Fatdog64-700 beta2 [Closed]

[stemsee]

My touch screen laptop works out of the box, no calibration needed (which doesn't work anyway!)

How is that achieved?

[gcmartin]

@StemSee, the FATDOG team will have to address the technical aspects which achieve this.

In general, though, Touch achieved similar to Lighthouse? Mouse-button1 need be finger click anywhere on an open desktop to bring up menu, too, for program launches. This is really bringing new appearance useful to the kennel! Lightening fast and touchable.

2015 is going to be a great year, thanks in much part to the leadership that this has brought us. @Kirk and @JamesBond innovations!

[SFR]

This one puzzles me - SpaceFM (from repo, but also compiled by myself) is missing icons (except one, to be pedantic) on the left pane (devices).
They're GTK stock icons, so I have no clue why aren't they displayed; they're all ok in Slacko, for instance.
Moreover, all those icons are shown correctly in drop-down menu (see screenshot).

A little progress on the issue...
For some reason, GTK in FatDog is refusing to display stock icons (but only in some cases) if we use their "classic" names, in favour to their new equivalents, for instance:

Code: Select all

gtkdialog -s <<< '<window icon-name="gtk-find"><button ok></button></window>'

should display 'find' icon in title bar and it does so in Slacko, but not FatDog, whereas:

Code: Select all

gtkdialog -s <<< '<window icon-name="edit-find"><button ok></button></window>'

displays it properly in both Slacko and FatDog.

It's weird, because I red here that this naming convention became deprecated since GTK-3.10, so GTK-2.x should have no probs with it...

Greetings!

[technosaurus]

For some reason, GTK in FatDog is refusing to display stock icons (but only in some cases) if we use their "classic" names, in favour to their new equivalents,

...because the gtk devs are the parts of human anatomy with red hats. They make sure to backport all the broken mis-features to gtk+-2.24, so using the latest version of gtk2.24 ensures maximal breakage to make it seem like the transition to gtk3 isn't that bad (it is that bad).

To maintain compatibility DO NOT compile gtk+-2.24 with --disable-deprecated

[SFR]

--disable-deprecated

Hmm, in 2.24.24 I didn't find any option in ./configure that alludes to this.
The problem appears to be more complex and mysterious, because for most of widgets (tree, menuitem, button, etc.) the old naming still does the job...

Greetings!

[technosaurus]

that change was made in 2011 IIRC and possibly part of the reason debian stayed on 2.20 (I prefer a patched 2.16 version) I'd be interested to see someone fork the gtk-xcb version on sourceforge though.

The new naming scheme is in conformance with _actual_ specs at least.
http://standards.freedesktop.org/icon-t ... atest.html
and
http://standards.freedesktop.org/icon-n ... atest.html

This is a good thing, since it will make the transition to another toolkit that uses the specification easier. I've been looking at tekui and previously libagar, but a toolkit could just as easily be built on netsurf's framebuffer toolkit.

[laurentius77]

Could anyone make eztables working in FD64-700-b2?

There is a wrong point out of iptables binary which shoud be changend in IPT=/usr/bin/iptables

But after that everything mess up with an error

Code: Select all

Staring eztables...
iptables v1.4.21: invalid port/service 'INPUT' specified

I think a firewall is a crucial piece of software in an operating system and Eztables needs an urgent fixing.

Do you have any ideea what triggers this error?

Thank you.

[LazY Puppy]

Hi.

What kind of format is the file: initrd?

How do I open this to edit its content?

Thanks

[smokey01]

To edit initrd in Fatdog just click on it. Follow the prompts but you have to be in Fatdog to do it.

[neerajkolte]

Hi LazY Puppy,

If you would just click on the initrd file while running Fatdog, it extracts it in a directory in /tmp and also gives you instructions to repack it after your changes.

Edit: Opps too late.
Thanks

- Neeraj.

[LazY Puppy]

To edit initrd in Fatdog just click on it. Follow the prompts but you have to be in Fatdog to do it.

Hi LazY Puppy,

If you would just click on the initrd file while running Fatdog, it extracts it in a directory in /tmp and also gives you instructions to repack it after your changes.

Edit: Opps too late.
Thanks

- Neeraj.

Thanks.
Will try this next time booting FatDog.
Just thought it might be somehow possible to open this initrd from within another Puppy running.

[stemsee]

Lazy Puppy
unpack

Code: Select all

cpio -i < $path$boot/initrd

pack

Code: Select all

find . | cpio -o -H newc > $path$drive/$dir2/initrd

[jamesbond]

Page 16
-------

@peebee ("quote: With Ted Dog's help, I've made a hybrid iso for LxPup64 using your components in a similar way to LH64. Hope this is ok? ") - Sure, please feel free to use Fatdog64's UEFI boot model/hybrid iso stuff In fact I said somewhere that Fatdog64's UEFI stuff should be able to boot all other puppies too, including 32-bit puppies.

@Ted Dog ("quote: I have reborrowed Fatdog64s shim keys for all other linuxes I install, hope they do not mind.") - No, I don't mind at all

@pcpunk: officially full install is not supported for Fatdog 600/700. I have written a guide for full-install of Fatdog 600 which should work on 700 too - but beware it is not supported and not fully tested. Some other forum member (I forgot the name) took my guide and enhanced it a bit, he posted on his own blog - I think you should be able to google it.

@SFR: nbd - you need to specify the max_part parameter on your kernel command line, e.g. "nbd.max_part=16". The reason why nbd is built-in in Fatdog kernel is because Fatdog has the ability to run off NBD disks for distributed deployment.

@SFR: "any chance to get perl-regexp compiled to grep?" - I will have to check. If it needs perl for this to build, then the answer is no. If it can live with PCRE library only, then okay. Beware of perl regexp, though, it has a few very nasty corner cases. Better stick with regular/extended regexp.

@SFR: "upx not compiled with LZMA" - thanks for the notice, I will check.

@Volhout: it seems that the problem is missing firmware. Get the missing firmware (those are the file named in your dmesg output) from here: http://git.kernel.org/cgit/linux/kernel ... .git/tree/ and save them under /lib/firmware, and reboot. You should be able to get your wifi going.

Page 15
-------
@SFR: $HOME/my-applications is deprecated. It is in in devx because it contains L18L's i18n helper scripts. If L18L moves the location of these tools elsewhere, then "my-applications" will be totally gone.

@technosaurus: I don't like $HOME/.local/{bin,lib,etc} because the ".local" part makes it invisible. When a user installs an application inside his/her own home directory, he/she should be able to see where they are. "$HOME/.config" may be okay with questionable reasons that some config files are not user-friendly, but hiding binaries is a bad practice.

@SFR: Thanks for the report on suspend. I'll give it another review. I have personally never managed to get suspend to work correctly (it can suspend all right but when waking up, the screen is always blank for me) - perhaps due to my quirky graphics. Kirk managed to get it to work. But in all cases, it should not prevent all other power-down activities to fail, so I'll look into that.

@SFR: thanks for the updated pbackup/pmirror scripts.

@L18L: thanks for your ongoing i18n effort for Fatdog! I'll definitely visit the thread again, it is way overdue (and your PM too). Thanks for the updated Xlock script too.

@SFR: "missing zoom.sh" - yes that's the script, I had that included in 600. I was seriously considering whether to include that or not because it only works on very simplistic cases (e.g. only one screen). If you are working with multiple screens then that script will screw your display configuration I think I'll get it in anyway and see how it goes.

@SFR: "Mouse: Control+4 & Control+5 (vol up & down) don't work for me" - I tried it, works on mine. But sven does have a bug, the mouse/key assignment doesn't seem to work. How about this - how about assigning Control+4/5 to run a program instead, and see if the program is launched? If it does, then we know that the hotkey works, it's just the volume controlling that doesn't work.

----
I still have to read page 14,13,12,11,10,9
----

Meanwhile, moving forward now - Page 19
----
@Karl Godt: loop max_part - same as nbd, you need to specify that in kernel command line: loop.max_part=16 or something.

@JustGreg, thanks for the contribution.

@SFR: thanks catching up pnscan.1 manual

@SFR: "Have you ever considered displaying labels under drive icons (not only in tooltips) like on the screenshot?", not until you mention it. Sounds like a good idea for someone who has to manage a lot of partitions. Will get that in place, thanks.

@SFR: I will address the font problem a bit later. For now try installing font-misc-misc from the packages and see if it helps (this package contains the entire *.pcf fonts - some of which got trimmed to NLS and devx sfs during build-time).

@stemsee: "My touch screen laptop works out of the box, no calibration needed (which doesn't work anyway!)" - Calibration is manual. You need to run it from terminal, and once calibration is done you need to take the output and edit the appropriate files. I have been intending to write a script to automate this but my touchscreen laptop is locked down somewhere so I haven't been motivated enough. Btw - does the "long press as right-click" work for you?

@techosaurus: "To maintain compatibility DO NOT compile gtk+-2.24 with --disable-deprecated" - Thanks for the info. GTK 2.24 is the last version before GTK3, so it makes sense that they "prepare" us for the incoming c**p that is GTK3. I checked my build script, it doesn't use that --disable-deprecated though ...

As a side note, 700 beta2 uses GTK2 2.24.24. I just checked their source and there *is* GTK2 2.24.25. So in a way, it is still maintained (=alive, if barely). That says a lot about GTK3.

Page 20
-------
@laurentius - please revert your change to eztables binary, and instead symlink /usr/bin/iptables to /sbin/iptables as noted by SFR here: http://murga-linux.com/puppy/viewtopic. ... 947#813947

@SFR: initrd is plain cpio archive. It's uncompressed, so as stemsee said "cpio -i < initrd" will expand it.

cheers

[stemsee]

Long press does not activate right-click menu! In DebianDog the right-click menu opens as soon as I touch the desktop using touchscreen, it must depend on touchscreen recognition.

[LazY Puppy]

Lazy Puppy
unpack

Code: Select all

cpio -i < $path$boot/initrd

pack

Code: Select all

find . | cpio -o -H newc > $path$drive/$dir2/initrd

Thanks.

I have now updated my Edit-Initrdgz package from 1.4 to 1.5. It is now able to unpack and repack initrd.gz (like in default Puppies), initrd.xz (like in Lighthouse64) and initrd (like in FatDog64). Both, GUI and right-click Action, are now able to unpack and repack the files initrd, initrd.gz and initrd.xz.

[SFR]

the max_part parameter on your kernel command line, e.g. "nbd.max_part=16"

Thanks, didn't know that.

@SFR: "Mouse: Control+4 & Control+5 (vol up & down) don't work for me" - I tried it, works on mine. But sven does have a bug, the mouse/key assignment doesn't seem to work. How about this - how about assigning Control+4/5 to run a program instead, and see if the program is launched? If it does, then we know that the hotkey works, it's just the volume controlling that doesn't work.

You're right, the shortcuts work with regular apps.
But, I reassigned CTRL+mouse-4 & CTRL+mouse-5 reusing the very same options, I believe (Mixer -> Volume Up/Down) and it just started to work.
Here's what have changed in /root/.sven/mouse_config:

Code: Select all

-type=mixerv.so
+type=mixer.so

@SFR: "any chance to get perl-regexp compiled to grep?" - I will have to check. If it needs perl for this to build, then the answer is no. If it can live with PCRE library only, then okay. Beware of perl regexp, though, it has a few very nasty corner cases. Better stick with regular/extended regexp.

I recompiled it with perl regexp and seems to work fine without devx.
I think it's just a name, not dependency.

@SFR: $HOME/my-applications is deprecated. It is in in devx because it contains L18L's i18n helper scripts. If L18L moves the location of these tools elsewhere, then "my-applications" will be totally gone.

Speaking of i18n - I can't get momanager to work.
No matter if langpack is loaded or not, all I'm getting is:
"Sorry, this needed basic langpack from puppy is missing: langpack_en. The common repository normally has this PET. Aborting."
Turns out that there's lots of references in the code to /root/.packages/* and other stuff, which is Puppy/Woof specific.
Even after commenting out the code that aborts the execution:

Code: Select all

#131206 base langpack from puppy required ex: langpack_es
if [ "`ls /root/.packages/langpack_${transLANG}-* 2>/dev/null`" = "" ] ; then
 pupdialog --title "MoManager: abort" --background pink --colors --msgbox "Sorry, this needed basic langpack from puppy is missing: \Zblangpack_${transLANG}\ZB. The \Zbcommon repository\ZB normally has this PET. Aborting." 0 0
 exit 1
fi #131206

it just hungs forever in some inf. loop...
Or perhaps I'm missing something..? L18L, could you please step in?

@SFR: I will address the font problem a bit later. For now try installing font-misc-misc from the packages and see if it helps (this package contains the entire *.pcf fonts - some of which got trimmed to NLS and devx sfs during build-time).

Yeah, that works with xosview. Earlier, I got these fonts from some other distro, didn't know they're already in the repo here.

@SFR: "Have you ever considered displaying labels under drive icons (not only in tooltips) like on the screenshot?", not until you mention it. Sounds like a good idea for someone who has to manage a lot of partitions. Will get that in place, thanks.

Will be nice, thanks. Just a quick heads up: printf '\n' alone wasn't good enough for me to insert newline (e.g. "label-less" partitions weren't aligned vertically to fit the others) - turned out that it had to be printf '\r\n'.

I have personally never managed to get suspend to work correctly (it can suspend all right but when waking up, the screen is always blank for me) - perhaps due to my quirky graphics.

Tell me about it... Spent days (since Slacko-5.4, k3.2.x-3.10.x) trying various kernel parameters and other things, without any significant results...
But in my case it's worse, because usually it works flawlessly for 1, 2 or even 10 times in a row...and then suddenly - dead. Unpredictable.
Therefore I'm not using this "feature" too often and if - only after saving session.

it should not prevent all other power-down activities to fail, so I'll look into that.

I think this is the reason of the issue:

Code: Select all

#Debounce switch 
[ -e /tmp/acpi_poweroff ] && exit
touch /tmp/acpi_poweroff #acpi_poweroff-delay.sh called later will delete this flag after a delay.

Later in the code is:

Code: Select all

# don't suspend when lid is open
grep -q open /proc/acpi/button/lid/*/state && exit

so the acpi_poweroff flag remains in /tmp and blocks all further requests.

Greetings!

[neerajkolte]

Hi SFR,

I really liked the four software bundles you posted in Contributed packages thread.
Especially the games in them.
Thanks

- Neeraj.

[neerajkolte]

Could anyone tell me how to make all ports stealthy using Eztables.

I have posted a Question Here
Thanks.

- Neeraj.

[stemsee]

@LazyPuppY

initrd.img? DebianDog also initrd1.img .xz.gz

[L18L]

@SFR: $HOME/my-applications is deprecated. It is in in devx because it contains L18L's i18n helper scripts. If L18L moves the location of these tools elsewhere, then "my-applications" will be totally gone.

Speaking of i18n - I can't get momanager to work.
No matter if langpack is loaded or not, all I'm getting is:
"Sorry, this needed basic langpack from puppy is missing: langpack_en. The common repository normally has this PET. Aborting."
Turns out that there's lots of references in the code to /root/.packages/* and other stuff, which is Puppy/Woof specific.
Even after commenting out the code that aborts the execution:

Code: Select all

#131206 base langpack from puppy required ex: langpack_es
if [ "`ls /root/.packages/langpack_${transLANG}-* 2>/dev/null`" = "" ] ; then
 pupdialog --title "MoManager: abort" --background pink --colors --msgbox "Sorry, this needed basic langpack from puppy is missing: \Zblangpack_${transLANG}\ZB. The \Zbcommon repository\ZB normally has this PET. Aborting." 0 0
 exit 1
fi #131206

it just hungs forever in some inf. loop...
Or perhaps I'm missing something..? L18L, could you please step in?

Fatdog's i18n is growing very slowly (we are slow and lazy)
and changes occur every day.

Fatdog-languagepack in addition to puppy langpack is no more relevant because the most translations for all available languages are alredy in the NLS.sfs

Fatdog-languagepack is just an addition to this independant of Puppy's langpack. Just for testing and then should be integrated into that big fat NLS.sfs.

My working actual version attached.
But please go to http://www.murga-linux.com/puppy/viewto ... 1&start=65 and use before.

Hoping to get a /usr/share/locale/pl/LC_MESSAGES/fatdog.mo uploaded as: MoManager-pl-UNKNOWN

EDITed later:

Change

Code: Select all

        LANG=${ORIGLANG} xgettext -o ${ATEXTDOMAIN}.pot -j --from-code=UTF-8  --keyword=_ -L Shell --no-wrap  ./xEXTRASCRIPTS	

to

Code: Select all

        LANG=${ORIGLANG} xgettext -o ${ATEXTDOMAIN}.pot -j --from-code=UTF-8  --keyword=_ -L Shell --no-wrap -f ./xEXTRASCRIPTS

in /usr/sbin/momanager