Bionic32-light - 27 April, 2019

[sheldonisaac]

I need help with getting PUPMODE=13 which works in other Puppys, including Precise Light.

Problem located and, kludged. There is something amiss in initrd.gz. I substituted in Peebee's initrd.gz from bionicpup32-8.0+3-uefi.iso and your kernel line (with my folder name and uuid) booted to "PUPMODE=13".

You can go ahead and do that substitution if you wish but I will post an ISO update once @davids45 gets back to me about his printer fix.

Cheers, J

Edit: Problem found on line 1256 of init in init.gz, will definitely be updated.

Many thanks, jrb, for tracking it down! I'll just wait, no hurry at all.

Sheldon

[davids45]

G'day jrb,

I'm glad you located the pet of rcrsn51's gstoraster, despite my typo (too many decimal points in 905 ).

Also, I'd like a bit more detail on the lib-ld-2.27 link. How do you decide what is needed.

Sorry for my second typo in that post - it's just "ld-" not "lib-ld-" .

Again, years ago, I think rcrsn51's pets for installing Epson printers often included creating a link to add ld-lsb.so.3 (or earlier, ld-lsb.so.2) into Pups where this was missed by the Pup developer. Seems now to be missed in every new Pup .

This /lib/ld-lsb.so.3 link omission sees the Epson printer showing "Filter fail" when CUPS is opened after a failure to print.
Over the years, I've added this link manually for a series of my Epson printers across a wide range of Pups (I'm a bit of an Epson-fanboy/OAP). So far, it's always worked after a "Filter fail".

What does change often are the numbers at the end of the ld-file in /lib. It's usually the first file in /lib as it doesn't have the prefix 'lib-' so easy to find.

In this morning's Pup (Slacko-6.9.6.7 with a 3.16.43 kernel), it's ld-2.23.so.
Your Light Bionic has increased the ld- to ld-2.27.so
But so far, whatever the numbering, making the link "ld-lsb.so.3" to ld-x.xx.so has been OK - I'm waiting til ld-lsb.so.4 is needed .

I'd welcome you adding the ghostscript+gstoraster files to your next Bionic32-Light, plus the /lib link - these are not going to "heavy-up" Light by much. And the libtiff links in /usr/lib/ too, of course.

Perhaps other owners/users of Epson printers and scanners could comment and test?

Thanks,
David S.

[jrb]

G'day jrb,

I'm glad you located the pet of rcrsn51's gstoraster, despite my typo (too many decimal points in 905 ).

Also, I'd like a bit more detail on the lib-ld-2.27 link. How do you decide what is needed.

Sorry for my second typo in that post - it's just "ld-" not "lib-ld-" .

Again, years ago, I think rcrsn51's pets for installing Epson printers often included creating a link to add ld-lsb.so.3 (or earlier, ld-lsb.so.2) into Pups where this was missed by the Pup developer. Seems now to be missed in every new Pup .

This /lib/ld-lsb.so.3 link omission sees the Epson printer showing "Filter fail" when CUPS is opened after a failure to print.
Over the years, I've added this link manually for a series of my Epson printers across a wide range of Pups (I'm a bit of an Epson-fanboy/OAP). So far, it's always worked after a "Filter fail".

What does change often are the numbers at the end of the ld-file in /lib. It's usually the first file in /lib as it doesn't have the prefix 'lib-' so easy to find.

In this morning's Pup (Slacko-6.9.6.7 with a 3.16.43 kernel), it's ld-2.23.so.
Your Light Bionic has increased the ld- to ld-2.27.so
But so far, whatever the numbering, making the link "ld-lsb.so.3" to ld-x.xx.so has been OK - I'm waiting til ld-lsb.so.4 is needed .

I'd welcome you adding the ghostscript+gstoraster files to your next Bionic32-Light, plus the /lib link - these are not going to "heavy-up" Light by much. And the libtiff links in /usr/lib/ too, of course.

Perhaps other owners/users of Epson printers and scanners could comment and test?

Thanks,
David S.

Thanks David, I'm sure that there are other printers out there that will need this ghostscript, perhaps with tweaking, and gstoraster is tiny so no great burden. The fact you got it to work is an excellent indicator that it will work for others.

[Sage]

This light offering is a stroke of genius, in keeping with the original aims of Puppy. Staying with 32bit, sine qua non - it's LIGHT, and still works with 64bit, of course, worth mentioning only because that fact seems to evade many folks! Perfect for diagnosis, repair and demonstration. Did I say "It works", unlike some other bodged attempts at 'slim' versions. Certainly worth bug-squashing but avoid the siren voices that plague distrospace with requests for more, bigger, & co. - send them to major releases; wood-for-trees syndrome abounds.

[musher0]

Hi, jrb.

You may wish to replace mplayer with madplay to play the chime at boot.

mplayer sends 9.7 Kb of useless warnings to /tmp/xerrs.log, while madplay
sends none. Evidence attached.

The sound quality is equivalent.

IHTH

[musher0]

Hi, jrb.

Looking at those red entries in /usr/lib...

... libtar is missing! No problem! We'll just fall back on good old zip!

So is libtidy. This one is for proper html / xml formatting.

And libical. Including this library may be important for lots of people
who use the iCal type of calendars online and offline.

Just a few thoughts.

[musher0]

Hello all.

This real lzop will replace the feeble busybox clone at /bin.

Compiled in and for upupbb-light. (It will probably work in
same level Pups, but no guarantees.)

Enjoy!

[jrb]

This light offering is a stroke of genius, in keeping with the original aims of Puppy. Staying with 32bit, sine qua non - it's LIGHT, and still works with 64bit, of course, worth mentioning only because that fact seems to evade many folks! Perfect for diagnosis, repair and demonstration. Did I say "It works", unlike some other bodged attempts at 'slim' versions. Certainly worth bug-squashing but avoid the siren voices that plague distrospace with requests for more, bigger, & co. - send them to major releases; wood-for-trees syndrome abounds.

Thanks Sage, I have to confess I'm a retro guy. I liked the old Puppy ethos of small, simple and efficient. I sometimes wonder how software file sizes can continue to increase with, in many cases, no apparent improvement in performance? Oh well, "The good old days".

[jrb]

Hi, jrb.

You may wish to replace mplayer with madplay to play the chime at boot.

mplayer sends 9.7 Kb of useless warnings to /tmp/xerrs.log, while madplay
sends none. Evidence attached.

The sound quality is equivalent.

IHTH

Good call musher! I have never examined /root/Startup/startup-sound. What if I hadn't included mplayer? Change made.

[jrb]

Hi, jrb.

Looking at those red entries in /usr/lib...

... libtar is missing! No problem! We'll just fall back on good old zip!

So is libtidy. This one is for proper html / xml formatting.

And libical. Including this library may be important for lots of people
who use the iCal type of calendars online and offline.

Just a few thoughts.

I looked but did not find any apps in Bionic-light that depended on these, including /usr/bin/tar. I'm sure that if a program is installed, in PPM or otherwise, the need will become apparent.

[jrb]

Hello all.

This real lzop will replace the feeble busybox clone at /bin.

Compiled in and for upupbb-light. (It will probably work in
same level Pups, but no guarantees.)

Enjoy!

I will encourage testers to try installing your collection of .pets and seeing how they perform. I'll mention that on Page 1 when I post the update. I have included replaceit in that update, small and looks useful. I'm just about to run it through Woofce now, but I won't be able to post it until tomorrow morning. My bandwidth shrinks in the afternoon and evening, luckily I'm an early bird.

[dancytron]

I did a manual frugal install with this. It booted fine, but my mouse didn't work. I tabbed down in first run dialog and tried to get it to work but it didn't.

It is nothing special, just a logitech B100 standard usb wired mouse.

Anything you suggest to try?

edit:

This is from Debian Dog:

root@ChromeDog64:~# lsusb
Bus 001 Device 010: ID 046d:c315 Logitech, Inc. Classic Keyboard 200
Bus 001 Device 005: ID 1a40:0101 Terminus Technology Inc. Hub
Bus 001 Device 004: ID 1a40:0101 Terminus Technology Inc. Hub
Bus 001 Device 003: ID 07ca:0826 AVerMedia Technologies, Inc.
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 005 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 004 Device 002: ID 046d:c077 Logitech, Inc. M105 Optical Mouse
Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 002 Device 002: ID 0bc7:0008 X10 Wireless Technology, Inc. Wireless Transceiver (ACPI-compliant)
Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub

I'm going to suggest starting with a "brute force" attack. Here is xorg_base_new-upup19.03.pet containing all the files I left out, plus some that I left in. Most of them I'm sure have nothing to do with this problem, but if it does fix it then we can start figuring out exactly what is needed. If it doesn't then we can look elsewhere.

Install this .pet, reboot and see if that fixes your mouse.

Good Luck, J

Okay, I didn't have much time to mess with it, but it is intermittent. I only got it to do it one more time out of about 6 or 7 boots. I think the init might just need another wait in it somewhere.

I'll mess with it more later, probably the beginning of next week or so.

I like your concept though. Have you thought about building an adrive with a browser in it (maybe Fred's firefox portable) and putting the mesa stuff in a ydrive?

[musher0]

Hello all.

This real lzop will replace the feeble busybox clone at /bin.

Compiled in and for upupbb-light. (It will probably work in
same level Pups, but no guarantees.)

Enjoy!

I will encourage testers to try installing your collection of .pets and seeing how
they perform. I'll mention that on Page 1 when I post the update. (...)

Thanks, jrb.

I'm sure the users will appreciate having the possibility to update major utilities
in upupbb-light to current versions.

~~~~~~~~~

In the same line of thought, please find the latest ( as of this writing ) mksquashfs and
unsquashfs utilities in the attached pet archive. This version includes lzma, lzo and lz4
compressions, in addition to the traditional gzip and xz copressions.

As evidence of how much it's changed, the help file for the newest version of
mksquashfs ( again, as of this writing ) is almost twice as long as the help file
generated by the mksquashfs currently in upupbb.

Enjoy.

~~~~~~~~~

BFN.

[jrb]

Have you thought about building an adrive with a browser in it (maybe Fred's firefox portable) and putting the mesa stuff in a ydrive?

A genius idea dancytron. I have started appending code to Fred's ffportable-inst.gz which will download and build Firefox-portable on a drive chosen by the user and then link it to /usr/local/bin where it can be activated from the internet menu. I will include the script and the activation script will launch the setup on first use and then launch Firefox-portable after that.

As for the ydrive, the idea is to keep ram usage to a minimum, but I could easily make the ydrive available as a separate download.

Edit: May involve a small delay in uploading first update

[musher0]

Hi jrb.

Don't hurry... We're still finding bugs: xlock has to be run directly to work.

BFN.

[musher0]

Hello, all.

Does anyone know the meaning of the message "no local support" appearing just after
boot in /tmp/xerrs.log? ( Please see attached capture. )

I find that weird because < locale -a > lists my locale, and I can write in any editor in
my language.

TIA for any insight.

[musher0]

Hi jrb.

Don't hurry... We're still finding bugs: xlock has to be run directly to work.

BFN.

Update --
I got the Xlock icon on the desktop to work after doing the following:

-- erase file .xlockrc in /root

-- open a console and type < xlock >

-- answer the prompts about a password.
Xlock will start, changing modes +/- every minute.

-- Interrupt xlock by typing your password.
You should now be back on the desktop.

-- Click on the Xlock icon on the desktop. It should now run as expected.

IHTH.

[davids45]

G'day jrb,

A bit more testing of the small ghostscript and gstoraster packages (earlier posts) that get CUPS working my Epson inkjet printer has been fine, but I see that the CUPS PDF printer is still not working.
So I've deleted the CUPS-PDF printer for now.

Quite a few packages will export-to-PDF their output so I just need to find an alternative for more 'stand-alone' PDF print needs.

If I can't find such an alternative, I could try the ghostscript package from peebee's 'normal' BionicPup32 versions.

David S.

[musher0]

Hi jrb and all.

I thought I had reported not being able to type any accented character in CLI -- and
obviously not in the joe or MP editors. Reviewing the pages of this thread, it seems
that I haven't.

In any case, please find the solution attached. Directory /usr/share/X11/locale is
missing in this upupbb32-light, and that is the cause of the console not accepting
accented characters when your system locale is not English.

Follows the procedure to regain the capacity of writing accented characters in
console and console apps:
-- Download the attached in a convenient and safe place;

-- Move or copy it to /usr/share/X11;

-- Once you have it there, either
----- open a console in that directory and type

Code: Select all

unzip locale.zip

OR
----- use SFR's UExtract or
----- the Pupzip utility
to unzip it.

That's it.

I have no merit in finding this solution. I was tipped off by somebody in an Arch Linux
forum thread who had a similar problem, although for a different reason. ( That
ArchLinux user had used Bleachbit to do a clean-up of his files, and Bleachbit
somehow erased the locale.dir file in /usr/share/X11/locale... Moral of this story: do not
use Bleachbit! )

Please note -- The attached locale directory from peebee's BionicPup32 provides the
en_US.utf-8 locale only. ( I.e. and internationalized US locale, that provides accented
characters for occasional use. ) I still need to find a source for the fr_CA.utf-8 and other
accented locales. I'll keep you posted.

~~~~~~~~~~~~~~

Since I'm on a bug-squashing roll...

When I rebooted upupbb32-light to do the above correction, the Pup absolutely refused
to load my customized PuppuPin: on boot-up, it keeps providing jrb's default PuppyPin.

jrb, can you please tell your Pup not to do that anymore? TIA.

( Screen capture attached, BTW. )

~~~~~~~~~~~~~~

Enough bug-squashing for today!

IHTH. Tah-dah.

[jrb]

G'day jrb,

A bit more testing of the small ghostscript and gstoraster packages (earlier posts) that get CUPS working my Epson inkjet printer has been fine, but I see that the CUPS PDF printer is still not working.
So I've deleted the CUPS-PDF printer for now.

Quite a few packages will export-to-PDF their output so I just need to find an alternative for more 'stand-alone' PDF print needs.

If I can't find such an alternative, I could try the ghostscript package from peebee's 'normal' BionicPup32 versions.

David S.

Tackling the Cups problem first. I don't remember CUPS-PDF ever working for me, but "Print to File"->PDF always has.

Have you tried this technique? Seems a shame that CUPS-PDF doesn't work but, as I say, it hasn't worked for me since Puppy4. Maybe I'll leave it out in the first update.