Dpup Stretch 7.5 CE (RC-5)

[radky]

Hi,
Here need to transfer picture and text files from a smartphone to puppy.
(running puppy Stretch 7.5 K 4.1.149)

I have installed all the files according to the following post:
http://www.murga-linux.com/puppy/viewtopic.php?t=109828
Afterwards have restarted puppy.

Activated Bluetooth ON: Both devices are seen from each other. But a file transfer (picture or text file) fails with comment: "file not transferred".

Browsing the forum, I found that the needed kernel module l2cap is missing.

How could I get that l2cap module?

Thanks for any advice.
Charlie

Hi Charlie,

If I understand correctly, the bluetooth l2cap module was a legacy configuration option (CONFIG_BT_L2CAP) available for 2.5.X and 2.6.X kernels, and this config is now obsolete in modern kernels.

https://cateee.net/lkddb/web-lkddb/BT_L2CAP.html

It seems functioning Bluetooth for Puppy is hit or miss (mostly miss). I do not use Buletooth devices, so my advice is limited.

Maybe someone with Bluetooth experience in Puppy will offer their recommendations.

[radky]

Im on a 64bit hp pavillion with a usb wifi rtl8192cu is the driver running.

it works on reboot (maybe a little slow and unstable) but it looses connection after a while?

otherwise its fast the dpup 7.5 rc4!

could it be a 64bit issue? (bionicpup 8.0 works wery well but is around 30-40 mb bigger ind use of ram).

is there a fix for this?

kindly Gnimmelf

Hi Gnimmelf,

Many Puppy users report good WIFI connection with BK's Simple Network Setup (SNS).

Alternatively, you could try Frisbee which may provide improved stability for wireless roaming in some scenarios.

[tallboy]

There is a message from Debian regarding kernel upgrades, is that something that should/could be done to Dpup Stretch-7.5?

Debian urges users to update their systems immediately

[tallboy]

radky, a few posts back I diagnosed my troubles to expanding the /etc/hosts file. I have noticed that every time I boot a multisession CD, I get a new puppypc identifying number, which is also set in the /etc/hosts file.
Is there some kind of dynamic creation of that number which causes my troubles?
I have just made a new multisession CD-R, and saved my setup to the CD. It will not boot properly, the desktop has a cursor, and two overlapping, empty trays at the bottom, no background, icons or menu, no reaction to mouse input. I followed my NoteCase list of settings which I formerly tested to work, X and jwm was restarted before the save to cd. I think that JWMDesk may be the reason, but I don't know how to diagnose my trouble.
BTW, I use the k4.1.48 version. When I opened the .jwmrc from terminal, it was empty.

[radky]

tallboy wrote:

a few posts back I diagnosed my troubles to expanding the /etc/hosts file. I have noticed that every time I boot a multisession CD, I get a new puppypc identifying number, which is also set in the /etc/hosts file.
Is there some kind of dynamic creation of that number which causes my troubles?

The Puppy code which ensures a unique hostname is found in /etc/rc.d/rc.sysinit as follows:

Code: Select all

if [ "`cat /etc/hostname`" = "puppypc" ];then
 echo -n "Updating unique hostname..." >/dev/console #hostname
 echo "puppypc${RANDOM}" > /etc/hostname
 PUPHOSTNAME="`cat /etc/hostname`"
 HOSTSFILEFIRST="`grep -w 'puppypc' /etc/hosts|sed 's% puppypc%%'`"
 HOSTSFILEREST="`grep -v 'puppypc' /etc/hosts`"
 echo "$HOSTSFILEFIRST $PUPHOSTNAME" > /etc/hosts
 echo "$HOSTSFILEREST" >> /etc/hosts
 status_func 0
fi

At the following link you can read more about the genesis of this code as provided by 01micko in Aug 2011:

http://www.murga-linux.com/puppy/viewto ... 932#550932

Once the unique hostname is established on first run, it is recorded in /etc/hostname and /etc/hosts, then subsequently included in your savefile (assuming you choose to save the current desktop session before shutdown).

---

tallboy wrote:

I have just made a new multisession CD-R, and saved my setup to the CD. It will not boot properly, the desktop has a cursor, and two overlapping, empty trays at the bottom, no background, icons or menu, no reaction to mouse input. I followed my NoteCase list of settings which I formerly tested to work, X and jwm was restarted before the save to cd... When I opened the .jwmrc from terminal, it was empty.

Tallboy, many of JWMDesk's functions are accomplished by on-the-fly SED editing of /etc/xdg/templates/_root_.jwmrc, and these changes are subsequently reflected in the desktop via /root/.jwmrc.

Please keep in mind, if /etc/xdg/templates/_root_.jwmrc is manually edited by the user, any syntax errors introduced in the XML code could result in failed or faulty creation of /root/.jwmrc. In this scenario, the JWM desktop may fail.

[tallboy]

Once the unique hostname is established on first run, it is recorded in /etc/hostname and /etc/hosts, then subsequently included in your savefile (assuming you choose to save the current desktop session before shutdown).

OK, then that is the problem when it comes to multisession CD. Once I have saved my setup back to the CD, all workfiles are read from and saved to a HDD, and there is usually no need for saving to the CD at shutdown, I just poweroff. The problem is that the multisession CD will always create a new hostname on startup, and that will crash with any saved hostname. I have tested it, and I got a new hostname every time. In my old Lucid, the hostname was only puppypc, whitout any number to it.

[charlie6]

Hi radky,

uname -i returns «unknown» instead of system architecture (see shot attached with k4.9.149)

Have encountered this while compiling gcad3D, a CAD software (http://gcad3d.org).
Also the case on DpupStretch 7.5 k4.1.48.

It would return 32 or 64 according to system architecture currently running.

HTH
Charlie

[radky]

uname -i returns «unknown» instead of system architecture.

Have encountered this while compiling gcad3D, a CAD software (http://gcad3d.org).

It would return 32 or 64 according to system architecture currently running.

Hi Charlie,

In Linux it seems 'uname -i' is non-portable, and there are output discrepancies across GNU/Linux distros.

For me, 'uname -i' and 'uname -p' return 'unknown' in Dpup Stretch.

If I understand correctly, some distros have patched the 'uname' utility in the coretutils package.

For example, running the Slackware 'uname -i' on my system returns 'GenuineIntel' rather than a numerical platform designation.

If the gcad3d install script depends on a numerical output of 'uname -i' before proceeding with the installation, maybe you could remove that paticular test.

[tallboy]

My stress level is raised again , after another attempt to make a working multisession save to CD. 33Mb of time-consuming modifications down the drain. When I rebooted to RAM and opened the save folder, I noted that there are several modifications that are not saved at all, and several of the modified and overwritten original files, are saved in their stock form instead of the modified version. I have made exact notes of the modifications, so I am certain that I did everything correct. Again!
There is somthing wrong in the Dpup Stretch-7.5 RC4 save code, somewhere.
One indication may be two empty trays on top of each other (the border of rear one is just visible), a cursor in an empty screen, no mouse and keyboard interaction, and a strugggling player trying to get squeeze some more out of the CD, again and again...

[radky]

I noted that there are several modifications that are not saved at all, and several of the modified and overwritten original files, are saved in their stock form instead of the modified version. I have made exact notes of the modifications, so I am certain that I did everything correct.

tallboy,

By private message, if you could send the 'exact notes of your modifications' I could test your method of multisession save to CD.

Thanks

[tallboy]

radky, I'll translate some of my norwegian comments to english first, and send it to you. It is a Notecase file (.ncd).

BTW, I have compiled the latest ImageMagic 7.0.8-46 in Dpup Stretch, and made both an .sfs and a .pet, They are each ~24Mb in size. They can be found at https://w123.no/linux/, but it is maybe not the ideal download site...
ImageMagick can read, write and modify more than 200 image formats. If you are unfamiliar with the program, I suggest you read the intro documentation before attempting to modify an image. There is a README, with links, on site and in in both .pet and .sfs.

[charlie6]

Hi radky,

uname -i returns «unknown» instead of system architecture...
If the gcad3d install script depends on a numerical output of 'uname -i' before proceeding with the installation, maybe you could remove that paticular test.

I found a workaround:
in the relevant script among the source files, I edited and replaced ,«uname -i» by «getconf LONG_BIT»; this allowed the build to complete without error, and the resulting built binary file to find the folder whose name has the "getconf LONG_BIT" in it (actualy 32).

Maybe I should report your comment to the gcad3ds developper. In earlier versions he indeed had used «getconf LONG_BIT» instead of «uname -i» to get 32 or 64 according to the used system.

Thanks a lot for your reply!
Best regards
Charlie

[Lobster]

Is the RC-4 ready to be an official Puppy 8.1 Community Edition?

[darry19662018]

Yes it is very stable and with 4.1.48 kernel ideal for old machines.

[radky]

Is the RC-4 ready to be an official Puppy 8.1 Community Edition?

@Lobster,

Its good to see you on the forum again. I hope all is well in the Paradox Realm.

Since inception in late 2017, it seems Dpup Stretch 7.5 has become a 'slow' rolling release rather than an incremental point release. To that end, RC5 should be available soon including kernel and application updates, plus multiple fixes and enhancements introduced recently by our Woof-CE devs.

Although Dpup Stretch is a Woof-CE build, it is unconventional in many ways and deviates somewhat from standard Puppy. I'm not sure how well it would port to an 8.1 designation, but thanks for your consideration.

[Lobster]

@Lobster,

Its good to see you on the forum again. I hope all is well in the Paradox Realm.

It is is a quantum place. Both existent and non-existent. Schrödinger's cat says hi/meow ...

Since inception in late 2017, it seems Dpup Stretch 7.5 has become a 'slow' rolling release rather than an incremental point release. To that end, RC5 should be available soon including kernel and application updates, plus multiple fixes and enhancements introduced recently by our Woof-CE devs.

Although Dpup Stretch is a Woof-CE build, it is unconventional in many ways and deviates somewhat from standard Puppy. I'm not sure how well it would port to an 8.1 designation, but thanks for your consideration.

An unconventional Puppy? My favourite sort.
Because of the stability, Woof CE build, rolling release, hard work, you are welcome to be part of the movement towards Puppy Nine, any time ...

Live long
and Puppy

[augras]

Hi radky,
I just download your puppy : great job again. Thank you very much.
I put it in french with the french langpack by esmourguit : it works fine. And i put also FbBox in french : I just modify your english2french script for some translations and it is ok.
I didn't read all the 43 pages and don't know if you already explain how to put FbBox in french.
In case, to get FbBox in french : https://augras.eu/puppy_forum/index.php ... 409#msg409.
Only the Menu entries are in french and PupShutdown.
Thank again Radky,
Philippe

[augras]

Hi,
i'm with your puppy since yesterday, stretch-7.5-uefi-k4.9.149, on a old laptop lenovo x201i from 2009, and it seems everything works fine. I'm still with my xenial by phil666 in which my ram seems always full but it's not the same here : i don't understand why but it's what i can see.
I just not Get java runtime doesn't work (it's the same in the other puppies), and transparency doesn't work with urxvt-control.
I'm with 2 screens, the one of the laptop and another 22" one and zarfy is not easy to use like lxrandr : each reboot my config isn't keep.
For a RC it's a very good one.
Thank you for this great work.
Philippe

EDIT : i found 2 other small things
- Get flash doen't work (like in the other puppies) : there is a Getflash-1.6d.pet that works http://murga-linux.com/puppy/viewtopic. ... 3&start=22
- the sylpheed entry in the menus is empty : in the sylpheed.desktop file there is Name[fr]= but it must be Name[fr]=Sylpheed

[radky]

Hi Philippe,

Thanks for testing Stretch RC4 (released 10 Jan 2019). The next release (RC5) is now in testing and should be available soon.

Thanks also for the link which provides instructions for French localization of FbBox.

https://augras.eu/puppy_forum/index.php ... 409#msg409

---

augras wrote:

Get java runtime doesn't work

I downloaded and tested rerwin's get_java-1.1.pet, but it seems the Oracle site parameters have changed and the download fails. Currently, I don't have time to troubleshoot those changes, but maybe one of our forum members could evaluate further.

http://murga-linux.com/puppy/viewtopic.php?t=101592

---
augras wrote:

(urxvt) transparency doesn't work with urxvt-control

Please try the following:
1 - Go to Menu -> Utility -> Urxvt Terminal Control
2 - Set Transparency to On
3 - Set Shading to 100
4 - Click Apply
5 - Restart/Relaunch Urxvt

---

augras wrote:

zarfy... each reboot my config isn't keep

You could create a small script which (when booting) automatically loads zarfy with your last screen configuration settings.

For the auto-start script, run the following code in a terminal to create /root/Startup/zarfy:

Code: Select all

echo '#!/bin/sh
zarfy -l' > $HOME/Startup/zarfy
chmod 755 $HOME/Startup/zarfy

---

augras wrote:

Get flash doesn't work (like in the other puppies) : there is a Getflash-1.6d.pet that works.

The sylpheed entry in the menus is empty : in the sylpheed.desktop file there is Name[fr]= but it must be Name[fr]=Sylpheed

Thanks, I'll make the correction for RC5.

Edit:

The default GetFlashPlayer v1.7-1 works fine for me.

[augras]

Your script for zarfy works fine : thank you.
I made a new test with getflash in fresh frugal install and it works ! I don't understand what was wrong before !
A RC5 : are you sure ? You are not joking ?
The RC4 is already very good.
Thank you for your answer and your hard work,
Philippe