Puppy Linux Discussion Forum

dancytron wrote: ↑Wed Mar 06, 2024 9:43 pm

Installed on brand new 10.0.6 frugal install and it worked paired with my phone with Android 10.
Would you share what you ran to get it into a .pet file? I'd like to know how to do that.

Thanks for the report, glad it works on later androids.
If you are compiling it yourself it would be "new" by default. I changed it by editing the meson_options text file in the source, so if that is left unaltered that setting would be theoretically what is optimum for later phones, but maybe not a critical thing.

Anyway the answer to making a .pet once everything is compiling and installing is to run new2dir ninja -C builddir install (equivalent of using "new2dir make install" if using autotools and gnu make) and then of course dir2pet on the package directory tree that is created.

I did have to tinker around with things a bit. Not all those dependencies in that list need to be installed as some are in the devx. When I tried to apt install libsdl2-dev it complained about some dependency version mismatch so I compiled latest libsdl2 from source. I did have to apt install libavdevice-dev with dependencies. The libusb-1.0-0-dev package goes in via apt but the shared library is already present.

In the install script I had to edit the download path to the prebuilt server from version 2.3.1 to version 2.4 as it was grabbing the wrong version which refused to run. I also had to edit the sha sum because the script does a check of that. I think the git master checkout has this corrected now, but in the release tarball it has not been updated.

I did initially use apt to install the Debian adb which did work and helped me to see where it went in the file system. It has a symlink into PATH. Sticking with this would have been an option but I decided to try substitution of the latest official version and removal of the then unneeded shared libs.

hello all,
i was playing around with some scripts and i realize a small bug on GTK some attributes looks like do take any affect at all ex: width-request height-request

here is an example how is final layout test it in lunapup, fossapup and CE, Bionic the only puppy that display extra height is bookworm and if you replace default-width to width-request gets more wierd it looks likes <text warp="true" width-char="35"> dont do nuthig as well

#!/bin/sh

GTKDIALOG="`which gtkdialog4 gtkdialog3 gtkdialog | head -1`"
export GTKDIALOG ; [ -z "$GTKDIALOG" ] && echo "$(gettext "GTKDIALOG is missing, please install it first.")" && exit

# grab Puppy name and version
[ -f "/etc/DISTRO_SPECS" ] && . /etc/DISTRO_SPECS ; P="$DISTRO_NAME - $DISTRO_VERSION " || P=
export JWM_VER=$(jwm -v | head -n1 | cut -d " " -f2- | cut -d " " -f1 | sed -e 's/[snvgit-]//g')

export JWM_LOGO=$(ls /usr/share/pixmaps/puppy/desktop.svg 2>/dev/null | head -n 1)
export COLOR="#FF0000"
export P_BTN_PATH="/usr/share/jwm/themes_window_buttons"

pixmap () { echo '<pixmap><variable>IMAGE</variable><height>'$2'</height><input file>'$1'</input></pixmap>' ; }
export -f pixmap

# width-request="450"  or default-width="450"
# height-request="300" or default-height="300"
export MAIN_GUI='
<window title="Window buttons theme" icon-name="gtk-convert" default-width="450">
	<vbox border-width="10">

	<hbox>
		<vbox homogeneous="true" space-expand="true" space-fill="true">'`pixmap "$JWM_LOGO" 64`'</vbox>
		<vbox>
			<text use-markup="true"><label>"<b><span size='"'large'"'>JWM THEME </span><span size='"'large'"' color='"'$COLOR'"'>SWITCHER</span></b>"</label></text>
			<text><label>"'$(gettext "For :")' '$P'"</label></text>
			<text><label>"'$(gettext 'Desktop Environment')': JWM v'$JWM_VER'"</label></text>
			<text warp="true" width-chars="45"><label>"'$(gettext 'The window buttons can be themed to your liking. If you want to build your own buttons, see')' '${P_BTN_PATH}'"</label></text>
		</vbox>
	</hbox>


	<hbox homogeneous="false" space-expand="false" space-fill="false">
		<button space-expand="false" space-fill="false" tooltip-text="Reinstates the stock standard icons">
			<label>"'$(gettext "Clear")'"</label>
			<input file icon="clear"></input>
			<action>clear_func</action>
		</button>
		<text space-expand="true" space-fill="true"><label>""</label></text>
		<button space-expand="false" space-fill="false">
			<label>"'$(gettext "Quit")'"</label>
			<input file icon="quit"></input>
			<action>exit:quit</action>
		</button>
	</hbox>

</vbox>
</window>'
$GTKDIALOG -pc MAIN_GUI

Best regards

width-request layout

@AntonioPt

Recent Woof-CE Pups (such as BookwormPup) are GTK3 builds rather than the familiar GTK2 builds of legacy Puppy Linux. Unfortunately, when comparing the GTK2 and GTK3 iterations of gtkdialog, the definition and implementation of widget parameters can vary significantly and the appearance of the rendered GUI can be quite different.

For example, the GTK3 iteration of gtkdialog applies line wrapping differently than the GTK2 version, and it also adds extra padding to GUI widgets such that excess spacing is often visualized at the bottom of the GUI -- as demonstrated by your example code in BW64. In some cases the gtkdialog code can be adjusted so GUIs rendered by GTK2 and GTK3 versions of gtkdialog will appear similar, but in other cases you may be disappointed.

As a workaround for GUIs that do not render properly with the GTK3 version of gtkdialog, modern Pups also include the companion GTK2 version (named gtk2dialog). If you prefer GTK2 rendering of your application in modern Pups, simply adjust the code to test first for the presence of gtk2dialog. If gtk2dialog is not found, then your application is running in legacy Puppy which defaults to gtkdialog4/gtkdialog3/gtkdialog.

Adding gtk2dialog to line 3 of your example application:

GTKDIALOG="`which gtk2dialog gtkdialog4 gtkdialog3 gtkdialog | head -1`"

Hi @radky, thxx for all the info gonna try it as soon as possible

Best regards António

hi @radky any suguestion about the text ? thxx in advange

@AntonioPt

Text wrap in gtkdialog can vary depending on code parameters of the current GTK theme and/or variations in the selected font family/style/size.

For example, in the images below the GTK theme is Flat-grey-rounded with font size 10 and regular style. The only GTK font difference is the choice of font family: The first image is DejaVu San while the second image is Noto Sans.

In this case, the text wrap variation is determined by the choice of font family. However, variations in the selected font style and size can produce similar differences in text wrap.

radky wrote: ↑Sat Mar 09, 2024 9:17 pm

In this case, the text wrap variation is determined by the choice of font family. However, variations in the selected font style and size can produce similar differences in text wrap.

For short text like in this case it might be a better idea not to rely on automatic word wrapping and instead do the wrapping manually at the desired wrap points:
<text width-chars="45"><label>"'$(gettext $'The window buttons can be themed to your\nliking. If you want to build your own buttons,\nsee')' '${P_BTN_PATH}'"</label></text>

To a certain degree text wrapping in gtkdialog can be configured with the "wrap-mode" attribute:

wrap-mode="0" # wrap lines at word boundaries. same a default wrap=true
wrap-mode="1" # wrap lines at character boundaries
wrap-mode="2" # wrap lines at word boundaries, but fall back to character boundaries if there is not enough space for a full word

Thxx again for all the info @MochiMoppel and @radky get it now gonna try and thankyou again

grub2config v2.01 fails on GPT drive

drive = 250gb hdd, GPT, UEFI only
partition 1 = 300mb fat32 ESP, flags = boot, esp
partition 2 = 40gb ntfs win, flags = none
partition 3 = 20gb ext3 linux, flags = none

Grub2config fails, gives error

Here's the log file:

/sbin/grub2config
version 2.0.1
Mon Mar 11 06:43:14 PM GMT 2024
splash start Searching.
sda|_ATA_SAMSUNG_HM250JI
1
splash end Searching.
Choosed:sda
Checking sda...
Boot Partition:None: none
Boot flag not found on sda
Missings:Boot flag on sda
Set the boot flag using GParted or something.
splash end

Have unsuccessfully tried:
-reset flags
-delete all partitions, start over
-change HDD

Has anyone else seen this error?

Thanks
wizard

wizard, "grub2config v2.01 fails...Has anyone else seen this error?"

Yes, the first time on an HPLaptop which came with Windows 10. I assumed that the failure was peculiar to it. Then on a Dell after I had installed (first) Manjaro, (later) Linux Mint. I assumed the Manjaro install had somehow altered the initial Fat32 partition preventing grub2 from writing to it. grub2config had previously worked.

But you know what they say about assumptions.

@mikeslr

I've tested on both a Dell and Lenovo laptop and confirmed it only happens if the drive is GPT.

Thanks
wizard

@wizard - I'm guessing you already tried resetting the partition table via gparted to gpt (which resets & deletes everything). Reason I ask is when I set up a new hdd, I do that first & then if I create an ntfs partition, it automatically is assigned the msftdata flag, which I see your ntfs partition doesn't have. I did once set msdos partition table, then set gpt, when having issues.

@ozsouth

I'm guessing you already tried resetting the partition table via gparted to gpt

,

Yes, changed back and forth between MSDOS & GPT during the testing. The msftdata flag doesn't seem to have any effect.

Thanks
wizard

I use Ventoy (as most everyone knows) and only have used Grub2Config twice in my past. I cannot remember what/which storage device it was used on, so I am not offering any Grub2Config solution that maybe @shinobar knows of, or any that other forum members know. BUT ...

I successful have used Ventoy to create all of my current boot disk (mostly USB, couple HDD, 1 SSD). In doing so, when creating a bootable, I choose GPT and Secure-Boot (for UEFI) and have no boot issues. And from this create, I often add/change partitions to meet my needs. But they are always bootable.

This comment is not suggesting to change from use of Grub2Config, rather it is offered for device comparison as there may be something minor that can be gleen from it.

@Clarity
I don't find any post by shinobar since Aug. 2022, so I think someone else with the skills, maybe @gyrog, will need to take a look at this issue. Seems grub2config does not detect the boot, esp flags on GPT drives and will not install.

Thanks
wizard

I agree!

@jamesbond and @wiak are 2 other forum members who I am aware that have the skills to review.

Its important as Grub2Config is being incorporated into distros, OOTB.

wizard wrote: ↑Mon Mar 11, 2024 6:48 pm

Checking sda...
Boot Partition:None: none
Boot flag not found on sda
Missings:Boot flag on sda
Set the boot flag using GParted or something.

@wizard
I seem to recall awhile back that setting the "boot" flag on the esp partition has become optional.

Nothing to do with whether drive is GPT or MBR.

Did you do what the window requested, run Gparted and set the boot flag?

Yes, gparted the entire drive creating a first fat32 partition and setting its boot flag. grub2config still reports the same error.

Suggest further discussion of this grub2 problem --which isn't specific to Bookworm-- be continued here, https://www.forum.puppylinux.com/viewto ... 83#p113883 unless the moderators think some other Forum locations (Users?) would be better. In which case please move that thread.

If I try & use grub2config I get the same error message as Wizard, even though gparted shows boot & esp flags,
and gdisk has the first partition, correctly, as EF00.

Perhaps a clue is that the grub2config.log file has:

/mnt/nvme1n1p4/grub2config-2.0.1/usr/sbin/grub2config
version 2.0.1
Wed Mar 13 03:01:53 AEDT 2024
fdisk: cannot open /dev/nvme0n: No such file or directory
fdisk: cannot open /dev/nvme0n: No such file or directory
Boot Partition:None: none
Boot flag not found on nvme0n1
Missings:Boot flag on nvme0n1
Set the boot flag using GParted or something.

When I select /dev/nvme0n1, why is fdisk being passed /dev/nvme0n,
which doesn't exist...is nvme0n1 being truncated somewhere? I tried
to look at the code...but gave up as its beyond me.

Also, while gparted shows the boot & esp flags, fdisk doesn't. See this thread:

https://askubuntu.com/questions/983154/ ... -boot-flag

Moved to https://www.forum.puppylinux.com/viewtopic.php?t=10951

I, again, suggest that further discussion of the grub2 boot-loader problem be continued here, https://www.forum.puppylinux.com/viewtopic.php?t=10951.

This is NOT a BookwormPup64 problem. It will occur without regard from which Puppy grub2config is run.

Running firefox as spot is not convenient IMHO on BookwormPup :
viewtopic.php?p=115085#p115085

Hi,

I'm facing an issue where my internal HDD are not begin detected. My device are working probably I'm certain the issue are from the new kernel builds.
Same issue exists with F96-CE_4 release.

What type drive?
Hard drive?
SSD?
NVME?

What setup does it have?
msdos or GPT partition table?
What partitions on it and what format is used on them?

Hello @mypuppy2
Please show the results of this terminal command when booted on either desktop.

lsblk -a

Copy the result to a text file and post back here.

@bigpup

My drive type is HDD, ATA/SATA connection, with MBR layout.

@Clarity

```shell
# lsblk -a
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
loop0 7:0 0 441.1M 1 loop /initrd/pup_ro2
loop1 7:1 0 15M 1 loop /initrd/pup_k
loop2 7:2 0 109.3M 1 loop /initrd/pup_f
loop3 7:3 0 71.6M 1 loop /initrd/pup_z
loop4 7:4 0 8.2M 1 loop /initrd/pup_b
loop5 7:5 0 74.9M 1 loop /initrd/pup_a
loop6 7:6 0 0B 0 loop
loop7 7:7 0 0B 0 loop
sdb 8:16 1 14.6G 0 disk
├─sdb1 8:17 1 14.6G 0 part
└─sdb4 8:20 1 760M 0 part
zram0 254:0 0 2.8G 0 disk [SWAP]
```

@mypuppy2
Restart your PC, enter the UEFI and look for the system drive. Is it there?