all physical ram made into ZRAM ? If so, why not a good idea to add such a feature into Fatdog701 and later versions of Fatdog64 ?

[Ted Dog]

The ram use on Fatdog64 and most non quirky version divides the RAM in half making /tmp and the 'magic' layered filesystem in one half and user files in second half. You are only shown user amount left. The great majority of the /tmp half is unused with such a large total of RAM. And can still be used but not shown against user section.

[MU]

Thanks for this great puplet

I have some Intel baytrail devices. They require a 64bit system, but a 32bit Grub bootloader.
There is a patched Ubuntu with such a loader (google for Linuxium).
I took the grub files and packaged them.
You can simply copy them to the stick on which you installed the iso.

Files+readme+screenshots:

https://drive.google.com/folderview?id= ... sp=sharing

Sound works with a cheap 11 Euro USB soundcard. Inbuilt wifi does not work. Will try at home if a USB to Ethernet adaptor or an external wifi dongle work.

Cheers, Mark

[Sage]

Welcome back, Mark!

[MU]

thanks Sage

Glad to see many well known names around here like you, Ted Dog, Kirk, Rarsa...

I used Puppy (precise) until some months ago as my main system, just had no time to be active any longer. Work as application manager for united internet now, have a girlfriend who lives in germany but with kenyan roots... so busy with other stuff. However because of my girlfriend I abandoned my desktop computer, too loud for the sleeping room where I spend most time.
As a result I bought some fanless devices based on the intel baytrail processors. Unfortunately there was no Linux for them until Linuxium found out that you have to add a 32 bit bootloader. So I was a "windows user" for some months, until I took a closer look at his solution.
He added a 32 bit loader to an installation CD of Ubuntu.
As I was afraid to brick my devices, I did not want to install on the inbuilt emmc drives, and finally was able to create an SD card using the installer. So I can run Ubuntu from external SD card.

summary: https://www.facebook.com/photo.php?fbid ... cation=ufi

link: download SD card image + readme.txt: https://drive.google.com/folderview?id= ... sp=sharing

Now with the running system I saw that it is no magic, and tried it with Puppy. The 32 bit versions like precise do not work with it. Quirky April boots, but offers only to save on CD. Finally Fatdog was the solution: I can save to USB with it.

My girlfriend will visit Kenya in november for 3 months, so I might have some more time to customize the Fatdog to my needs

Will not spend as much time as some years ago again, but still love Puppy and will follow especially fatdog and quirky

[MU]

I installed it on another Tablet, a Medion P2214T, it uses the Baytrail Celeron N2940 processor that supports 4 GB of RAM.

Same method with the 32bit Efi as above.

It had problems first to authentificate on my wlan network, so I added an external Wifi stick. After two reboots now both wifi devices work.
On this tablet, the internal soundchip works out of the box (not the case with Ubuntu).

Installed Icewm, added some themes, used gslapt to install Mplayer and Chrome and for Window-shadows "compton", and added a script to set gamma and brightness of the screen.

My preferred filemanager, XFE, is available too in gslapt

So after some minutes of configuration the system works as I like it, so it will replace the Ubuntu I had set up before.

Thanks again for this sweet Distro


If you plan to buy a baytrail device, I recommend the versions that support 4 GB of Ram, like the N2xxx celerons, or the new "Cherrytrail X8500" processors that are rare to find yet, but soon will be very common.

With the "z37xxx" processors that support just 2 GB, you quickly run out of memory using modern webbrowsers with more than 10 tabs open at once.

Mark