Page 23 of 44

Posted: Sat 01 Sep 2012, 09:45
by LateAdopter
Hello BarryK

Thanks for the answer.
Apart from that, why are you upgrading the package lists? There is no need to. You are testing Precise Puppy beta4, which is built with the latest package lists.
Since the package manager seems to handle dependencies well, I thought I would try the ultimate test. That is to install VLC with its hundreds of dependencies from the Ubuntu repo and see what happened!

The main precise repo has VLC 2.01 which is quite buggy, but the precise-updates has VLC 2.03, which is the current version.

There seems to be a problem with the package manager, though. Now that I have a lot of user loaded packages, it takes an infinite amount of time to do the package search. It worked OK to start with.

I treat installing .debs as a one-way ticket. So I'll delete the save file and start again!

Posted: Sat 01 Sep 2012, 11:55
by 01micko
I found an interesting OpenGL benchmark and compiled it for Slacko, but it does work in Precise with one added dependant lib:

libglew1.5_1.5.7.is.1.5.2 (type it in PPM search box, dont get the other version!)

Results for me are identical in Slacko and Precise on my Radeon Cedar card.

Get it from this post


-------------------------

Re: nvidia driver

Posted: Sat 01 Sep 2012, 11:59
by BarryK
shinobar wrote:nvidia-glx-304.43-k3.2.28.sfs is available.
http://shino.pos.to/party/bridge.cgi?puppy/precise/
Works on Puppy precise beta4 5.2.71 and 73, but does not with 5.2.72.
Made by the getnvidia-0.7.
http://www.murga-linux.com/puppy/viewtopic.php?t=72405

Usage:
1. Boot presice beta4(exept 72) with boot parameter 'puppy pfix=nox'.
2. Run 'xorgwizard', choose 'vesa'.
3. 'xwin' to get graphical desktop.
4. Click the nvidia-glx-304.43-k3.2.28.sfs and choose 'Install'.
5. Skip the launcher 'Nvidia config'.
6. Run 'xorgwizard', choose 'probe'.
Thanks, I have uploaded this, and your Gimp SFS, to ibiblio, see my blog post:

http://bkhome.org/blog/?viewDetailed=02980

Re: function4puppy4 corrections, take 4

Posted: Sat 01 Sep 2012, 12:25
by BarryK
rerwin wrote:Barry,
I tested forcing ("adding") use of a module that has a firmware package and found the the package still does not get copied into place. That was the major reason I updated functions4puppy4 -- to obtain firmware from a directory instead of a tarball. Here is a corrected version and the related difference file.

Please consider adding this to woof before releasing a precise candidate. Thanks.
Richard
Thanks, done, it will be in the next upload of Woof.

Re: Precise Puppy beta4, August 30, 2012

Posted: Sat 01 Sep 2012, 12:27
by BarryK
Billtoo wrote:I compiled vlc-2.0.3 and qt-4.8.2 in precise puppy.

vlc is about a 12mb download and qt-4.8.2 is about 45mb.

The download link for qt is:

http://www.datafilehost.com/download-97ac889a.html

The download link for vlc is:

http://www.datafilehost.com/download-bc2923b8.html
Nice, but I won't host these on ibiblio nor can they be added to PPM, as they need to be split into EXE, DEV, DOC and NLS pets.

82845G/GL[Brookdale-G]/GE works!

Posted: Sat 01 Sep 2012, 15:32
by rerwin
I have, at last, tried precise pup on my IBM NetVista with the problematic 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device. Much to my amazement, it "just works"! Even though the i915 modeset=1 is in effect. In lucid pup, I had to have it set at 0 to get past the black screen issue. Several other puppies with various kernels seemed to have the black screen problem. But I am pleased that this long-term-support puppy is not one of them.

Now the problem appears to be with the nouveau driver and black screens.
Richard

Posted: Sat 01 Sep 2012, 16:44
by rjbrewer
Using Htop to observe "load averages" while watching
videos; I get about 10 times the load average reading with
Precise compared to Lucid or Wary.

There isn't any temperature differences or or other
indications of overload??

Edit:
The videos are slightly smoother and more solid on the
other Puppies.

Posted: Sat 01 Sep 2012, 19:48
by BHINTZ
I would certainly support inclusion of suspend/resume as part of the standard OS. yr1945 suggested acpitool-0.5.1.2 which seems to work fine as long as I directly initiate the sleep function (I made a short cut to desktop). However, it does not respond to closing the lid.

pupshutdown does not include a suspend on my machine (dell latitude D630)
but I am leaving it installed as it might be necessary for current functioning.

This is very quick, but does not retain log in information when shut down and restarted later. It does retain this information over sleep/resume.

An excellent version. Many thanks

bob

Re: 82845G/GL[Brookdale-G]/GE works!

Posted: Sat 01 Sep 2012, 21:10
by 01micko
rerwin wrote:I have, at last, tried precise pup on my IBM NetVista with the problematic 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device. Much to my amazement, it "just works"! Even though the i915 modeset=1 is in effect. In lucid pup, I had to have it set at 0 to get past the black screen issue. Several other puppies with various kernels seemed to have the black screen problem. But I am pleased that this long-term-support puppy is not one of them.

Richard
Hi Richard

There are 2 reasons for this.

1) KMS was not configured properly in k2.6.33.2 which shipped with Lucid. I doubt that it would have made a difference for Brookdales anyway.

2) The last couple of years have seen some big improvements in KMS and the Xorg drivers. They have now deprecated UMS. So if KMS doesn't work the only option is Vesa.

Posted: Sun 02 Sep 2012, 00:13
by bigpup
rerwin wrote:Thanks, bigpup, for trying the pet package. Given the info you provided, it would seem that the logic should work as intended for your card. The main thing to verify is that there is a hidden flag file:
/tmp/.nouveau_modeset_0.
(Click the eye option in rox, to see it.)

If it is not present, then the rules file is faulty. If it is present, then either the config file is incorrect or the flag gets set too late for the loading of the module. Both pieces were cloned from the lucid pup files for the i915 module.

With the i915 case, when used outside of the lupu distro (with other kernels), the results were intermittent at best. So, this technique may be inadequate.

Another way to test the concept is to restore the version of nouveau.conf that sets the mode to 1 and modify it to set 0. If that works, it would confirm that the flag is too late.
Richard
/tmp/.nouveau_modeset_0 file is present.

/etc/modeprobe.d/nouveau.conf
At present has this:

Code: Select all

install nouveau \
 [ -e /tmp/.nouveau_modeset_0 ] \
  && /sbin/modprobe --ignore-install nouveau modeset=0 $CMDLINE_OPTS \
  || ( [ -e /tmp/.nouveau_nomodeset ] \
        && /sbin/modprobe --ignore-install nouveau $CMDLINE_OPTS \
        || /sbin/modprobe --ignore-install nouveau modeset=1 $CMDLINE_OPTS )
What should I change in this code?
I am using the Vesa driver, after booting pfix=nox and using Xorgwizard to select Vesa. Did it make the changes you are talking about to this file?

Re: Precise Puppy beta4, August 30, 2012

Posted: Sun 02 Sep 2012, 03:56
by Billtoo
BarryK wrote:
Billtoo wrote:I compiled vlc-2.0.3 and qt-4.8.2 in precise puppy.

vlc is about a 12mb download and qt-4.8.2 is about 45mb.

The download link for qt is:

http://www.datafilehost.com/download-97ac889a.html

The download link for vlc is:

http://www.datafilehost.com/download-bc2923b8.html
Nice, but I won't host these on ibiblio nor can they be added to PPM, as they need to be split into EXE, DEV, DOC and NLS pets.
I'll attach what I can to this post and a link to the vlc nls pet, qt-482 is too big to upload to datafilehost, there is a 100mb file size limit and the qt-482_doc pet is over that.

The download link to vlc-2.0.3_NLS is:

http://www.datafilehost.com/download-563ae99c.html

Posted: Sun 02 Sep 2012, 05:34
by James C
5.2.72 live pfix=ram on another old test box. Sound,internet and display all working and correct on initial boot.

# report-video
Precise Puppy, version 5.2.72 on Sun 2 Sep 2012

Chip description:
00
12.0 VGA compatible controller
NVIDIA Corporation C68 [GeForce 7050 PV / nForce 630a] (rev a2)
oem: NVIDIA
product: MCP67 - mcp68-01 Chip Rev

X Server: Xorg
Driver used: nouveau

X.Org version: 1.11.3
dimensions: 1024x768 pixels (270x203 millimeters)
depth of root window: 16 planes

...the above also recorded in /tmp/root/ as report-video,
and archived with xorg.conf and Xorg.0.log as report-video-full.gz
#
# glxgears
nvfx_screen_get_param:95 - Warning: unknown PIPE_CAP 30
nvfx_screen_get_param:95 - Warning: unknown PIPE_CAP 30
nvfx_screen_get_param:95 - Warning: unknown PIPE_CAP 55
nvfx_screen_get_param:95 - Warning: unknown PIPE_CAP 56
nvfx_screen_get_param:95 - Warning: unknown PIPE_CAP 59
nvfx_screen_get_param:95 - Warning: unknown PIPE_CAP 58
nvfx_screen_get_param:95 - Warning: unknown PIPE_CAP 30
Running synchronized to the vertical refresh. The framerate should be
approximately the same as the monitor refresh rate.
glxgears: nvfx_state_fb.c:124: nvfx_framebuffer_validate: Assertion `util_format_get_stride(fb->zsbuf->format, fb->width) <= nvfx->hw_zeta.pitch' failed.
Aborted
#
-Computer-
Processor : 2x AMD Athlon(tm) 64 X2 Dual Core Processor 3800+
Memory : 903MB (172MB used)
Operating System : Unknown distribution
User Name : root (root)
Date/Time : Sun 02 Sep 2012 05:32:23 AM CDT
-Display-
Resolution : 1024x768 pixels
OpenGL Renderer : Gallium 0.4 on NV67
X11 Vendor : The X.Org Foundation
-Multimedia-
Audio Adapter : HDA-Intel - HDA NVidia

Pmount problem report

Posted: Sun 02 Sep 2012, 18:34
by gcmartin
Ran into this issue that may not have been noticed.

Here to help

Posted: Sun 02 Sep 2012, 19:04
by pemasu
I do not think it is issue per se. Pmount works just fine without mut. Barry Kauler has posted in his blog several times that ancient mut feature is not in Puppies anymore. I think this "issue" just confirms it.
Mut2 is also very old, and I don't know if it even works properly anymore. It was developed years ago by Forum member Jesse, who also developed our 'elspci' and 'puppyinputdetect'. We do have the sources for those last two and still use them.

I don't have Mut2 in any of my Puppy builds.

Posted: Sun 02 Sep 2012, 20:10
by rerwin
bigpup wrote:/tmp/.nouveau_modeset_0 file is present.

/etc/modeprobe.d/nouveau.conf
At present has this:

Code: Select all

install nouveau \
 [ -e /tmp/.nouveau_modeset_0 ] \
  && /sbin/modprobe --ignore-install nouveau modeset=0 $CMDLINE_OPTS \
  || ( [ -e /tmp/.nouveau_nomodeset ] \
        && /sbin/modprobe --ignore-install nouveau $CMDLINE_OPTS \
        || /sbin/modprobe --ignore-install nouveau modeset=1 $CMDLINE_OPTS )
What should I change in this code?
I am using the Vesa driver, after booting pfix=nox and using Xorgwizard to select Vesa. Did it make the changes you are talking about to this file?
That is my code that replaces the existing nouveau.conf file, to make it "smart". It is unrelated to Xorgwizard. It intercepts the loading of nouveau to choose the mode to set. But I think all of it is moot!

I have been experimenting on my PC that uses nouveau, adjusting the mode in both the boot parameter and in that config file -- but in "reverse". I verified that the auto_modeset package performs as expected, so can be trusted. My PC gets the black screen if the mode is 0 in the config file. If I leave the config file's mode at "1" but use boot parameter nouveau.modeset=0, all is well. I arrived at a conclusion about how this works.

Apparently, the boot parameter is in effect when nouveau is used initially, but the config-file modeset overrides that for the continuation of the use of nouveau. For my PC, I infer that the initial usage (whatever that actually is) works no matter how the mode is set, or not set at all. After that, it works only if the mode is "1".

I imagine that your PC is not so tolerant for the inital usage of nouveau, but once the mode is set to "1" by the (existing) config file, it works just fine. So, I think you are forced to use the nouveau.modeset parameter to get past the initial use of nouveau.

As I have no idea what is being done in the setup of the graphics drivers, this is about as far as I can go. Obviously, my auto_modeset package is useless, since it affects only what happens after the initial usage of nouveau. I hope this helps 01micko to focus on that initial phase to determine what it is that is affected by the boot argument, before the config file supercedes it.

Thank you, again, for reporting and working this issue.
Richard

Posted: Mon 03 Sep 2012, 07:56
by feelsub
Billtoo wrote:
I compiled vlc-2.0.3 and qt-4.8.2 in precise puppy.

vlc is about a 12mb download and qt-4.8.2 is about 45mb.

The download link for qt is:

http://www.datafilehost.com/download-97ac889a.html

The download link for vlc is:

http://www.datafilehost.com/download-bc2923b8.html


Nice, but I won't host these on ibiblio nor can they be added to PPM, as they need to be split into EXE, DEV, DOC and NLS pets.
gimp-2.8.2-1-precise.sfs is available.
http://shino.pos.to/party/bridge.cgi?puppy/precise/
Based on the gimp
how do you know if one doesn't follow the forum???

General and easier for all puppies:
the ability to search & install applications. Why not centralize search on applications' name, its version and finally possible sources & install mode.

A CSV file can do that; for instance:
GIMP;2.8.2;pet; i686;URL1;URL2;URL3; ....
GIMP;2.6.12;sfs; i386;URL1;URL2;URL3; ....
[/code]

HP5200 network printer still unable to work

Posted: Mon 03 Sep 2012, 07:59
by tiangeng
HP5200 network printer still unable to work.
Slacko530 use CUPS1.4.8,it work fine.

Re: HP5200 network printer still unable to work

Posted: Mon 03 Sep 2012, 11:35
by rcrsn51
tiangeng wrote:HP5200 network printer still unable to work.
1. What is the full model name of your printer?

2. What driver are you using when you install the printer in CUPS? The built-in Gutenprint driver? An HPLIP package?

3. What is the exact nature of the problem? Does CUPS not detect the printer? Does it print garbage? From a CUPS test page? From an application?

4. Does it work when connected directly by USB?

Finding applications

Posted: Mon 03 Sep 2012, 14:47
by mikeslr
@ feelsub,

Devs, having spent long hours developing applications and discussing them on active Puplet threads, are too exhausted to note their existence on the "Additional Applications" sub-forum, http://murga-linux.com/puppy/index.php?f=63. Fortunately, the Puppy Forum is a do-acracy. A users, who finds something useful not cataloged on the Additional Applications Sub-Forum can, him-or-herself, write a short, descriptive post with a link to the post(s) containing the application's discussion and download link. If the user thinks about it. Or he or she can do what the rest of us do: complain about the lack of organization and fall back to a well-minded, http://www.wellminded.com/puppy/pupsearch.html, search to find something.

mikesLr

Finding applications --Deleted >Double Poost

Posted: Mon 03 Sep 2012, 14:49
by mikeslr
Deleted>Double post