Wary 0.9.8 (098) feedback

[rerwin]

rerwin
Seeing that the modem works now, I went ahead and did a full HD install.
As expected the modem was not found on first boot so I installed that last pet package you gave me and rebooted. I probed for the modem and still no modem, so I started looking around in bootmanager and found that the snd_intel8x0m was not loading so I added it manually and rebooted.
ttySL0 found and working fine!

But did you immediately reboot after the first bootup? Please try that before attempting a workaround. As it stands, I cannot make sense of what happened.

The first time, an HSF driver is used, but fails, setting things up so that the intel8x0m driver get used the next bootup. Give puppy a chance to handle it on its own, before intervening.

Would you verify that yours works that way? You can do that by doing a pupdial CHOOSE > ERASE ERASE. Click the ERASE (Okay) sequence twice, then reboot. You should see "no modem detected" in the connect wizard. Immediately reboot - the ttySL0 device should be detected that time.

This (so far undocumented, but someday) technique is needed only for certain built-in modems that might use one or the other (HSF, Intel) driver, depending on the modem chipset involved. The double ERASE corresponds to the double boot requirement. The first ERASEs the ttySL0 device - rebooting at that point would re-detect the ttySL0. The second ERASE clears the disabling of the HSF driver, so that it will be used on the next bootup. Normally, you would not do the double ERASE, because you would want ttySL0. It is mainly for testing the "two-boot" function.

Curiously, my laptop with the same modem as yours (hardware ID, anyway) used to behave as I describe here. But now the ttySL0 device is detected right away! That is a reason one cannot depend on which module gets the modem.
Richard

[rerwin]

Barry,
The HSF modem lockup during connection is now corrected in my fix package here: http://www.murga-linux.com/puppy/viewto ... 639#467639

The problem is caused by an incorrect build instruction regarding the "nvm" tarball creation. I have corrected the problem and will PM you the corrected instructions for archiving.

The issue also applies to the HCFPCI modem support, so I have corrected that, too. However, I cannot test that because I don't have a usable HCF modem.
Richard

[Sit Heel Speak]

...I installed the ati-driver-installer-10-10-x86.x86_64.run driver on this laptop in wary 098. I had the kernel_src-2.6.31.14-patched.sfs and the wary_devx_098.sfs files installed and I installed it by
exiting to the prompt and running the file.

Still not working here. Asus M2NPV-VM board, ATi HD 4550 pci-e vidcard.

The (new, 2nd version) .pet package installs OK. I reboot, it comes up with the yafsplash telling me to restart xwin, I exit X to the prompt, xorgwizard, XOrg, Probe, choose fglrx, test. Test does not succeed. No black screen, no "If you can see this it's OK", it just blinks and comes up showing blanks before the refresh rates and driver-name. Attempting to run xwin anyway gives a black screen.

The compile of ati-driver-installer-10-10-x86.x86_64.run succeeds, and I did then the /lib/modules/fglrx/make_install.sh.

Then did aticonfig --initial, and that produces what appears to be a good /etc/X11/xorg.conf.

But, nope, still won't run.

I have a vague memory of Mark Ulrich, last year, editing xorgwizard to deal with fglrx. Late night here, will look into that angle of it in the morning.

Is anyone else succeeding (or, failing similarly) with Catalyst in Wary?

[James C]

Wary 098 live pfix=ram on another old XP box that I had on the shelf.

Working sound and internet as well as the correct 1440x900 resolution on initial boot. As a side note, can't even get over 1024x768 in XP .



VIDEO REPORT: Wary Puppy, version 098

Chip description:
oem: ATI RAGE128

Driver used by Xorg:
r128

Video mode used by Xorg:
Resolution: Depth 24 Depth: "Display"

...the above also recorded in /tmp/report-video
#


-Computer-
Processor : Intel(R) Pentium(R) 4 CPU 1.60GHz
Memory : 513MB (135MB used)
Operating System : Puppy Linux 0.09
User Name : root (root)
Date/Time : Tue 16 Nov 2010 11:53:54 PM CST
-Display-
Resolution : 1440x900 pixels
OpenGL Renderer : Unknown
X11 Vendor : The X.Org Foundation
-Multimedia-
Audio Adapter : ICH - Intel 82801BA-ICH2
-

[DaveS]

Barry, for a long time now, selecting menu>desktop>JWM Configuration>Apply Current GTK Theme to JWM has caused the contents of /root/.jwmrc-tray to be deleted leaving the user with no tray at the bottom of the screen. I have posted about it before and helped a few folks recover their tray, but I dont know how to fix the basic fault.

[BarryK]

Barry, for a long time now, selecting menu>desktop>JWM Configuration>Apply Current GTK Theme to JWM has caused the contents of /root/.jwmrc-tray to be deleted leaving the user with no tray at the bottom of the screen. I have posted about it before and helped a few folks recover their tray, but I dont know how to fix the basic fault.

Yes, thanks for reporting it again. I guess that I have been hoping someone might fix it!

[shinobar]

Barry, for a long time now, selecting menu>desktop>JWM Configuration>Apply Current GTK Theme to JWM has caused the contents of /root/.jwmrc-tray to be deleted leaving the user with no tray at the bottom of the screen. I have posted about it before and helped a few folks recover their tray, but I dont know how to fix the basic fault.

Yes, thanks for reporting it again. I guess that I have been hoping someone might fix it!

I am not sure but i do not see that happens on LupQ.
The Jwmconfig2 in LupQ is different with only one file fixed by the Japanese team. The file is /usr/local/jwmconfig2/taskbar.
http://www.murga-linux.com/puppy/viewtopic.php?t=59058
I am not sure whether it affects on the issue DaveS reports.

[BarryK]

Video upgrade wizard recommended mesa-6.5.2-w5;installed it. The OpenGL portion of hardinfo/Display gives:

Code: Select all

Vendor	Unknown
Renderer	Unknown
Version	Unknown
Direct Rendering	No

Adding mesa-demos-7.6-w5 plus the dependency freeglut-2006-02-08-w5 (ignoring the xorg_base dependency which crashes X much like Shinobar's experience) gives

Code: Select all

Vendor	Tungsten Graphics, Inc
Renderer	Mesa DRI Intel(R) 845G 20061017 x86/MMX/SSE2
Version	1.3 Mesa 6.5.2
Direct Rendering	Yes

glxgears gives about 1800 frames per 5 sec, probably the highest I have seen on this laptop.

Well, freeglut is a fairly small package, so I am including it built-in to the next Wary. So all the openGL libraries, libGL, libGLU and libglut will be available built-in.

[BarryK]

Selecting Preferences then clicking OK kills Asunder.
See the log file attached

I have upgraded to Asunder 2.0 and it does not crash. Thanks for the feedback.

[glenco]

rerwin
Ok, I formatted sda4 and did a fresh FULL install off of my DVD/RW
This is how it went:
First boot:
-ran pupdial and it found ttySL0
-pupdial > CHOOSE > ERASE
-back to the CHOOSE window and tried ERASE again and the window went blank and froze
Killed pupdial!
-back into pupdial > CHOOSE >ERASE
-closed pupdial
(erased twice maybe three times??????)
-opened console and ran pmodemdiag
-renamed it to 1stBOOTdiag
shutdown & rebooted
2nd boot:
ran pupdial and it found ttySL0 right from the start
created an account
tried to connect and it went into that NO CARRIER loop again
ran pmodemdiag again and renamed it to 2ndBOOT diag
Installed modem_update_to_wary098-woof-1.pet
shutdown and rebooted
3rd boot
dialed out with no problem!!!

[Karl Godt]

# jwmconfig
0
23
sed: -e expression #1, char 2: invalid usage of line address 0

I had this vanishing jwm taskbar at Macpup-511 also , perhaps while running gtk-chtheme simultaneously.
Diffed /usr/local/jwmconfig2 with several puppies : no output ( = no difference )

[BarryK]

Thanks Karl Godt.
Found the version 1.6.0 makes Improper entry for full install with Wary-098 and in some case with Puppy 5.x. Please upgrade to the new v1.6.1.
http://www.murga-linux.com/puppy/viewtopic.php?t=51697
Your feed back is welcome.

Thanks, I have upgraded.

[BarryK]

# jwmconfig
0
23
sed: -e expression #1, char 2: invalid usage of line address 0

I had this vanishing jwm taskbar at Macpup-511 also , perhaps while running gtk-chtheme simultaneously.
Diffed /usr/local/jwmconfig2 with several puppies : no output ( = no difference )

I just tested jwmconfig and was able to run Apply GTK theme to JWM.

[BarryK]

01micko has requested the /usr/sbin/download_file script for testing.

Note to everyone else, this if for 01micko only. It is integrated into the Video Upgrade Wizard and PPM in the next release of Wary.

It can however be tested standalone, for example:

download_file http://distro.ibiblio.org/pub/linux/dis ... .14-w5.pet

...you just get minimal download windows. You only get more dialogs if something goes wrong, for example you specify a non-existent file:

download_file http://distro.ibiblio.org/pub/linux/dis ... odah-1.pet

Note that it only downloads to the current directory.

It returns various error messages (see comments in file) that the caller program can act upon.

[Sage]

SHS & ATI.
There is something really weird about 098 and ATI. Lupu seems to run on just about everything, new & old. Up to 094 runs on most kit including ATI, albeit with the known bugs, but at least they all boot.
There is something entirely different about 098 and ATI chipsets/graphics and I've no idea what it is, except it's serious and reproducible. Done all the usual checks and double checks, new d/ls, new burns, changed memory, changed boards, changed video cards (where these are discrete cards - even Vanta & Savage4 - the beasts of the genre work). Needs the gurus to crawl all over it.

[BarryK]

All kinds of error conditions can be tested with download_file.

For example, to test continuation of a partially downloaded file:

# touch alsa-utils_DOC-1.0.21-w5.pet

# download_file http://distro.ibiblio.org/pub/linux/dis ... .21-w5.pet

[Karl Godt]

just killed the taskbar at dpup-005 which has got gtk-chtheme !

dpup-009 hasn't got gtk-chtheme but lxappearance ( compiled somewhere in the long thread I think by iguleder ) and hitting the button
" Or, apply current GTK theme to JWM ' : the taskbar stays. It applies the Citrus Theme ( yellow ) because it's the first alphabetical in the row ( usr-share-themes )

[Karl Godt]

OK : It's not /usr/share/themes :

# find / -name *jwm*
/root/.jwmrc-previous
/root/.jwmrc
/root/.jwm
/root/.jwm/jwm_colors
/root/.jwm/jwmrc-personal
/root/.jwm/jwmrc-theme
/root/.jwm/jwmrc-personal2
/root/.jwm/themes/Aurora-jwmrc
/root/.jwm/themes/Black-vg-jwmrc
/root/.jwm/themes/Citrus-cut-jwmrc
/root/.jwm/themes/DarkGradient-jwmrc
/root/.jwm/themes/Flat-grey-jwmrc
/root/.jwm/themes/Gradient-blueish-jwmrc
/root/.jwm/themes/Gradient-brown-jwmrc
/root/.jwm/themes/Gradient-grey-jwmrc
/root/.jwm/themes/Original-jwmrc

I don't know why it applies Citrus-cut by ""default""

[Karl Godt]

NEXT :
Brute Force Method :
boot, menu > desktop > jwm configuration > Or, apply curr GTK th ...
dpup-001 : Taskbar away
dpup-002 : Taskbar away
dpup-003 : Taskbar away

.....
4. :
quirky 1.3 official : Taskbar stays with a lot of yellow
# jwmconfig
16
31

diff jwmconfig2 : no output

one of the diferences maybe : resolution 1024 against 1280

[Karl Godt]

Nope : changig the resolution to 1280x1024 doesn't reproduce the lost taskbar here , minor thing : uname -a complains that the
# uname -a
Linux puppypc 2.6.33.2 #1 SMP Mon Apr 5 12:55:04 Local time zone must be set--see zic m i686 GNU/Linux
#
but it is set and the time is local and date and time clicking the clock in the tray says : 17th Nov 13:32:31 so its a compiling thing for the kernel .
Another difference : german keyboard now and not an us-uk keyboard.