BookwormPup64 10.0.10

[OscarTalks]

Yes Sir, from Debian repos, using the out-of-the-box Synaptic.

OK, well I have never tested the Debian build in BookwormPup64 because I compiled my own which is "Puppified" and meets my needs. You may be able to set the Debian version to run as root from within Puppy, but it may refuse to run like that. I suggest you uninstall the Debian one and try installing the .pet I linked to instead. If you do that, let me know if it works.

[wanderer]

hi radky and bookworm team

thank you for doing all this work
and creating this masterpiece

in 10 months we are going to have to decide
what our new distrowatch candidate will be

do you think we should upgrade bookworm ?
or do you think we should consider another candidate ?

i will report what you suggest
on the whats up in puppy world thread so everyone can see it

wanderer

[greengeek]

I have not read the whole thread yet (just started trialling BookwormPup now) but just want to highlight some things that have stood out so far:

1) Conky sometimes seems to have some corruption/overwriting.
Not sure what it is trying to display:

BookwormPup64_conky.jpg (33.21 KiB) Viewed 3890 times

2) In most pups - executable bash scripts are shown with their title in green font. But in BW64 they just have black font titles - same as non-executable scripts.

executable_non_green.jpg (63.7 KiB) Viewed 3890 times

3) I find the default tray icons more confusing than I have seen in other pups.
- The "terminal" icon resembles the colour of the desktop so I confuse it with the "reduce to desktop" function
- The actual "reduce to desktop" (minimise windows) icon looks more like a leafpad or notepad/editor icon
- The default browser icon just confuses me. (Clouds and blue sky remind me of desktop images rather than browser function)

Tray_icons_confusing.png (29.29 KiB) Viewed 3890 times

- I don't like a browser to suddenly pop up an "updating" splash screen and hold up my browsing. Reminds me of Windows behaviour. I feel that browser updates should be totally under user control not vice versa.

Just my 2c worth based on half an hours usage. Have not used long term yet.

[Fossil]

@greengeek
The problem with overflowing text is the width of Conky's display. There are two easy solutions:
1) Go into the Menu-Desktop-Conky-setup, resize the font. Perhaps not so good if eyesight is a problem.
2) Open up Edit conkyrc. Resize the display's width, such as: # Minimum size of text area
minimum_size 500
maximum_width 333

Which works for me on a 1680 x 1050 display.

[Fossil]

@greengeek
The problem with overflowing text is the width of Conky's display. There are two easy solutions:
1) Go into the Menu-Desktop-Conky-setup, resize the font. Perhaps not so good if eyesight is a problem. Or extend the Windows width to a larger number.
Mine is using NotoSans 13 # Minimum size of text area
minimum_size 500
maximum_width 333

Which works for me on a 1680 x 1050 display.

[Governor]

BW64 has XFE file manager v1.45 with Edit>Preferences>Dialogs>Confirm execute text files enabled. This results in frequent spinning cursors delays. Unchecking this setting stops the delays, but in this version it also disables XFE's ability to run script files. This was not the case in prior versions. I down revved XFE to v1.43.2 and it fixed the issue.

Thanks
wizard

BookwormPup64_10.0.6

How do I fix this on my computer?

[bigpup]

I don't like a browser to suddenly pop up an "updating" splash screen and hold up my browsing. Reminds me of Windows behaviour. I feel that browser updates should be totally under user control not vice versa.

This is because firefox, in it's settings, is set to auto update.

The Firefox browser needs to have this set to, search for updates, and ask what to do.

would be good to have this setting, already in it.

This auto update setting, has caused issues for me, on first boot-up, on a slow computer.
It makes it seem something is wrong with the Firefox browser, until it completes downloading, and doing the update.

On a low RAM computer. I would like to first make a save and be using it, before doing the update to Firefox.

[wizard]

@Governor

How do I fix this on my computer?

Install this:
https://sourceforge.net/projects/friend ... t/download

wizard

[fredx181]

I don't like a browser to suddenly pop up an "updating" splash screen and hold up my browsing. Reminds me of Windows behaviour. I feel that browser updates should be totally under user control not vice versa.

This is because firefox, in it's settings, is set to auto update.

The Firefox browser needs to have this set to, search for updates, and ask what to do.

would be good to have this setting, already in it.

This auto update setting, has caused issues for me, on first boot-up, on a slow computer.
It makes it seem something is wrong with the Firefox browser, until it completes downloading, and doing the update.

On a low RAM computer. I would like to first make a save and be using it, before doing the update to Firefox.

I think best is to have installed by default the firefox-esr package from the Debian repos, then it will upgrade only if you do apt update; apt install firefox-esr (works only if an upgrade is available of course).
edit: or install / upgrade from Synaptic.

[dancytron]

Yes Sir, from Debian repos, using the out-of-the-box Synaptic.

OK, well I have never tested the Debian build in BookwormPup64 because I compiled my own which is "Puppified" and meets my needs. You may be able to set the Debian version to run as root from within Puppy, but it may refuse to run like that. I suggest you uninstall the Debian one and try installing the .pet I linked to instead. If you do that, let me know if it works.

There is a simple check for root in vlc that you can turn off.

To make Debian vlc run as root, run this from a terminal:

Code: Select all

sed -i 's/geteuid/getppid/' /usr/bin/vlc

See:

https://unix.stackexchange.com/question ... er-in-root

You will probably have to redo it when it upgrades.

[radky]

BookwormPup64 10.0.7

• New release
• See first post to download and second post for changelog

[Marv]

Updated my 10.0.6 install to 10.0.7 on my 2nd gen I5 Fujitsu S761 laptop. punionfs=overlay passed as a kernel parameter as usual. This frugal install has no savefile and a small ydrv with my customizations so it is very easy to update. Smooth update, clean dmesg, good idle resource use, good glxgears FPS, good normal use with all portables working as expected.

Also not unexpected, my hardware is not happy with the stock 6.1.94 kernel, both wrt connman persistence and synaptics touchpad re-initialization on resume from suspend. In general 6.6.x kernels and up have worked better for me. Running @ozsouth 6.9.4 kernel at the moment and seems perfect wrt both so far. Posting from the updated install running the current Brave portable.

Thanks

[mikeslr]

@radky

I haven't downloaded 10,0.7 yet. [Have to do that from laptop which is near to router]. But I have an issue with Synapse and don't recall if I specifically mentioned it on a Bookworm thread. The Changelog doesn't reflect any relevant change.

Synapse has some nifty features. But I keep all datafiles in folders on /mnt/home. I could find no way for Synapse to search beyond /root: no adjustment to preferences; and it did not follow symlinks in /root to elsewhere. Adding Synapse to Right-Click menu achieved nothing. Unless I've missed something, it's basically useless.

Compare Right-Click (folder)>Search with pfind.

@radky @rockedge @dimkr This is not just a Bookworm issue. I know I've mentioned it several times. Maybe it has to be fixed at woof. On initial boot QuickSet up is automatically run. It offers to sync to the User's time server before the User has configured wifi. [Maybe when it was written the assumption was everyone had a lan connection to the router]. At any rate, AS-IS the setup appears amateurish: asking someone who has to use wifi to do the impossible so that an error message will be displayed. Either leave that module out, or change the order, testing for an internet connection before triggering the time-server configuration module.

[radky]

Synapse has some nifty features. But I keep all datafiles in folders on /mnt/home. I could find no way for Synapse to search beyond /root: no adjustment to preferences; and it did not follow symlinks in /root to elsewhere.

Synapse Usage Guidelines:

BW64 Menu -> Help -> HOWTO use the Synapse Application Launcher

Synapse parses .desktop files in the /usr/share/applications folder and quickly launches the same applications available in the main Puppy menu. In BW64, Synapse also integrates the plocate utility to find and launch executable files -- or to open/view filetypes supported by appropriate mimetypes.

The plocate utility finds files by parsing a database of current system files generated by the updatedb.plocate utility at each boot -- or you can update on demand (Menu -> Filesystem -> Synapse update). If updatedb.plocate is called without additional configuration parameters, it will defer to parameters defined in /etc/updatedb.conf.

The plocate utility does not search the /mnt/home symlink to/initrd/mnt/dev_save but we can configure /etc/updatedb.conf to support direct scans of /initrd/mnt/dev_save (and exclude searching of the pup_* folders in /initrd).

If interested, you could try the following to support plocate searching of /initrd/mnt/dev_save:

(first, make a backup of /etc/updatedb.conf)

METHOD 1:

Run the following in a terminal to enable searching of /initrd/mnt/dev_save (and disable searching of /initrd/mnt/tmpfs and pup_*):

Code: Select all

sed -i 's%/initrd %/initrd/mnt/tmpfs /initrd/pup_a /initrd/pup_b /initrd/pup_dev /initrd/pup_f /initrd/pup_k /initrd/pup_loop /initrd/pup_ro1 /initrd/pup_ro2 /initrd/pup_rw /initrd/pup_z %' /etc/updatedb.conf

Then update the plocate database of system files: Menu -> Filesystem -> Synapse update

METHOD 2:

Alternatively, you can manually edit the blacklist directives in /etc/updatedb.conf

Go to line #3 (PRUNEPATHS) and replace

Code: Select all

/initrd

with

Code: Select all

/initrd/mnt/tmpfs /initrd/pup_a /initrd/pup_b /initrd/pup_dev /initrd/pup_f /initrd/pup_k /initrd/pup_loop /initrd/pup_ro1 /initrd/pup_ro2 /initrd/pup_rw /initrd/pup_z

Caveat:

The plocate database (/var/lib/plocate/plocate.db) is only as current as the most recent call to /usr/sbin/updatedb.plocate (which updates the database). If you create new files or delete old files, plocate will not be aware of the changes until the database is updated again.

[ozsouth]

I made a more compressed version of BookwormPup64 10.0.7, to fit on a CD (is 692mb). Also made a delta (246mb).

New version 17/9/24.

If your cpu temp display doesn't work, I made a new version (which allows for changes in this pup's base).
See: https://forum.puppylinux.com/viewtopic. ... 82#p125682

[Clarity]

Hello @radky and @rockedge. Very nice PUP.

To me, this appears as a work of Art. Very rapid executions.

On bare-metal Boots as expected in SG2D while a Ventoy launch will trap with/without a :psave=..." parm entry. Operates as expected on desktop and session saves as expected with WoofCE distro where the SAVESPEC file directs the location for saving and accessing the PUP's session. Thus, when launched from SG2D no user efforts are needed as the distro will progress (as expected) to desktop and will operate including shutdown without any issues.

QEMU report will be forthcoming.

Found a nagging issue in FirstRUN as it relates to the taskbar. See #1 in pic.

FirstRUN.jpg (95 KiB) Viewed 3595 times

[Clarity]

The #2 in above picture is not something tied to BKWP64. That item is a concern I've had for years and not sure which of the authors/contributors for FIrstRUN my concern should be directed.

Reason:
Some systems have an nvidia adapter (and others) for connecting monitor(s) to the system. And more and more users are using monitors-TVs connected via HDMI which also carries audio to the monitors.

When I first noticed this, I "thought" that the author(s) may at some point planned for that checkbox to ALSO allow a driver switch to occur.

So for years I had been hopeful that that option would expand to accommodate switching drivers.

But, as Linux systems progress more and more over time with nvidia continuing to give an allusion to becoming more Linux compliant, this box will probably never have its feature-set expanded for driver changes.

[Clarity]

I completed 80% of my package needs for long term use. It has taken a bit of time, today, to get to where I am for my uses. CPU with 5 Chrome windows open sits at 1% with NO 8G Swap partition storage used, as it sits at 0. Current system view on bare-metal with packages I've added is in this picture.:

FirstRUN(3).jpg (16.81 KiB) Viewed 3584 times

[bigpup]

BookwormPup64 10.0.7
Frugal install with save folder.

Pmount program

Right click on a listed partition that is not mounted.
Select check file system
Nothing happens.

Yes the partition is formatted in a format that should be able to be checked. (ext 3 or 4)

[peebee]

The BookwormPup64 build recipe is now available on Github Woof-CE and BookwormPup64 can be built using Github Actions:
https://github.com/puppylinux-woof-CE/w ... n/dpupbw64

(Note: the .sfs are larger due to the Github run environment providing an older version of zstd)

[radky]

BookwormPup64 10.0.7
Frugal install with save folder.

Pmount program

Right click on a listed partition that is not mounted.
Select check file system
Nothing happens.

Yes the partition is formatted in a format that should be able to be checked. (ext 3 or 4)

@bigpup

Please insert the ampersand symbol (&) at the end of the action command on line 554 of /usr/sbin/pmount.

"$ONEDEV\" &</action>

Then launch pmount, choose the partition to evaluate -- and select the 'check filesystem' option.

Hopefully, the terminal will open and we can see if fsck runs with or without errors.

[bigpup]

Made edit.

No help.

Still nothing happens.

No terminal opens.

I copied the complete 554 line (what is between the <action> and </action>) and entered it in a terminal and got this:
.

Screenshot(2).jpg (49.71 KiB) Viewed 3343 times

[radky]

@bigpup,

The $ONEDEV variable (i.e. the selected dev partition) is defined in the early part of the pmount script.

If you simply copy the line to a terminal the $ONEDEV variable is undefined, and the command sequence fails.

Please edit line 554 of pmount and make sure a space precedes the & symbol):

Code: Select all

		 <action>rxvt --hold --title \"pMount - $(gettext 'Puppy Drive Mounter')\" -geometry 80x10 -bg black -fg green -e fsck -Cv \"$ONEDEV\" &</action>

Please test on different partitions and different drives.

[bigpup]

That is the way I have had the edit.

Code: Select all

<action>rxvt --hold --title \"pMount - $(gettext 'Puppy Drive Mounter')\" -geometry 80x10 -bg black -fg green -e fsck -Cv \"$ONEDEV\" &</action>

Are you getting it to do a file system check?

If yes.

I am using a save that has had software added to it by using synaptic package manager.

I will do a clean no save boot and see if it is still not working.

Someone posting in another topic, found the issue, so I tried and got it also, so reported it to you.

Update:
Now very, very, sorry I even said anything about this!

With a clean boot and using a new save with nothing new added.
It is working to check file system on partitions not mounted.

Now I am going to have to find what is braking it in the old save.

I hate it when a save is causing the issue!
This old one has some good stuff in it that I really want to keep.
I may just keep the issue.
Never really used Pmount for this check anyway.

Maybe I will figure out something and at least be able to help the other person, having the same issue with Pmount.

Again, very sorry making you have to deal with this!!!

I am loving the changes you did make with BookwormPup64 10.0.7!

UPDATE to the UPDATE:

It seems to be all about what mouse click you use to bring up the selection menu and also use to select.
Bring up the selection menu with a left click, then select file system check, by also using the left click.
It works.

If you use a right click to bring up the selection menu.
Left click on the file system check does nothing.
Right click on file system check and it works.

[MochiMoppel]

It seems to be all about what mouse click you use to bring up the selection menu and also use to select.
Bring up the selection menu with a left click, then select file system check, by also using the left click.
It works.

If you use a right click to bring up the selection menu.
Left click on the file system check does nothing.
Right click on file system check and it works.

It's a bug (feature? ) in GTK3.
E.g. in Geany open the "File" menu with rightclick, then leftclick "New". Nothing happens.
It works as expected when Pmount is run with gtk2dialog, but this would break other features.

@radky Regarding the command fsck -Cv . Does this really work? I can't check because all my file systems are NTFS. The usage info for fsck lists a -V option (capital V!) but no -v.

[radky]

Thanks for testing and feedback.

@bigpup

GTK devs are adamant about the usefulness of features added (or removed) as GTK2 transitioned to GTK3. However, I think we all agree that GTK3 lacks many of the convenient capabilities of GTK2, and additional restrictions are implemented in GTK4. It seems these functional quirks (bugs) are considered new features by the GTK devs.

@MochiMoppel

The fsck -v parameter was implemented in the Puppy pmount script many years ago -- as early as 2013. I'm not sure if the lowercase -v was ever a supported parameter of fsck, or maybe just a typographical error introduced in pmount. In a quick test, I see no difference in fsck output when using or not using a lowercase -v parameter, so maybe lowercase -v is a long-standing (harmless) typographical error. However, using the uppercase -V parameter does provide (minimal) additional information about the evaluated partition.

[MochiMoppel]

In a quick test, I see no difference in fsck output when using or not using a lowercase -v parameter

Confirmed. Very unusual that an option with wrong capitalization is simply ignored. Normally it triggers an error, occasionally it is accepted as valid, and there seems to be no consistency among the util-linux tools:

Code: Select all

# whereis -v
whereis: bad usage
Try 'whereis --help' for more information.

# whereis -V
whereis from util-linux 2.38.1

# blkid -v
blkid from util-linux 2.38.1  (libblkid 2.38.1, 04-Aug-2022)

# blkid -V
blkid from util-linux 2.38.1  (libblkid 2.38.1, 04-Aug-2022)

However, using the uppercase -V parameter does provide (minimal) additional information about the evaluated partition.

I find the "verbose" information not minimal but rather essential (tested with a mounted sfs file, type ext2). However in my case the fsck command apparently runs e2fsck, and in e2fsck the option for verbose is lowercase -v, not uppercase as in fsck. Could be the reason for lowercase in pmount.

Code: Select all

# fsck -nV /dev/loop4
fsck from util-linux 2.38.1
[/sbin/fsck.ext2 (1) -- /dev/loop4] fsck.ext2 -n /dev/loop4 
e2fsck 1.47.0 (5-Feb-2023)
Warning!  /dev/loop4 is mounted.
/dev/loop4 was not cleanly unmounted, check forced.
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
Pass 4: Checking reference counts
Pass 5: Checking group summary information
/dev/loop4: 2917/65536 files (8.5% non-contiguous), 109855/262144 blocks

# fsck -nv /dev/loop4
fsck from util-linux 2.38.1
e2fsck 1.47.0 (5-Feb-2023)
Warning!  /dev/loop4 is mounted.
/dev/loop4 was not cleanly unmounted, check forced.
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
Pass 4: Checking reference counts
Pass 5: Checking group summary information

        2917 inodes used (4.45%, out of 65536)
         249 non-contiguous files (8.5%)
           0 non-contiguous directories (0.0%)
             # of inodes with ind/dind/tind blocks: 246/18/0
      109855 blocks used (41.91%, out of 262144)
           0 bad blocks
           1 large file

        2035 regular files
         695 directories
           0 character device files
           0 block device files
           3 fifos
           0 links
         172 symbolic links (165 fast symbolic links)
           3 sockets
------------
        2908 files

[ozsouth]

Interesting cputemp observation. I made a 6.9.9 usrmerge kernel (will share soon) & only had to run: modprobe coretemp (on my Intel PC - modprobe k10temp would work with most AMD), then restart X to get the taskbar thermometer going. Doesn't work in 6.1.94 stock kernel. I note that gatotray takes the first temp1_input it finds on my PC, which is static 25C. If I could make it search in reverse, would get a more meaningful temp to display.

[radky]

@ozsouth

By default, Gatotray probes CPU sensor temperature in the following sequence, and displays the first value found.

1 - /sys/class/hwmon/hwmon0/device/temp1_input
2 - /sys/class/hwmon/hwmon1/device/temp1_input
3 - /sys/class/hwmon/hwmon0/temp1_input
4 - /sys/class/hwmon/hwmon1/temp1_input
5 - /sys/class/thermal/thermal_zone0/temp
6 - /proc/acpi/thermal_zone/THM/temperature
7 - /proc/acpi/thermal_zone/THM0/temperature
8 - /proc/acpi/thermal_zone/THRM/temperature

[Clarity]

.
The Table of BookWormPup64 boot results, by version, are posted here.
..