Wary 0.9.8 (098) feedback

Please post any bugs you have found
Message
Author
User avatar
Béèm
Posts: 11763
Joined: Wed 22 Nov 2006, 00:47
Location: Brussels IBM Thinkpad R40, 256MB, 20GB, WiFi ipw2100. Frugal Lin'N'Win

#41 Post by Béèm »

Fresh frugal install (Lin'N'Win)
I downloaded the pidgin 2.7.5 pet from the wary package library and transformed it to a sfs.
Pidgin is working now, but the pidgin icon doesn't display in the system tray. Instead I have a kinda document icon with a small red cross.
Last edited by Béèm on Sun 14 Nov 2010, 17:52, edited 1 time in total.
Time savers:
Find packages in a snap and install using Puppy Package Manager (Menu).
[url=http://puppylinux.org/wikka/HomePage]Consult Wikka[/url]
Use peppyy's [url=http://wellminded.com/puppy/pupsearch.html]puppysearch[/url]

User avatar
Karl Godt
Posts: 4199
Joined: Sun 20 Jun 2010, 13:52
Location: Kiel,Germany

#42 Post by Karl Godt »

OK : One pup_event_frontend_d has to run in the background for hotplugging events like inserting an USB drive.

Thanks to zigbert :
/sbin/pup_event_frontend_d :
line 90+ :

Code: Select all

#parameters
VERSIONP1="`cat /etc/DISTRO_SPECS | grep DISTRO_NAME | cut -f 2 -d \'`"
VERSIONP2="`cat /etc/DISTRO_SPECS | grep DISTRO_VERSION | cut -f 2 -d "="`"
while [ $# != 0 ]; do
	I=1
	while [ $I -lt `echo $# | wc -c` ]; do
		case $1 in
			-r|--recover-drive-icon) RECOVERDRIVEICON=1;;
			-v|--version) echo pup_event_frontend_d $VERSIONP1" "$VERSIONP2; exit;;
			-h|--help)
echo 'Usage: pup_event_frontend_d [-OPTION]

Options
  -h          Show this message
  -v          Show pup_event_frontend_d version number 
  -r          Recovering of driveicons only, no deamon mode'; exit;;
		esac
		shift
		I=$(($I+1))
	done
done

#v424 experiment, automatic unmounting...
line 804+ :

Code: Select all

if test "`echo $RECOVERDRIVEICON`" == "1"; then
RECOVERDRIVEICON=0
break
fi
done #2 second loop.


###END###
/usr/sbin/eventmanager :
line 315+ :
#v403 /sbin/clean_desk_icons can read this, then wipe all current icons...
[ "$ICONPARTITIONS" != "$NEWICONPARTITIONS" ] && echo "ICONPARTITIONS" > /tmp/pup_event_icon_change_flag
[ "$ICONDESK" != "$NEWICONDESK" ] && echo "ICONDESK" > /tmp/pup_event_icon_change_flag

Code: Select all

[ "$NEWICONWIPE" = "true" ] && echo "ICONWIPE" > /tmp/pup_event_icon_change_flag ; exec `pup_event_frontend_d -r` &#v411
/sbin/clean_desk_icons seems not to run properly : didn't recover the drive icon but added a second USB icon on the floppy icon.

Sage
Posts: 5536
Joined: Tue 04 Oct 2005, 08:34
Location: GB

#43 Post by Sage »

The good news was that 096 got me to orgwiz, but then couldn't find a compatible card/res/driver combo that worked. The bad, bad news is that 098 won't even boot beyond the 'switch' before Col. Panic shows his ugly mug.
For YT, even worse news is that the only PCI-E cards I possess are nV, which are incompatible with this ATI-based 939 board (MS7184/7093) so I have to use the onboard graphics, which was also intended to save some amps by avoiding those grisly all-singing-dancing video cards. Absolute bummer as I've built the case into my new lounge setup to drive the tele. Back to the drawing board for us all...

User avatar
tronkel
Posts: 1116
Joined: Fri 30 Sep 2005, 11:27
Location: Vienna Austria
Contact:

#44 Post by tronkel »

@sage

Try booting from live CD using the kernel cheat code "puppy pfix=ram" and see if that will boot your system.

If it boots up successfully like this, then something has gone amiss inside your pupsave - assuming you have one. Sometimes upgrading a pupsave file to a bumped Puppy version can cause this.
Life is too short to spend it in front of a computer

Sage
Posts: 5536
Joined: Tue 04 Oct 2005, 08:34
Location: GB

#45 Post by Sage »

This is the liveCD, tronk - I always use pfix=ram in these situations. There is a swap partition available, but nothing installed that's relevant (a Slack-64bit on disk).
The good news is that Lupu boots up flawlessly with minimum intervention, on-board graphics, live-CD, nothing saved or installed (except Slack).
Sorry, chaps, this is a home-grown Wary feature. Guess The Boss will have some comments when he awakes tomorrow.

User avatar
tronkel
Posts: 1116
Joined: Fri 30 Sep 2005, 11:27
Location: Vienna Austria
Contact:

#46 Post by tronkel »

@sage
Hmmm, that's a bit odd to say the least.
Is there perhaps another wary_xxx.sfs already on the Slack partition that is interfering with the boot sequence? Long shot, but I would check it out anyway.
Life is too short to spend it in front of a computer

Sage
Posts: 5536
Joined: Tue 04 Oct 2005, 08:34
Location: GB

#47 Post by Sage »

Absolutely not. I don't believe in frugals, anyway. But the effect is reproducible. This is a new feature of 098. Shame, because it functions better than ever on most other kit. I am, however, totally confident in this piece of hardware - everything else is rock solid on it.

User avatar
gjuhasz
Posts: 422
Joined: Mon 29 Sep 2008, 14:28

Asunder crashes

#48 Post by gjuhasz »

Selecting Preferences then clicking OK kills Asunder.
See the log file attached
Attachments
asundercrash.log.gz
(2.2 KiB) Downloaded 661 times

User avatar
BarryK
Puppy Master
Posts: 9392
Joined: Mon 09 May 2005, 09:23
Location: Perth, Western Australia
Contact:

#49 Post by BarryK »

[quote="pakt"]Barry, the problem with the savefile/xorgwizard menu extending below the edge of the screen is still present in wary-098.

Your mod in rc.country, “FONTMAP='’ #101112“, doesn’t blank the variable for some strange reason. The contents of /etc/fontmap is still “lat1-12.psfu
[url]https://bkhome.org/news/[/url]

glenco
Posts: 37
Joined: Fri 14 May 2010, 13:19
Location: Milwaukee

NO CARRIER

#50 Post by glenco »

Bummer :(
The same as 096, 098 finds my Pctel internal modem but finds NO CARRIER when it tries to dial.
-> WvDial: Internet dialer version 1.53
--> Initializing modem.
--> Sending: ATZ
ATZ
OK
--> Sending: ATQ0 V1 E1 S0=0 &C1 &D2 +FCLASS=0
ATQ0 V1 E1 S0=0 &C1 &D2 +FCLASS=0
OK
--> Modem initialized.
--> Sending: ATX3DT9083735
--> Waiting for carrier.
ATX3DT9083735
NO CARRIER
ERROR
--> No Carrier! Trying again.
--> Sending: ATX3DT9083735
--> Waiting for carrier.
ATX3DT9083735
NO CARRIER
ERROR
--> No Carrier! Trying again.
--> Sending: ATX3DT9083735
--> Waiting for carrier.
ATX3DT9083735
NO CARRIER
ERROR
--> No Carrier! Trying again.
--> Sending: ATX3DT9083735
--> Waiting for carrier.
ATX3DT9083735
NO CARRIER
ERROR
--> No Carrier! Trying again.

User avatar
BarryK
Puppy Master
Posts: 9392
Joined: Mon 09 May 2005, 09:23
Location: Perth, Western Australia
Contact:

#51 Post by BarryK »

[quote="pakt"]Barry, the problem with the savefile/xorgwizard menu extending below the edge of the screen is still present in wary-098.

Your mod in rc.country, “FONTMAP='’ #101112“, doesn’t blank the variable for some strange reason. The contents of /etc/fontmap is still “lat1-12.psfu
[url]https://bkhome.org/news/[/url]

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

#52 Post by rerwin »

glenco,
Thanks for the diag file. The mention of "PCTel" threw me off for a while, expecting the problem to be related to the pctel driver, which it is not.

Your internal modem uses a chip that has three drivers claiming it. It requires a second boot, after puppy reports no modem found. I have implemented logic to sort that out, but I think it is messed up in the transition to the modprobe.d files. I will work up a fix and post it soon, for you to test. I hope you can stick with me until we get it right, for the final.
Richard

glenco
Posts: 37
Joined: Fri 14 May 2010, 13:19
Location: Milwaukee

#53 Post by glenco »

rerwin
Thanks :) I should have mentioned that this is a DELL oem PCTEL modem in this laptop.
Windows reports this modem as:
[ Modems / PCTEL 2304WT V.92 MDC Modem ]



Device Properties:

Driver Description PCTEL 2304WT V.92 MDC Modem

Driver Date 11/21/2002

Driver Version 2.50.43.0

Driver Provider PCtel

INF File oem8.inf

Hardware ID PCI\VEN_8086&DEV_2486&SUBSYS_4C21134D&REV_02

Location Information PCI bus 0, device 31, function 6

PCI Device Intel 82801CAM ICH3-M - AC'97 Modem Controller [B-1]



Device Resources:

IRQ 05

Port D400-D4FF

Port DC00-DC7F

And Puppy 4.21 only uses the snd_intel8x0 & snd_intel8x0m modules??
No slamr is loaded at all.

User avatar
James C
Posts: 6618
Joined: Thu 26 Mar 2009, 05:12
Location: Kentucky

#54 Post by James C »

Another fresh manual frugal install of Wary 098,this time on my P4 test box.Initial boot had working sound and internet as well as the correct screen resolution.

The video upgrade wizard will not download on this install either....I just went to the repo and downloaded the Mesa package.

After clicking the button to download the Mesa package it's only about 1 or 2 seconds before the install dialog appears...... not enough time to download the package.

Anyway,



VIDEO REPORT: Wary Puppy, version 098

Chip description:
Intel Corporation 82865G Integrated Graphics Controller

Driver used by Xorg:
intel

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 2.80GHz
Memory : 512MB (241MB used)
Operating System : Puppy Linux 0.09
User Name : root (root)
Date/Time : Sun 14 Nov 2010 09:23:59 PM CST
-Display-
Resolution : 1024x768 pixels
OpenGL Renderer : Unknown
X11 Vendor : The X.Org Foundation
-Multimedia-
Audio Adapter : ICH4 - Intel ICH5

Sage
Posts: 5536
Joined: Tue 04 Oct 2005, 08:34
Location: GB

#55 Post by Sage »

New day, double check.
Yup, 098 doesn't work on this system.
All previous and all other distros work just fine (at least up to res/etc selection).
Clearly there is an issue here.

User avatar
pakt
Posts: 1157
Joined: Sat 04 Jun 2005, 16:54
Location: Sweden

#56 Post by pakt »

BarryK wrote:Sometimes I'm really dumb. I put that FONTMAP='' in the wrong place in /etc/rc.d/rc.country. Fixed.
Well, you’re certainly smarter than me. I can’t see why your previous fix doesn’t work (but that's probably why you're the Puppy guru and I just manage to find a fault or two :wink: ).
Methinks Raspberry Pi were ideal for runnin' Puppy Linux

User avatar
shinobar
Posts: 2672
Joined: Thu 28 May 2009, 09:26
Location: Japan
Contact:

Grub4DosConfig BUG

#57 Post by shinobar »

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.
Downloads for Puppy Linux [url]http://shino.pos.to/linux/downloads.html[/url]

User avatar
Sit Heel Speak
Posts: 2595
Joined: Fri 31 Mar 2006, 03:22
Location: downwind

Has anyone made ATi Catalyst work under Wary?

#58 Post by Sit Heel Speak »

Has anyone got ati_fglrx-10.10-k2.6.31.14-w5.pet (in the PPM repository) working? I suspect that it is too new for Wary's Xorg 7.3+ and I'll have to compile fglrx 9.8 if I want Catalyst.

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

"Pctel" built-in modem handling fixed

#59 Post by rerwin »

Thanks to glenco for getting me to focus on his built-in modem problem. I ended up cleaning up my code for that. A quick fix for that is the following edit command:

Code: Select all

sed -i -e 's/slamr:intel8x0m/slamr:snd_intel8x0m/' /etc/rc.d/MODULESCONFIG
If that alone does not help, the attached update package should.

The package also cleans up some 'loose ends":

1. start_cpu_freq - Avoids error message on old PCs without a BIOS display capability.

2. 10alsa - Adds "-s" parameter to grep of alsa_cards.conf, to avoid error message if no such file present on old PCs.

3. MODULESCONFIG - Corrects preference for slamr module.

4. rc.services - Inserted 1-second wait before running any init script, to allow/ensure that loading of all modules gets initiated. Deleted the 3-second "gross workaround" for loading to complete, which is no longer necessary because the more precise 'protect' function is used to determine when or whether a module completes loading.

5. 51-modprobe-usbserial.rules - Added detection of Huawei E160E, which is not yet in the usb_modeswitch data base or rules. (muggins)

6. Firmware tarballs for HSF and HCFPCI drivers - Corrected scripts and conf-file names to ensure correct ordering of particular modprobe.d statements. This plus the preference patch restores built-in modem functioning.

7. modemprobe_erase - Synchonized to new HSF/HCFPCI conf-file names, per tarballs.

8. pupdial - Updated label to "Phone/Access number". (nap44)

UPDATE 11/16/2010: Re-uploaded package as "-2" because it now fixes the HSF/HCF modem lockup when attempting to connect. Be sure to use pupdial's CHOOSE > ERASE after installation, to complete the correction with the next boot-up.

UPDATE 11/17/2010: Re-uploaded package "-2" after download number 3, to update pupdial. It now disables the "connect" buttons if the modem is ttySHSF? or ttySHCF? and the corresponding "nvm" directory is missing. Although the user could do nothing about it, pupdial prevents a connection attempt that would result in a locked-up PC.

UPDATE 11/17/2010: Re-uploaded the package as "-3" because it adds a change to the usb-modem mode switcher, so that it does not indicate completion if it encounters temporary conditions that prevent it from performing. A subsequent invocation might find those conditions no longer true, permitting mode switching.

UPDATE 11/18/2010: Uploaded "-4", which adds an attempt to prevent missed sound cards on some boot-ups. It also has some trace logging for troubleshooting the "ALSA-2barks" issue.

UPDATE 11/19/2010: Uploaded "-5", which fixes firmware timeout when "module hack" (for rt8139cp) is performed.

UPDATE 11/23/2010: Uploaded "-6-delta" to be installed after the "-5" package. This addresses the CD-boot Welcome-woofwoof problem reported by Barry. The underlying problem was with the busybox-depmod-generated modules.dep file, which does not screen out the ".koHIDE" suffix. Changing the suffix to ".HIDE" allowed the agrmodem and intelmodem variants to appear as designed. And the welcoming woofwoof is restored!

UPDATE 11/24/2010: Added the pupdial HCF fix to the "6-delta" package after download number 3. Also added a modeswitch rule for Barry's and muggins' Vodaphone/Huawei E160E wireless modems. (The rule for loading the driver is in "-5".)

UPDATE 11/25/2010: I hope this is my last replacement of "-6". I have fixed 3 items:
1. Removed the rules for E160E modem mode-switching and driver loading.
2. Changed the usbserial (3G modem) initialization script to choose the higher-numbered ttyUSB port if two (or more) are found.
3. Improved pusbmodem's implementation of retaining information on mode switches, so a device can be loaded even if its mode is not switched.

UPDATE 11/26/2010: To keep my promise to not re-upload "-6-delta". I am attaching a supplemental fix package, "-6-delta-pusbmodeswx", to make the soon-to-be-obsolete (I hope) pusbmodeswitch script compatible with its zzz-package replacement. This version resumes using the "standard" modeswitch database directory, but also handles a database tarball if present.
Richard
Attachments
modem_update_to_wary098-woof-6-delta-pusbmodeswx.pet
pusbmodeswitch updated to restore using the /etc/usb_modeswitch directory.
(3.5 KiB) Downloaded 545 times
modem_update_to_wary098-woof-6-delta.pet
Supplement to the "-5" package.
Corrects Agere/Intel variant "hide" names to avoid busybox depmod corruption.
Removes ALSA debug logging, but retains "event-gated" pacing of startup.
(30.57 KiB) Downloaded 521 times
Last edited by rerwin on Fri 26 Nov 2010, 17:47, edited 14 times in total.

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

Modem driver misbehavior - slusb, slamr, hsfmodem

#60 Post by rerwin »

Now the bad news. The above drivers fail as follows:

1. hsfmodem (hsfpcibasic2) - locks up the computer while starting a connection. This is the hybrid version, so I plan to compile and try the "full" version to isolate the problem, somewhat.
UPDATE 11/17/2010: Solved! The driver is OK, but a directory it needs was in the wrong place. Fixed by package here: http://www.murga-linux.com/puppy/viewto ... 639#467639

2. slamr - For a built-in (8086:2486) modem, when connecting, retries on "no carrier" response indefinitely!

3. slusb - Spews trace messages and fails to set up its port, but does not kill the PC. I attach the message log as evidence, in case someone can make sense of it.

EDIT: I think support for the SL USB modem has stopped, due to a proprietary code issue. The recent SL driver packages do not generate the slusb module.

So far, my DGC/ACM USB modem and PCtel, ESS and Lucent PCI modems work. I assume the Ageres and Intels still do, but have not verified with 098.
Richard
Attachments
messages-SL_USB_modem.tar.gz
Message log from boot-up with only a SmartLink USB modem plugged in.
(7.5 KiB) Downloaded 530 times
Last edited by rerwin on Wed 24 Nov 2010, 04:22, edited 2 times in total.

Post Reply