Puppy 4.1 Alpha 2

Please post any bugs you have found
Message
Author
JB4x4
Posts: 256
Joined: Sun 30 Jul 2006, 22:44
Contact:

#16 Post by JB4x4 »

@Barry

This is in regards to the "icon stacking" I found when inserting my USB key. I had this happening when I was running my screens native resolution of 1440x900. I switched it to 1024x768 and everything worked fine (I did have to move the sda and sdb icons as they were almost off of the desktop when I restarted X).

Hope this helps,
JB

User avatar
capoverde
Posts: 232
Joined: Wed 28 Jun 2006, 21:36
Location: Sanremo (Italy) with fine seaview

#17 Post by capoverde »

First test of Puppy 4.1 Alpha2 with pfix=ram.

OK with: boot, video (1280x1024 XVesa), USB mouse and PS2 keyboard, sound, ADSL cable network connection + browsing, audio and video streaming, wallpaper and clock setting, icon switching, mounting/unmounting drives - only hitch for now is the overlapping of drive icons, which now appear immediately with all other icons; just the topscreen message for first-time help, with associated double-bark, doesn't appear till after restarting X, as on previous "4" versions (not even after setting new wallpaper and icon type).
PMount still lists drive sda after drive sdb.

Asrock 775i65GV MoBo with Celeron 2800, 512MB RAM, 2 PATA 80GB drives.
Last edited by capoverde on Sat 07 Jun 2008, 12:49, edited 1 time in total.

NathanO
Posts: 210
Joined: Fri 23 Feb 2007, 00:03
Location: San Antonio, TX

#18 Post by NathanO »

Barry,

You say on your blog that you will not have full SCSI support, what do you mean?

My tower has both IDE and SCSI drives as well as a SCSI PCMCIA adapter that have worked thru 3.x series and 4.0.

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

#19 Post by tronkel »

Just tried 4.02alpha on my wife's HP Compaq nx6110 laptop, 500MB RAM, 2-3 years old.

Runs very fast, wireless fine, video restricted to 1024 x 768, sound OK. Running from live CD.

In common with other Puppy 4 versions, it occasionally exits to a black screen on shutdown. This has also happened on the Classmate PC as well.

Prehistoric has reported something similar in Chihuahua and has not as yet fully solved it. This problem has been around for a while in recent versions.

There is a 168MB sfs file loaded via unionfs. I wonder if the problem could be caused by the sfs file not being unmounted properly.

I'll try this with layerfs=aufs and see what that does.

edit: oops no - no aufs in 4.02Alpha

Anyone else had this problem?
Life is too short to spend it in front of a computer

User avatar
dogone
Posts: 202
Joined: Tue 22 Apr 2008, 02:53
Location: Arizona, USA

Alpha 2 drive management need work

#20 Post by dogone »

I've found drive handling in Alpha2 much less intuitive than in 4.00.

In 400:
- each partition ("drive") receives an icon.
- clicking an icon mounts the partition and opens ROX.
- closing ROX offers the unmount option (could be configurable)

In 402:
- drive icons represent physical devices, not logical ones (as in Windows).
- users have to "deal" with mounting/unmounting partitions.
- clicking a partition icon opens Pmount, not the partition.
- icon context menu lacks an unmount option

402 takes a big leap backwards by requiring the user to understand mounting and the distinction between drives and partitions. Windows addresses both (effectively, not intelligently).

Suggestions for 4.1 final:
- mimic 400's Hotplug behavior with enhancements.
- employ Event Manager.
- permit access to drives via a tray icon or icon drawer.
- permit user to disable unmount dialog at ROX close.
- permit user to assign persistent drive/partition labels.

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

#21 Post by tronkel »

Here are a few problems I found when testing 4.02alpha

On attempting to install to usb stick, universal installer fails with message that it cannot see the device - even although it was in fact plugged in. This was on the HP laptop nx6110. Also had a frozen black screen at shutdown.

Classmate PC can't see WLAN0. If the rt73 module is then manually loaded, then the wireless interface does become visible. The hotspot is then detected OK and DHCP seems to claim to bring the interface up with an IP address, but in fact the interface is dead. Black screen is intermittently a problem here too. This version of the rt73 usb driver appears to be the fault here. Not included in the the latest module set maybe in the pup-402.sfs?

Better news - my old compaq PC now has sound and Skype working. Also now retains the ALSA configuration after reboot.
Life is too short to spend it in front of a computer

JustGreg
Posts: 782
Joined: Tue 24 May 2005, 10:55
Location: Connecticut USA

#22 Post by JustGreg »

I have tried Puppy 4.1 alpha #2 on three systems:
1. my desktop, Cybernet ZPC-9000,
2. Intel Classmate (Generation 1),
3. 2goPC (Classmate Generation 2).

On the desktop, I did a manual frugal installation to ext2 partition on the hard disk by copying from the CD iso. Rebooted and Puppy 4.1 alpha started up. I checked and found the new rt73usb drivers were installed. I first tried the wifi connection scripts that I used to test Dingo-test with kernel 2.6.25 (http://www.murga-linux.com/puppy/viewtopic.php?t=29205). They worked and I was able to connect to my wpa2 encrypted network. I rebooted in Puppy 4.1 alpha2 to try network wizard. It did work after setting ap to wpa_supplicant, but, after defining the profile and trying to use, the network wizard said it was unable to connect. It did a network scan and said a network was found and one could get IP address. Auto DHCP did work and there was a network
connection to the wpa2 network.

I tried to do an installation to a USB flash drive to be used with the other two systems with the desktop machine installation. I plugged the USB flash drive in and saw its icon on the desktop. I check to see what files were present on it by mounting it. I deleted the files and unmounted it. When I used the universal installer, the universal installer could not find the USB flash drive. It was not possible to start the installation. I did a manual installation using Syslinux (no problems) and copying the needed files from the CDROM with Puppy 4.1 alpha #2. I now had my usb flash drive to test on the other two version of the Classmate. Both versions have the same wireless
and mouse hardware.

Using the 2goPC (Classmate Generation 2), I booted the usb flash drive. The normal setup occurred. When configuring xorg for video, the wizard presented many multiple conditions rather than just the normal 800x480. I selected 800x600 and xorg started normally. The screen display was fine. I check with "HardInfo" and found the display was set to 800x48. I had to manually load the rt73usb driver, the wifi hardware was not found initially. I tried the network wizard and connected with the same problem (wpa reported not found, but was available) as above. I was able to connect to google.com to check the network.

I used an external USB CD R/W drive to boot the Classmate. It booted and xorg setup correctly only had 800x480 resolution for the display. After manually loading the rt73usb wifi driver, I could not get the Classmate with either the network wizard or my scripts to connect to my wpa2 wifi network. I will have to do some more testing.
Enjoy life, Just Greg
Live Well, Laugh Often, Love Much

User avatar
capoverde
Posts: 232
Joined: Wed 28 Jun 2006, 21:36
Location: Sanremo (Italy) with fine seaview

#23 Post by capoverde »

Same problem as reported by tronkel and JustGreg: trying to install Puppy to an SD card with the Universal Installer, the card gets not detected (message "Nothing to choose. If a plugin device, did you plug it in?"). However, Pmount sees it and it can be mounted, read and written to, so a manual install should work.
Puppy 4 Dingo installed OK to an SD card with the Univ. Installer.

JustGreg
Posts: 782
Joined: Tue 24 May 2005, 10:55
Location: Connecticut USA

#24 Post by JustGreg »

I got a chance to test some more with the Intel Classmate. With Puppy alpha 2, the Intel Classmate can not connect to a network that is open (i.e. no encryption). I tried doing from the command line using iwconfig commands, no luck. It fails to associate, even with server ap address.

I also tried with Dingo test (kernel 2.6.25), which has worked in the past. With a wpa2 encrypted network, I could connect. Using wpa_cli in a console window, I notice there was a cycle of a disconnect event / authenicate / connect. The connection would stay for about a minute and the cycle would start again. The disconnect event said 4 way handshake failed. I do not have a clue of what to do next.

I hope this helps.
Enjoy life, Just Greg
Live Well, Laugh Often, Love Much

User avatar
dogone
Posts: 202
Joined: Tue 22 Apr 2008, 02:53
Location: Arizona, USA

Excessive CD spin-up at boot

#25 Post by dogone »

This one's for Barry.
My NEC CD/DVD drive spins up excessively while loading Puppy 400 (final). It spins up to a slightly lower rate while loading Alpha 2 (different kernel). The noise is annoying in both cases.

Here's the kicker. The drive spins slowly and silently while booting Puppy 4 NOP r1. Loading is just as fast but completely silent.

What's Grey's secret? Let's find out and implement it in the next alpha. Please?

User avatar
boscobearbank
Posts: 63
Joined: Thu 06 Apr 2006, 15:13
Location: MN

#26 Post by boscobearbank »

Averatec 3250 laptop (Ralink rt2500 wireless) glitch - wireless card was properly detected. Internet Connection Wizard claimed to have made a successful connection, but, in actuality, failed :cry: . I unloaded rt2500pci and its dependencies, reloaded them, then re-executed the Internet Connection Wizard. Success :D - I'm posting from it now.
Bosco Bearbank

User avatar
dogone
Posts: 202
Joined: Tue 22 Apr 2008, 02:53
Location: Arizona, USA

Puppy should default to saving user data

#27 Post by dogone »

I'd like some feedback on this one..
Puppy Event Manager v4 Alpha2 defaults to "never" (0) saving RAM to the pup_save file. I'm not comfortable with this default. Unless I'm mistaken, default to "never" means the user's data is not saved during a session until/unless she discovers this option in Event Manager or performs a manual save. I feel Puppy ought assume neither will happen and default to periodic saves.

User avatar
dogone
Posts: 202
Joined: Tue 22 Apr 2008, 02:53
Location: Arizona, USA

One Pmount preference cleared everytime Pmount restarts

#28 Post by dogone »

A small problem with Pmount in v4 Alpha2. The "Tick for Pmount to quit after mount/unmount" option is cleared when Pmount quits/restarts.

Everitt
Posts: 331
Joined: Tue 19 Dec 2006, 21:59
Location: Leeds,UK or Birmingham, UK

Re: Puppy should default to saving user data

#29 Post by Everitt »

dogone wrote:I'd like some feedback on this one..
Puppy Event Manager v4 Alpha2 defaults to "never" (0) saving RAM to the pup_save file. I'm not comfortable with this default. Unless I'm mistaken, default to "never" means the user's data is not saved during a session until/unless she discovers this option in Event Manager or performs a manual save. I feel Puppy ought assume neither will happen and default to periodic saves.
Data will always be saved at shutdown too.

User avatar
dogone
Posts: 202
Joined: Tue 22 Apr 2008, 02:53
Location: Arizona, USA

Puppy installer not detecting USB flash drives

#30 Post by dogone »

I can confirm Tronkel's observation. Puppy Universal Installer is not detecting attached flash sticks.

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

pup_bootd eval syntax error - SOLVED with suggested fix

#31 Post by rerwin »

When booting 4.0.2 on my home-built tower PC, the bootsysinit.log file shows that pup_bootd evaluated strings that caused it to fail. The messages are:

Code: Select all

/sbin/pup_bootd: eval: line 23: unexpected EOF while looking for matching `"'
/sbin/pup_bootd: eval: line 28: syntax error: unexpected end of file
Three sets of these messages appear.

To troubleshoot, I inserted some echo commands into pup_bootd:

Code: Select all

hotplug2stdout |
while read ONEUEVENT
do
 #graceful exit upon request (pup_eventd creates this file)...
 [ -f /tmp/pup_bootd_kill ] && exit
 #the uevent has space-separated fields...
 MAJOR="";MINOR="";DEVPATH="";SUBSYSTEM="";FIRMWARE=""
 if [ "$ONEUEVENT" != "TWOSECTIMEOUT" ];then
  EXTRAFIELDS="`echo "$ONEUEVENT" | cut -f 2-99 -d ' ' | tr ' ' '\n' | grep '='`" #grep is precaution.
  echo "" >> /tmp/bootsysinit.log  #rse
  echo "ONEUEVENT: $ONEUEVENT -"  >> /tmp/bootsysinit.log  #rse
  echo " EXTRAFIELDS: $EXTRAFIELDS" >> /tmp/bootsysinit.log  #rse
  eval "$EXTRAFIELDS" 2>> /tmp/bootsysinit.log   #rse
  ONEUEVENT="`echo -n "$ONEUEVENT" | cut -f 1 -d ' '`"
 fi
The new output shows the data that causes the eval error:

Code: Select all

ONEUEVENT: add@/devices/LNXSYSTM:00/LNXPWRBN:00/input/input2 ACTION=add DEVPATH=/devices/LNXSYSTM:00/LNXPWRBN:00/input/input2 SUBSYSTEM=input PRODUCT=19/0/2/0 NAME="Power Button (FF)" PHYS="LNXPWRBN/button/input0" EV==3 KEY==100000 0 0 0 MODALIAS=input:b0019v0000p0002e0000-e0,1,k74,ramlsfw SEQNUM=975 -
 EXTRAFIELDS: ACTION=add
DEVPATH=/devices/LNXSYSTM:00/LNXPWRBN:00/input/input2
SUBSYSTEM=input
PRODUCT=19/0/2/0
NAME="Power  < < < < < < < < < < < < < < < < < < The Problem!
PHYS="LNXPWRBN/button/input0"
EV==3
KEY==100000
MODALIAS=input:b0019v0000p0002e0000-e0,1,k74,ramlsfw
SEQNUM=975
/sbin/pup_bootd: eval: line 26: unexpected EOF while looking for matching `"'
/sbin/pup_bootd: eval: line 31: syntax error: unexpected end of file

ONEUEVENT: add@/devices/LNXSYSTM:00/device:00/PNP0C0C:00/input/input3 ACTION=add DEVPATH=/devices/LNXSYSTM:00/device:00/PNP0C0C:00/input/input3 SUBSYSTEM=input PRODUCT=19/0/1/0 NAME="Power Button (CM)" PHYS="PNP0C0C/button/input0" EV==3 KEY==100000 0 0 0 MODALIAS=input:b0019v0000p0001e0000-e0,1,k74,ramlsfw SEQNUM=976 -
 EXTRAFIELDS: ACTION=add
DEVPATH=/devices/LNXSYSTM:00/device:00/PNP0C0C:00/input/input3
SUBSYSTEM=input
PRODUCT=19/0/1/0
NAME="Power  < < < < < < < < < < < < < < < < < < The Problem!
PHYS="PNP0C0C/button/input0"
EV==3
KEY==100000
MODALIAS=input:b0019v0000p0001e0000-e0,1,k74,ramlsfw
SEQNUM=976
/sbin/pup_bootd: eval: line 26: unexpected EOF while looking for matching `"'
/sbin/pup_bootd: eval: line 31: syntax error: unexpected end of file

ONEUEVENT: add@/devices/LNXSYSTM:00/device:00/PNP0C0E:00/input/input4 ACTION=add DEVPATH=/devices/LNXSYSTM:00/device:00/PNP0C0E:00/input/input4 SUBSYSTEM=input PRODUCT=19/0/3/0 NAME="Sleep Button (CM)" PHYS="PNP0C0E/button/input0" EV==3 KEY==4000 0 0 0 0 MODALIAS=input:b0019v0000p0003e0000-e0,1,k8E,ramlsfw SEQNUM=977 -
 EXTRAFIELDS: ACTION=add
DEVPATH=/devices/LNXSYSTM:00/device:00/PNP0C0E:00/input/input4
SUBSYSTEM=input
PRODUCT=19/0/3/0
NAME="Sleep  < < < < < < < < < < < < < < < < < < The Problem!
PHYS="PNP0C0E/button/input0"
EV==3
KEY==4000
MODALIAS=input:b0019v0000p0003e0000-e0,1,k8E,ramlsfw
SEQNUM=977
/sbin/pup_bootd: eval: line 26: unexpected EOF while looking for matching `"'
/sbin/pup_bootd: eval: line 31: syntax error: unexpected end of file
(The line number difference reflects the insertion of the added echo lines.)

EDIT 6/9/08: The problem lies in the /sbin/pup_bootd logic starting at line 15:

Code: Select all

 MAJOR="";MINOR="";DEVPATH="";SUBSYSTEM="";FIRMWARE=""
 if [ "$ONEUEVENT" != "TWOSECTIMEOUT" ];then
  EXTRAFIELDS="`echo "$ONEUEVENT" | cut -f 2-99 -d ' ' | tr ' ' '\n' | grep '='`" #grep is precaution.
The last line separates space-delimited assignment statements into individual lines, but can be tricked by embedded spaces in a quoted string.

Considering that line 15 initializes only five variables, I assume only they are important. Changing the grep in the last line to find only those five variables avoids the variables that might contain embedded space characters. Consider the following as a replacement for that line, line 17:

Code: Select all

  EXTRAFIELDS="`echo "$ONEUEVENT" | cut -f 2-99 -d ' ' | tr ' ' '\n' | grep -E '^MAJOR=|^MINOR=|^DEVPATH=|^SUBSYSTEM=|^FIRMWARE='`" #v4.0.3 retain only needed variable assignments.
Are three of the variables actually needed? I see only DEVPATH and FIRMWARE being used within the while-loop.
EDIT END
Richard
Last edited by rerwin on Mon 09 Jun 2008, 21:15, edited 2 times in total.

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

4.1alpha2's handling of partitions is awkward

#32 Post by pakt »

dogone wrote:I've found drive handling in Alpha2 much less intuitive than in 4.00.

402 takes a big leap backwards by requiring the user to understand mounting and the distinction between drives and partitions. Windows addresses both (effectively, not intelligently).
I agree with dogone that drive handling is not intuitive in 4.1a2 - in fact I find it quite awkward. Why should the user be forced to manipulate Pmount when a drive icon on the desktop is clicked? It should work like in HotPup where the drive is mounted and a ROX window opens. When the window is closed, ask if the partition should be unmounted - nice and simple!

If having many drive partitions causes a problem with too many icons being displayed, have a setting (as in HotPup) where just the removable partition icons are shown. Then Pmount can be used to mount/unmount those fixed partitions that don't have an icon. All in all, I think Dougal got it right with HotPup. I still find it the simplest way of dealing with partitions ;-)
Methinks Raspberry Pi were ideal for runnin' Puppy Linux

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

#33 Post by Sage »

The number of folks using more than three or four partitions is probably limited to <0.1% of punters reading this Forum? [Nine million users of forty six partitions take one step forward.]
One sympathises with Our Dear Leader who works so close to the bleeding edge, but I do get the impression from his blogs that he values this kind of feedback from users as well as fellow developers. All too easy in academia to get carried away with the intellectual minutia!
All of which is to agree that Puppy is tending to follow the guru trail of greater and greater complexity, coding sophistication, bloat, w.h.y., without reference back to core user requirements, nay understanding, - and I don't mean pandering to WiFi, laptops or even scsi(!) which, in the case of the former two are unnecessary commercial drivers and better outlawed?! In the present case, the issue stretches back all the way to Snr. Torvalds with his new drives designators - itself a major bone of contention with his nearest and dearest.
Tip of the iceberg, pakt. And your nation knows all about icebergs.
K.I.S.S. me, Hardy - although even that utterance is contentious.

User avatar
boscobearbank
Posts: 63
Joined: Thu 06 Apr 2006, 15:13
Location: MN

#34 Post by boscobearbank »

Sage wrote:The number of folks using more than three or four partitions is probably limited to <0.1% of punters reading this Forum?
Count me in that < 0.1%. On the other hand, I have yet to determine a solution to the drive mounting/display problem that I consider ideal, I'll take whatever Barry's giving.
Bosco Bearbank

ICPUG
Posts: 1308
Joined: Mon 25 Jul 2005, 00:09
Location: UK

#35 Post by ICPUG »

When Hotpup first appeared it got a lot of rave comments.

I first saw it in one of the community 3.02 versions. I thought it was brilliant too. Why?

- it is intuitive
- takes away the mystery of mounting and unmounting (helpful for Windows refugees) without removing the capability to keep unmounted for security purposes (helpful for those who like a bit of security)
- It worked like virtually every other similar concept I had come across. This was one area where Puppy was deficient compared to Knoppix. One of my tests of a Live CD is how easy is it to access my hard drive. Hotpup improved Puppy usability no end.

I thought Hotpup was the killer app that would distinguish Dingo from all that went before.

Then, when Dingo comes out, Hotpup is not on by default? Odd decision, presumably in deference to those with many partitions.

Now Barry decides to use a Hotpup alternative centred on drives rather than partitions.

I read the developer blog and the reason for this decision is a bit beyond me. I guess it is kinda wrapped up with this event recognition thingy? Certainly I could not see any drawback of Hotpup except for the users with large numbers of partitions. If it worked and had rave reviews why is it 'fixed' in the next upgrade?

My feeling, as a user, is the new approach is not as as intuitive as Hotpup.

Occasionally, Puppy has these conflicts as to whether it is a distro for developers/super users or a distro for users. I think that people with many partitions are not your average user. They are more savvy and can turn Hotpup off and mount manually with Pmount when required.

As a user I come down in favour of the Hotpup aproach.

ICPUG

Post Reply