Lucid Puppy Revitalized as 5.2.8.7 - Nov., 2017 - Final

[sindi]

2008 Puppy Forum topic "invisible cursor", says problem is not just with nvidia chips but any
video chip using onboard RAM. Solution is to add to xorg.conf "HWCursor" "off" but this is no
longer the correct format to disable hardware cursor.

I tried "HWCursor" "False" or "Off" or "off"
"SWCursor" "True" etc.


The 2008 discussion mentioned Rerwin and Xorgwizard.
I am upgrading to that era of hardware (2006 desktops, 2008 netbook).
A 2010 linux is ideal (with the generous offer of updated browser).

A separate graphics card might also work.

[rerwin]

I'm doing some at-home testing on the "unstable build" of the next Pale Moon for linux version. It requires, absolute minimum, ffmpeg-0.8 and associated libavcodec-53. Lucid Puppy doesn't seem to have sufficiently up-to-date versions.. I dove into PPM, and came up with the display as shown in the attached screenshot. The main "Puppy Package Manager" dialogue claims that ffmpeg 0.5.9 and associated libavcodec52 0.5.9 is installed.

But the "PPM: version differences" dialogue lists ffmpeg 0.8.11 as being installed. Am I misreading this?

I see the same, so dug into the story behind that. Ffmpeg is normally installed from ubuntu .deb packages; that is what the PPM is seeing. In the first version of 5.2.8.6, a .pet package was added, ffmpeg-0.8.11-lupu528u005. That replaced the versions of libpostproc51 and libswscale0. Wherever the information is about installed deb packages (which I cannot find) it did not get updated, probably because the package was not a .deb. So, we have a hybrid, ffmpeg-0.8.11 with two 0.5.1 dependencies. It is a risk we have taken from the beginning of 5.2.8.6.

The question, now, is whether to update libpostproc51 and libswscale0, because they appear to be newer than the .pet versions. But they are associated with ffmpeg 0.5.9. Or maybe they were what was available when the .pet was built, but have moved on. Their descriptions are:

libpostproc is a library containing older h263 based video postprocessing routines.
libswscale is a library containing video image scaling and colorspace/pixelformat conversion routines.

The 2 updates probably contain security updates. I plan to add them. I attach them here, in case anyone wants to try them.

[Snail]

Walter and Rerwin,

I have never succeeded in getting Sulu4 to run on my T400 laptop or my similar-vintage HP-Compaq desktop. Conversely, I've never had any real difficulty with manual- frugal installs of Sulu2 with either machine, firstboot always comes up properly with the correct resolution. Why would you need to do a full install on the T400, it runs frugal just fine?. I must admit that with Sulu2 being available, I haven't tried very hard to overcome the black screen issue with Sulu4,

[Snail]

The storage device desktop icons are one of the really convenient things about Puppy. When you plug in a USB drive the quick appearance of an appropriate icon is far better than the windows way and much faster too. However, when I plug in my Motorola Moto G2, running Marshmallow using a usb cable and selecting file transfer mode MTP, no convenient icon appears. Hardinfo shows that Sulu2 has detected the phone correctly as XT1068, but I can't see how to mount or access it.. PMount doesn't find it at all.

In Tahr64-6053, I can access the drive. However it doesn't present nicely in the normal Puppy way. There is a different, more windows-like approach of popup panels etc. This is adequate but clunky,when compared to the usual puppy way.

[rerwin]

The modeset fix on the previous page was a start on making Lucid's Intel/nVidia/radeon mode settings similar to those of other puppies, mode 1. I now have a more complete resolution to some of the bootup problems with the video display. I test it on my nVidia GeForce4 MX 460, where without mode 0 it presents lower resolution desktops, sometimes with messed up colors. Even if it boots to full resolution with a fresh pupsave environment, saving and rebooting results in a lower resolution. Mode 0 prevents that.

The attached package initializes the i915, nouveau and radeon .conf files to use mode 1, like other puppies. It also has a udev rules file and command-line script to manage setting of mode 0 for special cases, allowing a user to set mode 0 for currently installed VGA controllers that use those drivers.

The script, modeset0, is for use when a user finds that the display is incorrect or not presented, at all. If viewing an incorrect video display, the user can ctrl-alt-bksp to get to the command prompt (outside of X). At the command prompt, the user can enter "modeset0" to create a udev rule to set mode 0 for the particular VGA controller.

At subsequent bootups, the rule file in the package, /etc/udev/rules/00-modeset0.rules, resets the modes to 1, then executes any rules it has for particular VGA controllers (currently only my MX460), to force mode 0. After that, the user's newly created rule file sets mode 0 for the current display.

This setup can be removed by re-running modeset0 from the command line (outside of X). It detects the presence of the users rule file and offers to remove it. One might ask why all this to simply set a configuration file.

This arrangement adapts to the change-out of display controllers, so that the modeset0.conf file gets removed when the controller hardware is absent. The user can run modeset0 again if the replacement controller needs it.

I request that anyone successfully using the tool to resolve a display issue, please, tell us and give the content of the created rule file (or attach it), so that it can be added to the 00-modeset0.rules file (in a generalized form), to avoid others needing to create their own special rule file.

EDIT: Re-uploaded the package after removing redundant initialization script, after download #2.

[rerwin]

Regarding the T400 and its VGA controller:

Code: Select all

00:02.0 VGA compatible controller [0300]: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller [8086:2a42] (rev 07)

During my research I ran across the following suggestion:

intel mobile 4 chipset (GMA500. etc.) - with the black screen after the splash from the infamous intel mobile 4 chipset -
actually it is the backlight - it is backwards? Boot the computer, and as root, edit the file /etc/rc.local
just before the exit 0 line, add the line "setpci -s 00:02.0 F4.B=00"

Has anyone tried this?

[belham2]

Regarding the T400 and its VGA controller:

Code: Select all

00:02.0 VGA compatible controller [0300]: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller [8086:2a42] (rev 07)

During my research I ran across the following suggestion:

intel mobile 4 chipset (GMA500. etc.) - with the black screen after the splash from the infamous intel mobile 4 chipset -
actually it is the backlight - it is backwards? Boot the computer, and as root, edit the file /etc/rc.local
just before the exit 0 line, add the line "setpci -s 00:02.0 F4.B=00"

Has anyone tried this?

Hi Rerwin,

Do you think that a problem I am seeing in my setup Lupu is related to what you've been discussing these past few msgs? I am AMD w/ motherboards that all have integrated ATI Radeon (RS880/Radeon HD versions). These are ~5-6 yr old machines, a bit old I guess but still they scream for my needs. Depending on which one I am using, they'll have a minimum of 2GB and max 8GB RAM. On all my diff machines (I abhor Intel on desktops, haha, whether rightly or wrongly, but luv Intel on laptops) my setup Lupu-----running off a USB frugally-installed with savefile---motors along great except for this annoying thing:

every dialog box that comes back for any application and/or system setting, where I can enter things to change whatever, that dialog box entry line goes completely black with screen-gibberish the moment the mouse clicks into that particular line. If I click on another entry line in the dialog box, that other line immediately also goes to blacked-out gibberish and the line that just had the mouse-click w/ gibberish goes back to being perfectly clear and normal appearing. I have to literally guess at what I am entering, because you can't see it. Slow, careful keyboard pecking is what happens, and then testing the setting and.or application to see if it entered like I typed. It's frustrating (browser and email apps, though, are immune to this). Or, dialog boxes will have sections completely blacked out and/or made to look gibberish See the pic below, using the "fbpanel configuration" as an example. It's like this for everything save the browser and email. Any tips and hints you could give me to stop this behavior would be greatly appreciated. I am at the point where I am wondering if it is the Xorg, but everything in Xorg is set correctly, from screen resolution to all that I can see. I've tried many different Xorg settings (dropping into X), but nothing is helping


(P.S. Also, weirdly, this is the first pup I've ever had that doesn't allow me to make the background in conky transparent..no matter how I enter things into conky, it has to be a color background...makes me wonder if the behavior I am describing here is related)
[edit: spelling corrections and to post correct pic]

[Les Kerf]

I am running lupusuper4plus from a USB thumb drive on a 2009 iMac with mostly good success.

I have done quite a bit of searching over the past week, trying to find a way to get this system to go to sleep (suspend) rather than doing a full shutdown.

There seems to be plenty of options for laptops, but very little for a desktop unit.

It seems to me that there should be a way, as the hardware certainly allows this when operating from OSX.

Any help will be appreciated.
Thanks,
Les

[backi]

Hi Les Kerf !

If you want to set your machine to sleep........

Install with Packetmanager acpitool (or acpitools not quite sure )

make a little script with Geany either :

acpitool -s

or:

bin!/sh
acpitool -s

make this scripts executable ...... hope you know how to do it .

Click on these scripts to execute them .

[rerwin]

Do you think that a problem I am seeing in my setup Lupu is related to what you've been discussing these past few msgs? I am AMD w/ motherboards that all have integrated ATI Radeon (RS880/Radeon HD versions). These are ~5-6 yr old machines, a bit old I guess but still they scream for my needs. Depending on which one I am using, they'll have a minimum of 2GB and max 8GB RAM. On all my diff machines (I abhor Intel on desktops, haha, whether rightly or wrongly, but luv Intel on laptops) my setup Lupu-----running off a USB frugally-installed with savefile---motors along great except for this annoying thing:

every dialog box that comes back for any application and/or system setting, where I can enter things to change whatever, that dialog box entry line goes completely black with screen-gibberish the moment the mouse clicks into that particular line. If I click on another entry line in the dialog box, that other line immediately also goes to blacked-out gibberish and the line that just had the mouse-click w/ gibberish goes back to being perfectly clear and normal appearing. I have to literally guess at what I am entering, because you can't see it. Slow, careful keyboard pecking is what happens, and then testing the setting and.or application to see if it entered like I typed. It's frustrating (browser and email apps, though, are immune to this). Or, dialog boxes will have sections completely blacked out and/or made to look gibberish See the pic below, using the "fbpanel configuration" as an example. It's like this for everything save the browser and email. Any tips and hints you could give me to stop this behavior would be greatly appreciated. I am at the point where I am wondering if it is the Xorg, but everything in Xorg is set correctly, from screen resolution to all that I can see. I've tried many different Xorg settings (dropping into X), but nothing is helping

That looks like a hardware/driver problem. The black-and-white part is what I would expect for the entire screen for a modeset issue. The first thing to try would be booting with radeon.modeset=0, to eliminate that as a possibility.

To help me research this, please get the card info for each VGA device, using the command string:

Code: Select all

lspci -nnvd $(lspci -nn | grep VGA | grep -Eo '[0-9a-f]+:[0-9a-f]+' | grep '.........')

For my device, that shows:

• 01:00.0 VGA compatible controller [0300]: NVIDIA Corporation NV17 [GeForce4 MX 460] [10de:0170] (rev a3) (prog-if 00 [VGA controller])
  Subsystem: CardExpert Technology Device [10b0:0002]
  Flags: 66MHz, medium devsel, IRQ 11
  Memory at c2000000 (32-bit, non-prefetchable) [size=16M]
  Memory at e0000000 (32-bit, prefetchable) [size=128M]
  Memory at e9000000 (32-bit, prefetchable) [size=512K]
  [virtual] Expansion ROM at e8000000 [disabled] [size=128K]
  Capabilities: [60] Power Management version 2
  Capabilities: [44] AGP version 2.0
  Kernel modules: nouveau

That should improve chances of finding a solution.
Richard

[prehistoric]

Hi rerwin,

This may help you find what is going on with machines having the nvidia 6150/6150LE video graphics chips with shared video memory.

I am running sulu2 with the proprietary nvidia 304 driver right now, and it seems flawless. Here's the kernel and C library version:

Code: Select all

Kernel		: Linux 3.2.48 (i686)
Version		: #1 SMP Thu Jul 11 12:20:55 GMT-8 2013
C Library		: GNU C Library version 2.11.1 (stable)
Distribution		: Puppy Linux

The kernel video module is naturally just called nvidia, but comes from the nvidia 304.117 driver package you linked on the first page of this topic.

When I originally booted this system the nouveau module seems to have screwed up, and I ended up running vesa at 800x600 until I replaced the driver. I blacklisted nouveau on the kernel line, even if that isn't strictly necessary.

There's a reason nvidia interacts with all kinds of things on this machine. I've attached the output from lspci to illustrate.

If you like I can run experiments to determine what is going wrong on that initial boot, but the real solution has to be that proprietary module.

[belham2]

To help me research this, please get the card info for each VGA device, using the command string:

Code: Select all

lspci -nnvd $(lspci -nn | grep VGA | grep -Eo '[0-9a-f]+:[0-9a-f]+' | grep '.........')

Richard

Hi Rerwin,

Thanks for responding. Here's the lspci output:

Code: Select all

# lspci -nnvd $(lspci -nn | grep VGA | grep -Eo '[0-9a-f]+:[0-9a-f]+' | grep '........')
01:05.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] RS880 [Radeon HD 4200] [1002:9710] (prog-if 00 [VGA controller])
	Subsystem: Biostar Microtech Int'l Corp Device [1565:1704]
	Flags: bus master, fast devsel, latency 0, IRQ 18
	Memory at d0000000 (32-bit, prefetchable) [size=256M]
	I/O ports at d000 [size=256]
	Memory at feaf0000 (32-bit, non-prefetchable) [size=64K]
	Memory at fe900000 (32-bit, non-prefetchable) [size=1M]
	Capabilities: [50] Power Management version 3
	Capabilities: [a0] Message Signalled Interrupts: Mask- 64bit+ Queue=0/0 Enable-

# lspci -nnvd $(lspci -nn | grep HDMI | grep -Eo '[0-9a-f]+:[0-9a-f]+' | grep '........')
01:05.1 Audio device [0403]: Advanced Micro Devices, Inc. [AMD/ATI] RS880 HDMI Audio [Radeon HD 4200 Series] [1002:970f]
	Subsystem: Advanced Micro Devices, Inc. [AMD/ATI] RS880 HDMI Audio [Radeon HD 4200 Series] [1002:970f]
	Flags: bus master, fast devsel, latency 0, IRQ 26
	Memory at feae8000 (32-bit, non-prefetchable) [size=16K]
	Capabilities: [50] Power Management version 3
	Capabilities: [a0] Message Signalled Interrupts: Mask- 64bit+ Queue=0/0 Enable+
	Kernel driver in use: HDA Intel
	Kernel modules: snd-hda-intel

# 

Also, I tried the "radeon.modeset=0" in my grub4dos entry. Wasn't sure if I put in there correctly, or even if grub4dos can even handle it (like grub & grub2 do). My Lupu grub4dos entry, after I put the radeon.modeset=0 in, is this:

Code: Select all

title Lupu 528 (sda1)
  find --set-root --ignore-floppies --ignore-cd /Lupu-dec-2017/lupu_528.sfs
  kernel /Lupu-dec-2017/vmlinuz psubdir=Lupu-dec-2017 pmedia=usbflash pfix=fsck radeon.modeset=0
  initrd /Lupu-dec-2017/initrd.gz

and I tried this, leaving the radeon part out: modeset=0:

Code: Select all

title Lupu 528 (sda1)
  find --set-root --ignore-floppies --ignore-cd /Lupu-dec-2017/lupu_528.sfs
  kernel /Lupu-dec-2017/vmlinuz psubdir=Lupu-dec-2017 pmedia=usbflash pfix=fsck modeset=0
  initrd /Lupu-dec-2017/initrd.gz

Both ways booted up quickly, but the problem still persisted.

Thanks again

[rerwin]

belham2,
Thanks for those lspci items.

For setting mode 0, the first syntax is correct. Since the default is neither 0 or 1, have you tried radeon.modeset=1? Gotta rule that out, too. However, what you are seeing is occurring long after bootup, so probably is not affected by the KMS mode.

Unfortunately, I don't have a link to a proprietary driver for radeon. I looked into that quite some time ago and had to "give up", as it didn't seem to be working out.

prehistoric,
Thanks for reminding us to try a nVidia driver when all else fails.

[belham2]

belham2,
Thanks for those lspci items.

For setting mode 0, the first syntax is correct. Since the default is neither 0 or 1, have you tried radeon.modeset=1? Gotta rule that out, too. However, what you are seeing is occurring long after bootup, so probably is not affected by the KMS mode.

Unfortunately, I don't have a link to a proprietary driver for radeon. I looked into that quite some time ago and had to "give up", as it didn't seem to be working out.

Hi Rerwin,

Been sitting here Googling and reading the past few hours, and have come to the conclusion this particular problem and motherboard (that I like to use the pups on) is getting passed on. What I mean is I am going to swap out my youngest's Biostar motherboard and give 'em this one (the youngest runs full Ubuntu, and this pup-problematic motherboard were dealing with here has no problems w/ full Ubuntu judging by the past Ubuntu-iterations I have ran on it at various times).

Thanks again for taking the time, but the more I read about issues like this, the more I realized this problem both isn't worth spending time on, plus it is also possibly a dead end. So, hand it on down to one of my rugrats, they'll never know (haha), and we can let it go

[Les Kerf]

Hi Les Kerf !

If you want to set your machine to sleep........

Install with Packetmanager acpitool (or acpitools not quite sure )

make a little script with Geany either :

acpitool -s

or:

bin!/sh
acpitool -s

make this scripts executable ...... hope you know how to do it .

Click on these scripts to execute them .

Thanks for the help.

I installed acpitools but this is the result when I run it:

# acpitool -s
Function Do_Suspend : could not open file : /proc/acpi/sleep.
You must have write access to /proc/acpi/sleep to suspend your computer.

When I look in /proc/ascpi there is no file named "sleep", only alarm, event, and wakeup.

Is this distro not capable of sleep?
Les

[backi]

Hello Les Kerf !

Don`t use Lucid Puppy in the moment ..
but never had any trouble in every Puppy i used (Tahrpup,Precise ,DpupExprimo,Xenial Pup,Raring Pup) ,even in Lucid Pup (Three Headed Dog) acpitool worked for me .

Can not give an answer ......maybe depends on Hardware ............

Maybe somebody else does know .

Greetings

[Les Kerf]

......maybe depends on Hardware ............

Possibly.

But the lack of /proc/acpi/sleep doesn't seem hardware related

Thanks for the help,
Les

[prehistoric]

...prehistoric,
Thanks for reminding us to try a nVidia driver when all else fails.

I would not recommend including all these possible drivers with the standard Lupu release, but it might be possible to make recommendations based on the results of hardware analysis and the kernel in use. If the files are at stable locations on the Internet, you could even offer to download the correct one. Older hardware is not going to change requirements.

Here's the relevant line from lspci which told me which driver to use:

Code: Select all

VGA compatible controller: NVIDIA Corporation C51PV [GeForce 6150] (rev a2)

Collect a number of these and the code to select a driver would be straightforward.

A little bit of instruction text would need to be added, but I think this would get past the point where many people give up. I can testify that most people are baffled by the idea of different kernels for the same OS, plus the sheer range of different drivers offered by nvidia. (Those telling us the latest drivers are backwards compatible are often the same people selling new hardware, which makes such advice suspect.)

Installing a recommended driver should be easier than having people compile drivers from source, which is what you do when really at the point where all else fails. (I did this on Fatdog 710.)

As before, I am willing to run experiments to test proposed fixes to allow nouveau to work OOTB. I still have another machine with the same nvidia chip set that is not required for everyday use.

[rerwin]

I would not recommend including all these possible drivers with the standard Lupu release, but it might be possible to make recommendations based on the results of hardware analysis and the kernel in use. If the files are at stable locations on the Internet, you could even offer to download the correct one.

Lucid pup does that, up to a point. Click on the Quickpet desktop icon and select the 'Drivers' tab, then on "Click here to test your graphics card". The window that appears describes your card and suggests the appropriate nVidia driver. It doesn't go further, but you know you can find the links in the third message in this thread.

[prehistoric]

...Lucid pup does that, up to a point. Click on the Quickpet desktop icon and select the 'Drivers' tab, then on "Click here to test your graphics card". The window that appears describes your card and suggests the appropriate nVidia driver. It doesn't go further, but you know you can find the links in the third message in this thread.

I know, but I no longer consider myself a naive user. It might be good to specifically mention a place to find these pre-compiled drivers, though I'm not sure how to keep this stable.

I'm afraid I had trouble with Quickpet, and quickly did my own search. Now I realize I should have made notes on what went wrong. Part of my problem was that programs offered through Quickpet like browsers are way out of date. You aren't likely to find any recent browser weighing in at 15 MB. Even going from Seamonkey 2.40 which I have currently installed, to 2.46 requires version 2.15 of the C libraries while we are stuck at 2.11. I'm starting to wonder if we could mine Precise Puppy for code without doing a complete rewrite. Tahrpup is much more up to date, but has largely dropped compatibility with old machines.

My concern is largely with security issues in old browsers. I am really appalled by some of the tricks I see sites trying to play. Myself, I'm running the latest Internet software I can get on Fatdog, and still running these as "spot". Also, I am determined to keep running text email (Sylpheed) to avoid running untrusted code every time I exchange messages. Every day I encounter spam with some very sneaky tricks in it.

Maybe BarryK's experiments with containers will give us a better solution. How did we get into a situation where you have to be paranoid to survive?